pacemaker  2.0.5-ba59be712
Scalable High-Availability cluster resource manager
unpack.c
Go to the documentation of this file.
1 /*
2  * Copyright 2004-2020 the Pacemaker project contributors
3  *
4  * The version control history for this file may have further details.
5  *
6  * This source code is licensed under the GNU Lesser General Public License
7  * version 2.1 or later (LGPLv2.1+) WITHOUT ANY WARRANTY.
8  */
9 
10 #include <crm_internal.h>
11 
12 #include <stdio.h>
13 #include <string.h>
14 #include <glib.h>
15 #include <time.h>
16 
17 #include <crm/crm.h>
18 #include <crm/services.h>
19 #include <crm/msg_xml.h>
20 #include <crm/common/xml.h>
21 #include <crm/common/xml_internal.h>
22 
23 #include <crm/common/util.h>
24 #include <crm/pengine/rules.h>
25 #include <crm/pengine/internal.h>
26 #include <crm/common/iso8601_internal.h>
27 #include <pe_status_private.h>
28 
29 CRM_TRACE_INIT_DATA(pe_status);
30 
31 /* This uses pcmk__set_flags_as()/pcmk__clear_flags_as() directly rather than
32  * use pe__set_working_set_flags()/pe__clear_working_set_flags() so that the
33  * flag is stringified more readably in log messages.
34  */
35 #define set_config_flag(data_set, option, flag) do { \
36  const char *scf_value = pe_pref((data_set)->config_hash, (option)); \
37  if (scf_value != NULL) { \
38  if (crm_is_true(scf_value)) { \
39  (data_set)->flags = pcmk__set_flags_as(__func__, __LINE__, \
40  LOG_TRACE, "Working set", \
41  crm_system_name, (data_set)->flags, \
42  (flag), #flag); \
43  } else { \
44  (data_set)->flags = pcmk__clear_flags_as(__func__, __LINE__,\
45  LOG_TRACE, "Working set", \
46  crm_system_name, (data_set)->flags, \
47  (flag), #flag); \
48  } \
49  } \
50  } while(0)
51 
52 static void unpack_rsc_op(pe_resource_t *rsc, pe_node_t *node, xmlNode *xml_op,
53  xmlNode **last_failure,
54  enum action_fail_response *failed,
55  pe_working_set_t *data_set);
56 static void determine_remote_online_status(pe_working_set_t *data_set,
57  pe_node_t *this_node);
58 static void add_node_attrs(xmlNode *attrs, pe_node_t *node, bool overwrite,
59  pe_working_set_t *data_set);
60 static void determine_online_status(xmlNode *node_state, pe_node_t *this_node,
61  pe_working_set_t *data_set);
62 
63 static void unpack_lrm_resources(pe_node_t *node, xmlNode *lrm_state,
64  pe_working_set_t *data_set);
65 
66 
67 // Bitmask for warnings we only want to print once
68 uint32_t pe_wo = 0;
69 
70 static gboolean
71 is_dangling_guest_node(pe_node_t *node)
72 {
73  /* we are looking for a remote-node that was supposed to be mapped to a
74  * container resource, but all traces of that container have disappeared
75  * from both the config and the status section. */
76  if (pe__is_guest_or_remote_node(node) &&
77  node->details->remote_rsc &&
78  node->details->remote_rsc->container == NULL &&
79  pcmk_is_set(node->details->remote_rsc->flags,
80  pe_rsc_orphan_container_filler)) {
81  return TRUE;
82  }
83 
84  return FALSE;
85 }
86 
87 
96 void
97 pe_fence_node(pe_working_set_t * data_set, pe_node_t * node,
98  const char *reason, bool priority_delay)
99 {
100  CRM_CHECK(node, return);
101 
102  /* A guest node is fenced by marking its container as failed */
103  if (pe__is_guest_node(node)) {
104  pe_resource_t *rsc = node->details->remote_rsc->container;
105 
106  if (!pcmk_is_set(rsc->flags, pe_rsc_failed)) {
107  if (!pcmk_is_set(rsc->flags, pe_rsc_managed)) {
108  crm_notice("Not fencing guest node %s "
109  "(otherwise would because %s): "
110  "its guest resource %s is unmanaged",
111  node->details->uname, reason, rsc->id);
112  } else {
113  crm_warn("Guest node %s will be fenced "
114  "(by recovering its guest resource %s): %s",
115  node->details->uname, rsc->id, reason);
116 
117  /* We don't mark the node as unclean because that would prevent the
118  * node from running resources. We want to allow it to run resources
119  * in this transition if the recovery succeeds.
120  */
121  node->details->remote_requires_reset = TRUE;
122  pe__set_resource_flags(rsc, pe_rsc_failed|pe_rsc_stop);
123  }
124  }
125 
126  } else if (is_dangling_guest_node(node)) {
127  crm_info("Cleaning up dangling connection for guest node %s: "
128  "fencing was already done because %s, "
129  "and guest resource no longer exists",
130  node->details->uname, reason);
131  pe__set_resource_flags(node->details->remote_rsc,
132  pe_rsc_failed|pe_rsc_stop);
133 
134  } else if (pe__is_remote_node(node)) {
135  pe_resource_t *rsc = node->details->remote_rsc;
136 
137  if ((rsc != NULL) && !pcmk_is_set(rsc->flags, pe_rsc_managed)) {
138  crm_notice("Not fencing remote node %s "
139  "(otherwise would because %s): connection is unmanaged",
140  node->details->uname, reason);
141  } else if(node->details->remote_requires_reset == FALSE) {
142  node->details->remote_requires_reset = TRUE;
143  crm_warn("Remote node %s %s: %s",
144  node->details->uname,
145  pe_can_fence(data_set, node)? "will be fenced" : "is unclean",
146  reason);
147  }
148  node->details->unclean = TRUE;
149  // No need to apply `priority-fencing-delay` for remote nodes
150  pe_fence_op(node, NULL, TRUE, reason, FALSE, data_set);
151 
152  } else if (node->details->unclean) {
153  crm_trace("Cluster node %s %s because %s",
154  node->details->uname,
155  pe_can_fence(data_set, node)? "would also be fenced" : "also is unclean",
156  reason);
157 
158  } else {
159  crm_warn("Cluster node %s %s: %s",
160  node->details->uname,
161  pe_can_fence(data_set, node)? "will be fenced" : "is unclean",
162  reason);
163  node->details->unclean = TRUE;
164  pe_fence_op(node, NULL, TRUE, reason, priority_delay, data_set);
165  }
166 }
167 
168 // @TODO xpaths can't handle templates, rules, or id-refs
169 
170 // nvpair with provides or requires set to unfencing
171 #define XPATH_UNFENCING_NVPAIR XML_CIB_TAG_NVPAIR \
172  "[(@" XML_NVPAIR_ATTR_NAME "='" XML_RSC_ATTR_PROVIDES "'" \
173  "or @" XML_NVPAIR_ATTR_NAME "='" XML_RSC_ATTR_REQUIRES "') " \
174  "and @" XML_NVPAIR_ATTR_VALUE "='unfencing']"
175 
176 // unfencing in rsc_defaults or any resource
177 #define XPATH_ENABLE_UNFENCING \
178  "/" XML_TAG_CIB "/" XML_CIB_TAG_CONFIGURATION "/" XML_CIB_TAG_RESOURCES \
179  "//" XML_TAG_META_SETS "/" XPATH_UNFENCING_NVPAIR \
180  "|/" XML_TAG_CIB "/" XML_CIB_TAG_CONFIGURATION "/" XML_CIB_TAG_RSCCONFIG \
181  "/" XML_TAG_META_SETS "/" XPATH_UNFENCING_NVPAIR
182 
183 static void
184 set_if_xpath(uint64_t flag, const char *xpath, pe_working_set_t *data_set)
185 {
186  xmlXPathObjectPtr result = NULL;
187 
188  if (!pcmk_is_set(data_set->flags, flag)) {
189  result = xpath_search(data_set->input, xpath);
190  if (result && (numXpathResults(result) > 0)) {
191  pe__set_working_set_flags(data_set, flag);
192  }
193  freeXpathObject(result);
194  }
195 }
196 
197 gboolean
198 unpack_config(xmlNode * config, pe_working_set_t * data_set)
199 {
200  const char *value = NULL;
201  GHashTable *config_hash = crm_str_table_new();
202 
203  pe_rule_eval_data_t rule_data = {
204  .node_hash = NULL,
205  .role = RSC_ROLE_UNKNOWN,
206  .now = data_set->now,
207  .match_data = NULL,
208  .rsc_data = NULL,
209  .op_data = NULL
210  };
211 
212  data_set->config_hash = config_hash;
213 
214  pe__unpack_dataset_nvpairs(config, XML_CIB_TAG_PROPSET, &rule_data, config_hash,
215  CIB_OPTIONS_FIRST, FALSE, data_set);
216 
217  verify_pe_options(data_set->config_hash);
218 
219  set_config_flag(data_set, "enable-startup-probes", pe_flag_startup_probes);
220  if (!pcmk_is_set(data_set->flags, pe_flag_startup_probes)) {
221  crm_info("Startup probes: disabled (dangerous)");
222  }
223 
224  value = pe_pref(data_set->config_hash, XML_ATTR_HAVE_WATCHDOG);
225  if (value && crm_is_true(value)) {
226  crm_info("Watchdog-based self-fencing will be performed via SBD if "
227  "fencing is required and stonith-watchdog-timeout is nonzero");
228  pe__set_working_set_flags(data_set, pe_flag_have_stonith_resource);
229  }
230 
231  /* Set certain flags via xpath here, so they can be used before the relevant
232  * configuration sections are unpacked.
233  */
234  set_if_xpath(pe_flag_enable_unfencing, XPATH_ENABLE_UNFENCING, data_set);
235 
236  value = pe_pref(data_set->config_hash, "stonith-timeout");
237  data_set->stonith_timeout = (int) crm_parse_interval_spec(value);
238  crm_debug("STONITH timeout: %d", data_set->stonith_timeout);
239 
240  set_config_flag(data_set, "stonith-enabled", pe_flag_stonith_enabled);
241  crm_debug("STONITH of failed nodes is %s",
242  pcmk_is_set(data_set->flags, pe_flag_stonith_enabled)? "enabled" : "disabled");
243 
244  data_set->stonith_action = pe_pref(data_set->config_hash, "stonith-action");
245  if (!strcmp(data_set->stonith_action, "poweroff")) {
246  pe_warn_once(pe_wo_poweroff,
247  "Support for stonith-action of 'poweroff' is deprecated "
248  "and will be removed in a future release (use 'off' instead)");
249  data_set->stonith_action = "off";
250  }
251  crm_trace("STONITH will %s nodes", data_set->stonith_action);
252 
253  set_config_flag(data_set, "concurrent-fencing", pe_flag_concurrent_fencing);
254  crm_debug("Concurrent fencing is %s",
255  pcmk_is_set(data_set->flags, pe_flag_concurrent_fencing)? "enabled" : "disabled");
256 
257  value = pe_pref(data_set->config_hash,
258  XML_CONFIG_ATTR_PRIORITY_FENCING_DELAY);
259  if (value) {
260  data_set->priority_fencing_delay = crm_parse_interval_spec(value) / 1000;
261  crm_trace("Priority fencing delay is %ds", data_set->priority_fencing_delay);
262  }
263 
264  set_config_flag(data_set, "stop-all-resources", pe_flag_stop_everything);
265  crm_debug("Stop all active resources: %s",
266  pcmk__btoa(pcmk_is_set(data_set->flags, pe_flag_stop_everything)));
267 
268  set_config_flag(data_set, "symmetric-cluster", pe_flag_symmetric_cluster);
269  if (pcmk_is_set(data_set->flags, pe_flag_symmetric_cluster)) {
270  crm_debug("Cluster is symmetric" " - resources can run anywhere by default");
271  }
272 
273  value = pe_pref(data_set->config_hash, "no-quorum-policy");
274 
275  if (pcmk__str_eq(value, "ignore", pcmk__str_casei)) {
276  data_set->no_quorum_policy = no_quorum_ignore;
277 
278  } else if (pcmk__str_eq(value, "freeze", pcmk__str_casei)) {
279  data_set->no_quorum_policy = no_quorum_freeze;
280 
281  } else if (pcmk__str_eq(value, "demote", pcmk__str_casei)) {
282  data_set->no_quorum_policy = no_quorum_demote;
283 
284  } else if (pcmk__str_eq(value, "suicide", pcmk__str_casei)) {
285  if (pcmk_is_set(data_set->flags, pe_flag_stonith_enabled)) {
286  int do_panic = 0;
287 
288  crm_element_value_int(data_set->input, XML_ATTR_QUORUM_PANIC,
289  &do_panic);
290  if (do_panic || pcmk_is_set(data_set->flags, pe_flag_have_quorum)) {
291  data_set->no_quorum_policy = no_quorum_suicide;
292  } else {
293  crm_notice("Resetting no-quorum-policy to 'stop': cluster has never had quorum");
294  data_set->no_quorum_policy = no_quorum_stop;
295  }
296  } else {
297  pcmk__config_err("Resetting no-quorum-policy to 'stop' because "
298  "fencing is disabled");
299  data_set->no_quorum_policy = no_quorum_stop;
300  }
301 
302  } else {
303  data_set->no_quorum_policy = no_quorum_stop;
304  }
305 
306  switch (data_set->no_quorum_policy) {
307  case no_quorum_freeze:
308  crm_debug("On loss of quorum: Freeze resources");
309  break;
310  case no_quorum_stop:
311  crm_debug("On loss of quorum: Stop ALL resources");
312  break;
313  case no_quorum_demote:
314  crm_debug("On loss of quorum: "
315  "Demote promotable resources and stop other resources");
316  break;
317  case no_quorum_suicide:
318  crm_notice("On loss of quorum: Fence all remaining nodes");
319  break;
320  case no_quorum_ignore:
321  crm_notice("On loss of quorum: Ignore");
322  break;
323  }
324 
325  set_config_flag(data_set, "stop-orphan-resources", pe_flag_stop_rsc_orphans);
326  crm_trace("Orphan resources are %s",
327  pcmk_is_set(data_set->flags, pe_flag_stop_rsc_orphans)? "stopped" : "ignored");
328 
329  set_config_flag(data_set, "stop-orphan-actions", pe_flag_stop_action_orphans);
330  crm_trace("Orphan resource actions are %s",
331  pcmk_is_set(data_set->flags, pe_flag_stop_action_orphans)? "stopped" : "ignored");
332 
333  set_config_flag(data_set, "remove-after-stop", pe_flag_remove_after_stop);
334  crm_trace("Stopped resources are removed from the status section: %s",
335  pcmk__btoa(pcmk_is_set(data_set->flags, pe_flag_remove_after_stop)));
336 
337  set_config_flag(data_set, "maintenance-mode", pe_flag_maintenance_mode);
338  crm_trace("Maintenance mode: %s",
339  pcmk__btoa(pcmk_is_set(data_set->flags, pe_flag_maintenance_mode)));
340 
341  set_config_flag(data_set, "start-failure-is-fatal", pe_flag_start_failure_fatal);
342  crm_trace("Start failures are %s",
343  pcmk_is_set(data_set->flags, pe_flag_start_failure_fatal)? "always fatal" : "handled by failcount");
344 
345  if (pcmk_is_set(data_set->flags, pe_flag_stonith_enabled)) {
346  set_config_flag(data_set, "startup-fencing", pe_flag_startup_fencing);
347  }
348  if (pcmk_is_set(data_set->flags, pe_flag_startup_fencing)) {
349  crm_trace("Unseen nodes will be fenced");
350  } else {
351  pe_warn_once(pe_wo_blind, "Blind faith: not fencing unseen nodes");
352  }
353 
354  pcmk__score_red = char2score(pe_pref(data_set->config_hash, "node-health-red"));
355  pcmk__score_green = char2score(pe_pref(data_set->config_hash, "node-health-green"));
356  pcmk__score_yellow = char2score(pe_pref(data_set->config_hash, "node-health-yellow"));
357 
358  crm_debug("Node scores: 'red' = %s, 'yellow' = %s, 'green' = %s",
359  pe_pref(data_set->config_hash, "node-health-red"),
360  pe_pref(data_set->config_hash, "node-health-yellow"),
361  pe_pref(data_set->config_hash, "node-health-green"));
362 
363  data_set->placement_strategy = pe_pref(data_set->config_hash, "placement-strategy");
364  crm_trace("Placement strategy: %s", data_set->placement_strategy);
365 
366  set_config_flag(data_set, "shutdown-lock", pe_flag_shutdown_lock);
367  crm_trace("Resources will%s be locked to cleanly shut down nodes",
368  (pcmk_is_set(data_set->flags, pe_flag_shutdown_lock)? "" : " not"));
369  if (pcmk_is_set(data_set->flags, pe_flag_shutdown_lock)) {
370  value = pe_pref(data_set->config_hash,
371  XML_CONFIG_ATTR_SHUTDOWN_LOCK_LIMIT);
372  data_set->shutdown_lock = crm_parse_interval_spec(value) / 1000;
373  crm_trace("Shutdown locks expire after %us", data_set->shutdown_lock);
374  }
375 
376  return TRUE;
377 }
378 
379 static void
380 destroy_digest_cache(gpointer ptr)
381 {
382  op_digest_cache_t *data = ptr;
383 
384  free_xml(data->params_all);
385  free_xml(data->params_secure);
386  free_xml(data->params_restart);
387 
388  free(data->digest_all_calc);
389  free(data->digest_restart_calc);
390  free(data->digest_secure_calc);
391 
392  free(data);
393 }
394 
395 pe_node_t *
396 pe_create_node(const char *id, const char *uname, const char *type,
397  const char *score, pe_working_set_t * data_set)
398 {
399  pe_node_t *new_node = NULL;
400 
401  if (pe_find_node(data_set->nodes, uname) != NULL) {
402  pcmk__config_warn("More than one node entry has name '%s'", uname);
403  }
404 
405  new_node = calloc(1, sizeof(pe_node_t));
406  if (new_node == NULL) {
407  return NULL;
408  }
409 
410  new_node->weight = char2score(score);
411  new_node->fixed = FALSE;
412  new_node->details = calloc(1, sizeof(struct pe_node_shared_s));
413 
414  if (new_node->details == NULL) {
415  free(new_node);
416  return NULL;
417  }
418 
419  crm_trace("Creating node for entry %s/%s", uname, id);
420  new_node->details->id = id;
421  new_node->details->uname = uname;
422  new_node->details->online = FALSE;
423  new_node->details->shutdown = FALSE;
424  new_node->details->rsc_discovery_enabled = TRUE;
425  new_node->details->running_rsc = NULL;
426  new_node->details->type = node_ping;
427 
428  if (pcmk__str_eq(type, "remote", pcmk__str_casei)) {
429  new_node->details->type = node_remote;
430  pe__set_working_set_flags(data_set, pe_flag_have_remote_nodes);
431  } else if (pcmk__str_eq(type, "member", pcmk__str_null_matches | pcmk__str_casei)) {
432  new_node->details->type = node_member;
433  }
434 
435  new_node->details->attrs = crm_str_table_new();
436 
437  if (pe__is_guest_or_remote_node(new_node)) {
438  g_hash_table_insert(new_node->details->attrs, strdup(CRM_ATTR_KIND),
439  strdup("remote"));
440  } else {
441  g_hash_table_insert(new_node->details->attrs, strdup(CRM_ATTR_KIND),
442  strdup("cluster"));
443  }
444 
445  new_node->details->utilization = crm_str_table_new();
446 
447  new_node->details->digest_cache = g_hash_table_new_full(crm_str_hash,
448  g_str_equal, free,
449  destroy_digest_cache);
450 
451  data_set->nodes = g_list_insert_sorted(data_set->nodes, new_node, sort_node_uname);
452  return new_node;
453 }
454 
455 static const char *
456 expand_remote_rsc_meta(xmlNode *xml_obj, xmlNode *parent, pe_working_set_t *data)
457 {
458  xmlNode *attr_set = NULL;
459  xmlNode *attr = NULL;
460 
461  const char *container_id = ID(xml_obj);
462  const char *remote_name = NULL;
463  const char *remote_server = NULL;
464  const char *remote_port = NULL;
465  const char *connect_timeout = "60s";
466  const char *remote_allow_migrate=NULL;
467  const char *is_managed = NULL;
468 
469  for (attr_set = pcmk__xe_first_child(xml_obj); attr_set != NULL;
470  attr_set = pcmk__xe_next(attr_set)) {
471 
472  if (!pcmk__str_eq((const char *)attr_set->name, XML_TAG_META_SETS,
473  pcmk__str_casei)) {
474  continue;
475  }
476 
477  for (attr = pcmk__xe_first_child(attr_set); attr != NULL;
478  attr = pcmk__xe_next(attr)) {
479  const char *value = crm_element_value(attr, XML_NVPAIR_ATTR_VALUE);
480  const char *name = crm_element_value(attr, XML_NVPAIR_ATTR_NAME);
481 
482  if (pcmk__str_eq(name, XML_RSC_ATTR_REMOTE_NODE, pcmk__str_casei)) {
483  remote_name = value;
484  } else if (pcmk__str_eq(name, "remote-addr", pcmk__str_casei)) {
485  remote_server = value;
486  } else if (pcmk__str_eq(name, "remote-port", pcmk__str_casei)) {
487  remote_port = value;
488  } else if (pcmk__str_eq(name, "remote-connect-timeout", pcmk__str_casei)) {
489  connect_timeout = value;
490  } else if (pcmk__str_eq(name, "remote-allow-migrate", pcmk__str_casei)) {
491  remote_allow_migrate=value;
492  } else if (pcmk__str_eq(name, XML_RSC_ATTR_MANAGED, pcmk__str_casei)) {
493  is_managed = value;
494  }
495  }
496  }
497 
498  if (remote_name == NULL) {
499  return NULL;
500  }
501 
502  if (pe_find_resource(data->resources, remote_name) != NULL) {
503  return NULL;
504  }
505 
506  pe_create_remote_xml(parent, remote_name, container_id,
507  remote_allow_migrate, is_managed,
508  connect_timeout, remote_server, remote_port);
509  return remote_name;
510 }
511 
512 static void
513 handle_startup_fencing(pe_working_set_t *data_set, pe_node_t *new_node)
514 {
515  if ((new_node->details->type == node_remote) && (new_node->details->remote_rsc == NULL)) {
516  /* Ignore fencing for remote nodes that don't have a connection resource
517  * associated with them. This happens when remote node entries get left
518  * in the nodes section after the connection resource is removed.
519  */
520  return;
521  }
522 
523  if (pcmk_is_set(data_set->flags, pe_flag_startup_fencing)) {
524  // All nodes are unclean until we've seen their status entry
525  new_node->details->unclean = TRUE;
526 
527  } else {
528  // Blind faith ...
529  new_node->details->unclean = FALSE;
530  }
531 
532  /* We need to be able to determine if a node's status section
533  * exists or not separate from whether the node is unclean. */
534  new_node->details->unseen = TRUE;
535 }
536 
537 gboolean
538 unpack_nodes(xmlNode * xml_nodes, pe_working_set_t * data_set)
539 {
540  xmlNode *xml_obj = NULL;
541  pe_node_t *new_node = NULL;
542  const char *id = NULL;
543  const char *uname = NULL;
544  const char *type = NULL;
545  const char *score = NULL;
546 
547  pe_rule_eval_data_t rule_data = {
548  .node_hash = NULL,
549  .role = RSC_ROLE_UNKNOWN,
550  .now = data_set->now,
551  .match_data = NULL,
552  .rsc_data = NULL,
553  .op_data = NULL
554  };
555 
556  for (xml_obj = pcmk__xe_first_child(xml_nodes); xml_obj != NULL;
557  xml_obj = pcmk__xe_next(xml_obj)) {
558 
559  if (pcmk__str_eq((const char *)xml_obj->name, XML_CIB_TAG_NODE, pcmk__str_none)) {
560  new_node = NULL;
561 
562  id = crm_element_value(xml_obj, XML_ATTR_ID);
563  uname = crm_element_value(xml_obj, XML_ATTR_UNAME);
564  type = crm_element_value(xml_obj, XML_ATTR_TYPE);
565  score = crm_element_value(xml_obj, XML_RULE_ATTR_SCORE);
566  crm_trace("Processing node %s/%s", uname, id);
567 
568  if (id == NULL) {
569  pcmk__config_err("Ignoring <" XML_CIB_TAG_NODE
570  "> entry in configuration without id");
571  continue;
572  }
573  new_node = pe_create_node(id, uname, type, score, data_set);
574 
575  if (new_node == NULL) {
576  return FALSE;
577  }
578 
579 /* if(data_set->have_quorum == FALSE */
580 /* && data_set->no_quorum_policy == no_quorum_stop) { */
581 /* /\* start shutting resources down *\/ */
582 /* new_node->weight = -INFINITY; */
583 /* } */
584 
585  handle_startup_fencing(data_set, new_node);
586 
587  add_node_attrs(xml_obj, new_node, FALSE, data_set);
588  pe__unpack_dataset_nvpairs(xml_obj, XML_TAG_UTILIZATION, &rule_data,
589  new_node->details->utilization, NULL,
590  FALSE, data_set);
591 
592  crm_trace("Done with node %s", crm_element_value(xml_obj, XML_ATTR_UNAME));
593  }
594  }
595 
596  if (data_set->localhost && pe_find_node(data_set->nodes, data_set->localhost) == NULL) {
597  crm_info("Creating a fake local node");
598  pe_create_node(data_set->localhost, data_set->localhost, NULL, 0,
599  data_set);
600  }
601 
602  return TRUE;
603 }
604 
605 static void
606 setup_container(pe_resource_t * rsc, pe_working_set_t * data_set)
607 {
608  const char *container_id = NULL;
609 
610  if (rsc->children) {
611  GListPtr gIter = rsc->children;
612 
613  for (; gIter != NULL; gIter = gIter->next) {
614  pe_resource_t *child_rsc = (pe_resource_t *) gIter->data;
615 
616  setup_container(child_rsc, data_set);
617  }
618  return;
619  }
620 
621  container_id = g_hash_table_lookup(rsc->meta, XML_RSC_ATTR_CONTAINER);
622  if (container_id && !pcmk__str_eq(container_id, rsc->id, pcmk__str_casei)) {
623  pe_resource_t *container = pe_find_resource(data_set->resources, container_id);
624 
625  if (container) {
626  rsc->container = container;
627  pe__set_resource_flags(container, pe_rsc_is_container);
628  container->fillers = g_list_append(container->fillers, rsc);
629  pe_rsc_trace(rsc, "Resource %s's container is %s", rsc->id, container_id);
630  } else {
631  pe_err("Resource %s: Unknown resource container (%s)", rsc->id, container_id);
632  }
633  }
634 }
635 
636 gboolean
637 unpack_remote_nodes(xmlNode * xml_resources, pe_working_set_t * data_set)
638 {
639  xmlNode *xml_obj = NULL;
640 
641  /* Create remote nodes and guest nodes from the resource configuration
642  * before unpacking resources.
643  */
644  for (xml_obj = pcmk__xe_first_child(xml_resources); xml_obj != NULL;
645  xml_obj = pcmk__xe_next(xml_obj)) {
646 
647  const char *new_node_id = NULL;
648 
649  /* Check for remote nodes, which are defined by ocf:pacemaker:remote
650  * primitives.
651  */
652  if (xml_contains_remote_node(xml_obj)) {
653  new_node_id = ID(xml_obj);
654  /* The "pe_find_node" check is here to make sure we don't iterate over
655  * an expanded node that has already been added to the node list. */
656  if (new_node_id && pe_find_node(data_set->nodes, new_node_id) == NULL) {
657  crm_trace("Found remote node %s defined by resource %s",
658  new_node_id, ID(xml_obj));
659  pe_create_node(new_node_id, new_node_id, "remote", NULL,
660  data_set);
661  }
662  continue;
663  }
664 
665  /* Check for guest nodes, which are defined by special meta-attributes
666  * of a primitive of any type (for example, VirtualDomain or Xen).
667  */
668  if (pcmk__str_eq((const char *)xml_obj->name, XML_CIB_TAG_RESOURCE, pcmk__str_none)) {
669  /* This will add an ocf:pacemaker:remote primitive to the
670  * configuration for the guest node's connection, to be unpacked
671  * later.
672  */
673  new_node_id = expand_remote_rsc_meta(xml_obj, xml_resources, data_set);
674  if (new_node_id && pe_find_node(data_set->nodes, new_node_id) == NULL) {
675  crm_trace("Found guest node %s in resource %s",
676  new_node_id, ID(xml_obj));
677  pe_create_node(new_node_id, new_node_id, "remote", NULL,
678  data_set);
679  }
680  continue;
681  }
682 
683  /* Check for guest nodes inside a group. Clones are currently not
684  * supported as guest nodes.
685  */
686  if (pcmk__str_eq((const char *)xml_obj->name, XML_CIB_TAG_GROUP, pcmk__str_none)) {
687  xmlNode *xml_obj2 = NULL;
688  for (xml_obj2 = pcmk__xe_first_child(xml_obj); xml_obj2 != NULL;
689  xml_obj2 = pcmk__xe_next(xml_obj2)) {
690 
691  new_node_id = expand_remote_rsc_meta(xml_obj2, xml_resources, data_set);
692 
693  if (new_node_id && pe_find_node(data_set->nodes, new_node_id) == NULL) {
694  crm_trace("Found guest node %s in resource %s inside group %s",
695  new_node_id, ID(xml_obj2), ID(xml_obj));
696  pe_create_node(new_node_id, new_node_id, "remote", NULL,
697  data_set);
698  }
699  }
700  }
701  }
702  return TRUE;
703 }
704 
705 /* Call this after all the nodes and resources have been
706  * unpacked, but before the status section is read.
707  *
708  * A remote node's online status is reflected by the state
709  * of the remote node's connection resource. We need to link
710  * the remote node to this connection resource so we can have
711  * easy access to the connection resource during the scheduler calculations.
712  */
713 static void
714 link_rsc2remotenode(pe_working_set_t *data_set, pe_resource_t *new_rsc)
715 {
716  pe_node_t *remote_node = NULL;
717 
718  if (new_rsc->is_remote_node == FALSE) {
719  return;
720  }
721 
722  if (pcmk_is_set(data_set->flags, pe_flag_quick_location)) {
723  /* remote_nodes and remote_resources are not linked in quick location calculations */
724  return;
725  }
726 
727  remote_node = pe_find_node(data_set->nodes, new_rsc->id);
728  CRM_CHECK(remote_node != NULL, return;);
729 
730  pe_rsc_trace(new_rsc, "Linking remote connection resource %s to node %s",
731  new_rsc->id, remote_node->details->uname);
732  remote_node->details->remote_rsc = new_rsc;
733 
734  if (new_rsc->container == NULL) {
735  /* Handle start-up fencing for remote nodes (as opposed to guest nodes)
736  * the same as is done for cluster nodes.
737  */
738  handle_startup_fencing(data_set, remote_node);
739 
740  } else {
741  /* pe_create_node() marks the new node as "remote" or "cluster"; now
742  * that we know the node is a guest node, update it correctly.
743  */
744  g_hash_table_replace(remote_node->details->attrs, strdup(CRM_ATTR_KIND),
745  strdup("container"));
746  }
747 }
748 
749 static void
750 destroy_tag(gpointer data)
751 {
752  pe_tag_t *tag = data;
753 
754  if (tag) {
755  free(tag->id);
756  g_list_free_full(tag->refs, free);
757  free(tag);
758  }
759 }
760 
773 gboolean
774 unpack_resources(xmlNode * xml_resources, pe_working_set_t * data_set)
775 {
776  xmlNode *xml_obj = NULL;
777  GListPtr gIter = NULL;
778 
779  data_set->template_rsc_sets = g_hash_table_new_full(crm_str_hash,
780  g_str_equal, free,
781  destroy_tag);
782 
783  for (xml_obj = pcmk__xe_first_child(xml_resources); xml_obj != NULL;
784  xml_obj = pcmk__xe_next(xml_obj)) {
785 
786  pe_resource_t *new_rsc = NULL;
787 
788  if (pcmk__str_eq((const char *)xml_obj->name, XML_CIB_TAG_RSC_TEMPLATE, pcmk__str_none)) {
789  const char *template_id = ID(xml_obj);
790 
791  if (template_id && g_hash_table_lookup_extended(data_set->template_rsc_sets,
792  template_id, NULL, NULL) == FALSE) {
793  /* Record the template's ID for the knowledge of its existence anyway. */
794  g_hash_table_insert(data_set->template_rsc_sets, strdup(template_id), NULL);
795  }
796  continue;
797  }
798 
799  crm_trace("Beginning unpack... <%s id=%s... >", crm_element_name(xml_obj), ID(xml_obj));
800  if (common_unpack(xml_obj, &new_rsc, NULL, data_set) && (new_rsc != NULL)) {
801  data_set->resources = g_list_append(data_set->resources, new_rsc);
802  pe_rsc_trace(new_rsc, "Added resource %s", new_rsc->id);
803 
804  } else {
805  pcmk__config_err("Ignoring <%s> resource '%s' "
806  "because configuration is invalid",
807  crm_element_name(xml_obj), crm_str(ID(xml_obj)));
808  if (new_rsc != NULL && new_rsc->fns != NULL) {
809  new_rsc->fns->free(new_rsc);
810  }
811  }
812  }
813 
814  for (gIter = data_set->resources; gIter != NULL; gIter = gIter->next) {
815  pe_resource_t *rsc = (pe_resource_t *) gIter->data;
816 
817  setup_container(rsc, data_set);
818  link_rsc2remotenode(data_set, rsc);
819  }
820 
821  data_set->resources = g_list_sort(data_set->resources, sort_rsc_priority);
822  if (pcmk_is_set(data_set->flags, pe_flag_quick_location)) {
823  /* Ignore */
824 
825  } else if (pcmk_is_set(data_set->flags, pe_flag_stonith_enabled)
826  && !pcmk_is_set(data_set->flags, pe_flag_have_stonith_resource)) {
827 
828  pcmk__config_err("Resource start-up disabled since no STONITH resources have been defined");
829  pcmk__config_err("Either configure some or disable STONITH with the stonith-enabled option");
830  pcmk__config_err("NOTE: Clusters with shared data need STONITH to ensure data integrity");
831  }
832 
833  return TRUE;
834 }
835 
836 gboolean
837 unpack_tags(xmlNode * xml_tags, pe_working_set_t * data_set)
838 {
839  xmlNode *xml_tag = NULL;
840 
841  data_set->tags = g_hash_table_new_full(crm_str_hash, g_str_equal, free,
842  destroy_tag);
843 
844  for (xml_tag = pcmk__xe_first_child(xml_tags); xml_tag != NULL;
845  xml_tag = pcmk__xe_next(xml_tag)) {
846 
847  xmlNode *xml_obj_ref = NULL;
848  const char *tag_id = ID(xml_tag);
849 
850  if (!pcmk__str_eq((const char *)xml_tag->name, XML_CIB_TAG_TAG, pcmk__str_none)) {
851  continue;
852  }
853 
854  if (tag_id == NULL) {
855  pcmk__config_err("Ignoring <%s> without " XML_ATTR_ID,
856  crm_element_name(xml_tag));
857  continue;
858  }
859 
860  for (xml_obj_ref = pcmk__xe_first_child(xml_tag); xml_obj_ref != NULL;
861  xml_obj_ref = pcmk__xe_next(xml_obj_ref)) {
862 
863  const char *obj_ref = ID(xml_obj_ref);
864 
865  if (!pcmk__str_eq((const char *)xml_obj_ref->name, XML_CIB_TAG_OBJ_REF, pcmk__str_none)) {
866  continue;
867  }
868 
869  if (obj_ref == NULL) {
870  pcmk__config_err("Ignoring <%s> for tag '%s' without " XML_ATTR_ID,
871  crm_element_name(xml_obj_ref), tag_id);
872  continue;
873  }
874 
875  if (add_tag_ref(data_set->tags, tag_id, obj_ref) == FALSE) {
876  return FALSE;
877  }
878  }
879  }
880 
881  return TRUE;
882 }
883 
884 /* The ticket state section:
885  * "/cib/status/tickets/ticket_state" */
886 static gboolean
887 unpack_ticket_state(xmlNode * xml_ticket, pe_working_set_t * data_set)
888 {
889  const char *ticket_id = NULL;
890  const char *granted = NULL;
891  const char *last_granted = NULL;
892  const char *standby = NULL;
893  xmlAttrPtr xIter = NULL;
894 
895  pe_ticket_t *ticket = NULL;
896 
897  ticket_id = ID(xml_ticket);
898  if (pcmk__str_empty(ticket_id)) {
899  return FALSE;
900  }
901 
902  crm_trace("Processing ticket state for %s", ticket_id);
903 
904  ticket = g_hash_table_lookup(data_set->tickets, ticket_id);
905  if (ticket == NULL) {
906  ticket = ticket_new(ticket_id, data_set);
907  if (ticket == NULL) {
908  return FALSE;
909  }
910  }
911 
912  for (xIter = xml_ticket->properties; xIter; xIter = xIter->next) {
913  const char *prop_name = (const char *)xIter->name;
914  const char *prop_value = crm_element_value(xml_ticket, prop_name);
915 
916  if (pcmk__str_eq(prop_name, XML_ATTR_ID, pcmk__str_none)) {
917  continue;
918  }
919  g_hash_table_replace(ticket->state, strdup(prop_name), strdup(prop_value));
920  }
921 
922  granted = g_hash_table_lookup(ticket->state, "granted");
923  if (granted && crm_is_true(granted)) {
924  ticket->granted = TRUE;
925  crm_info("We have ticket '%s'", ticket->id);
926  } else {
927  ticket->granted = FALSE;
928  crm_info("We do not have ticket '%s'", ticket->id);
929  }
930 
931  last_granted = g_hash_table_lookup(ticket->state, "last-granted");
932  if (last_granted) {
933  ticket->last_granted = crm_parse_int(last_granted, 0);
934  }
935 
936  standby = g_hash_table_lookup(ticket->state, "standby");
937  if (standby && crm_is_true(standby)) {
938  ticket->standby = TRUE;
939  if (ticket->granted) {
940  crm_info("Granted ticket '%s' is in standby-mode", ticket->id);
941  }
942  } else {
943  ticket->standby = FALSE;
944  }
945 
946  crm_trace("Done with ticket state for %s", ticket_id);
947 
948  return TRUE;
949 }
950 
951 static gboolean
952 unpack_tickets_state(xmlNode * xml_tickets, pe_working_set_t * data_set)
953 {
954  xmlNode *xml_obj = NULL;
955 
956  for (xml_obj = pcmk__xe_first_child(xml_tickets); xml_obj != NULL;
957  xml_obj = pcmk__xe_next(xml_obj)) {
958 
959  if (!pcmk__str_eq((const char *)xml_obj->name, XML_CIB_TAG_TICKET_STATE, pcmk__str_none)) {
960  continue;
961  }
962  unpack_ticket_state(xml_obj, data_set);
963  }
964 
965  return TRUE;
966 }
967 
968 static void
969 unpack_handle_remote_attrs(pe_node_t *this_node, xmlNode *state, pe_working_set_t * data_set)
970 {
971  const char *resource_discovery_enabled = NULL;
972  xmlNode *attrs = NULL;
973  pe_resource_t *rsc = NULL;
974 
975  if (!pcmk__str_eq((const char *)state->name, XML_CIB_TAG_STATE, pcmk__str_none)) {
976  return;
977  }
978 
979  if ((this_node == NULL) || !pe__is_guest_or_remote_node(this_node)) {
980  return;
981  }
982  crm_trace("Processing remote node id=%s, uname=%s", this_node->details->id, this_node->details->uname);
983 
984  this_node->details->remote_maintenance =
985  crm_atoi(crm_element_value(state, XML_NODE_IS_MAINTENANCE), "0");
986 
987  rsc = this_node->details->remote_rsc;
988  if (this_node->details->remote_requires_reset == FALSE) {
989  this_node->details->unclean = FALSE;
990  this_node->details->unseen = FALSE;
991  }
992  attrs = find_xml_node(state, XML_TAG_TRANSIENT_NODEATTRS, FALSE);
993  add_node_attrs(attrs, this_node, TRUE, data_set);
994 
995  if (pe__shutdown_requested(this_node)) {
996  crm_info("Node %s is shutting down", this_node->details->uname);
997  this_node->details->shutdown = TRUE;
998  if (rsc) {
999  rsc->next_role = RSC_ROLE_STOPPED;
1000  }
1001  }
1002 
1003  if (crm_is_true(pe_node_attribute_raw(this_node, "standby"))) {
1004  crm_info("Node %s is in standby-mode", this_node->details->uname);
1005  this_node->details->standby = TRUE;
1006  }
1007 
1008  if (crm_is_true(pe_node_attribute_raw(this_node, "maintenance")) ||
1009  ((rsc != NULL) && !pcmk_is_set(rsc->flags, pe_rsc_managed))) {
1010  crm_info("Node %s is in maintenance-mode", this_node->details->uname);
1011  this_node->details->maintenance = TRUE;
1012  }
1013 
1014  resource_discovery_enabled = pe_node_attribute_raw(this_node, XML_NODE_ATTR_RSC_DISCOVERY);
1015  if (resource_discovery_enabled && !crm_is_true(resource_discovery_enabled)) {
1016  if (pe__is_remote_node(this_node)
1017  && !pcmk_is_set(data_set->flags, pe_flag_stonith_enabled)) {
1018  crm_warn("Ignoring %s attribute on remote node %s because stonith is disabled",
1019  XML_NODE_ATTR_RSC_DISCOVERY, this_node->details->uname);
1020  } else {
1021  /* This is either a remote node with fencing enabled, or a guest
1022  * node. We don't care whether fencing is enabled when fencing guest
1023  * nodes, because they are "fenced" by recovering their containing
1024  * resource.
1025  */
1026  crm_info("Node %s has resource discovery disabled", this_node->details->uname);
1027  this_node->details->rsc_discovery_enabled = FALSE;
1028  }
1029  }
1030 }
1031 
1032 static bool
1033 unpack_node_loop(xmlNode * status, bool fence, pe_working_set_t * data_set)
1034 {
1035  bool changed = false;
1036  xmlNode *lrm_rsc = NULL;
1037 
1038  for (xmlNode *state = pcmk__xe_first_child(status); state != NULL;
1039  state = pcmk__xe_next(state)) {
1040 
1041  const char *id = NULL;
1042  const char *uname = NULL;
1043  pe_node_t *this_node = NULL;
1044  bool process = FALSE;
1045 
1046  if (!pcmk__str_eq((const char *)state->name, XML_CIB_TAG_STATE, pcmk__str_none)) {
1047  continue;
1048  }
1049 
1050  id = crm_element_value(state, XML_ATTR_ID);
1051  uname = crm_element_value(state, XML_ATTR_UNAME);
1052  this_node = pe_find_node_any(data_set->nodes, id, uname);
1053 
1054  if (this_node == NULL) {
1055  crm_info("Node %s is unknown", id);
1056  continue;
1057 
1058  } else if (this_node->details->unpacked) {
1059  crm_trace("Node %s was already processed", id);
1060  continue;
1061 
1062  } else if (!pe__is_guest_or_remote_node(this_node)
1063  && pcmk_is_set(data_set->flags, pe_flag_stonith_enabled)) {
1064  // A redundant test, but preserves the order for regression tests
1065  process = TRUE;
1066 
1067  } else if (pe__is_guest_or_remote_node(this_node)) {
1068  bool check = FALSE;
1069  pe_resource_t *rsc = this_node->details->remote_rsc;
1070 
1071  if(fence) {
1072  check = TRUE;
1073 
1074  } else if(rsc == NULL) {
1075  /* Not ready yet */
1076 
1077  } else if (pe__is_guest_node(this_node)
1078  && rsc->role == RSC_ROLE_STARTED
1079  && rsc->container->role == RSC_ROLE_STARTED) {
1080  /* Both the connection and its containing resource need to be
1081  * known to be up before we process resources running in it.
1082  */
1083  check = TRUE;
1084  crm_trace("Checking node %s/%s/%s status %d/%d/%d", id, rsc->id, rsc->container->id, fence, rsc->role, RSC_ROLE_STARTED);
1085 
1086  } else if (!pe__is_guest_node(this_node)
1087  && ((rsc->role == RSC_ROLE_STARTED)
1088  || pcmk_is_set(data_set->flags, pe_flag_shutdown_lock))) {
1089  check = TRUE;
1090  crm_trace("Checking node %s/%s status %d/%d/%d", id, rsc->id, fence, rsc->role, RSC_ROLE_STARTED);
1091  }
1092 
1093  if (check) {
1094  determine_remote_online_status(data_set, this_node);
1095  unpack_handle_remote_attrs(this_node, state, data_set);
1096  process = TRUE;
1097  }
1098 
1099  } else if (this_node->details->online) {
1100  process = TRUE;
1101 
1102  } else if (fence) {
1103  process = TRUE;
1104 
1105  } else if (pcmk_is_set(data_set->flags, pe_flag_shutdown_lock)) {
1106  process = TRUE;
1107  }
1108 
1109  if(process) {
1110  crm_trace("Processing lrm resource entries on %shealthy%s node: %s",
1111  fence?"un":"",
1112  (pe__is_guest_or_remote_node(this_node)? " remote" : ""),
1113  this_node->details->uname);
1114  changed = TRUE;
1115  this_node->details->unpacked = TRUE;
1116 
1117  lrm_rsc = find_xml_node(state, XML_CIB_TAG_LRM, FALSE);
1118  lrm_rsc = find_xml_node(lrm_rsc, XML_LRM_TAG_RESOURCES, FALSE);
1119  unpack_lrm_resources(this_node, lrm_rsc, data_set);
1120  }
1121  }
1122  return changed;
1123 }
1124 
1125 /* remove nodes that are down, stopping */
1126 /* create positive rsc_to_node constraints between resources and the nodes they are running on */
1127 /* anything else? */
1128 gboolean
1129 unpack_status(xmlNode * status, pe_working_set_t * data_set)
1130 {
1131  const char *id = NULL;
1132  const char *uname = NULL;
1133 
1134  xmlNode *state = NULL;
1135  pe_node_t *this_node = NULL;
1136 
1137  crm_trace("Beginning unpack");
1138 
1139  if (data_set->tickets == NULL) {
1140  data_set->tickets = g_hash_table_new_full(crm_str_hash, g_str_equal,
1141  free, destroy_ticket);
1142  }
1143 
1144  for (state = pcmk__xe_first_child(status); state != NULL;
1145  state = pcmk__xe_next(state)) {
1146 
1147  if (pcmk__str_eq((const char *)state->name, XML_CIB_TAG_TICKETS, pcmk__str_none)) {
1148  unpack_tickets_state((xmlNode *) state, data_set);
1149 
1150  } else if (pcmk__str_eq((const char *)state->name, XML_CIB_TAG_STATE, pcmk__str_none)) {
1151  xmlNode *attrs = NULL;
1152  const char *resource_discovery_enabled = NULL;
1153 
1154  id = crm_element_value(state, XML_ATTR_ID);
1155  uname = crm_element_value(state, XML_ATTR_UNAME);
1156  this_node = pe_find_node_any(data_set->nodes, id, uname);
1157 
1158  if (uname == NULL) {
1159  /* error */
1160  continue;
1161 
1162  } else if (this_node == NULL) {
1163  pcmk__config_warn("Ignoring recorded node status for '%s' "
1164  "because no longer in configuration", uname);
1165  continue;
1166 
1167  } else if (pe__is_guest_or_remote_node(this_node)) {
1168  /* online state for remote nodes is determined by the
1169  * rsc state after all the unpacking is done. we do however
1170  * need to mark whether or not the node has been fenced as this plays
1171  * a role during unpacking cluster node resource state */
1172  this_node->details->remote_was_fenced =
1173  crm_atoi(crm_element_value(state, XML_NODE_IS_FENCED), "0");
1174  continue;
1175  }
1176 
1177  crm_trace("Processing node id=%s, uname=%s", id, uname);
1178 
1179  /* Mark the node as provisionally clean
1180  * - at least we have seen it in the current cluster's lifetime
1181  */
1182  this_node->details->unclean = FALSE;
1183  this_node->details->unseen = FALSE;
1184  attrs = find_xml_node(state, XML_TAG_TRANSIENT_NODEATTRS, FALSE);
1185  add_node_attrs(attrs, this_node, TRUE, data_set);
1186 
1187  if (crm_is_true(pe_node_attribute_raw(this_node, "standby"))) {
1188  crm_info("Node %s is in standby-mode", this_node->details->uname);
1189  this_node->details->standby = TRUE;
1190  }
1191 
1192  if (crm_is_true(pe_node_attribute_raw(this_node, "maintenance"))) {
1193  crm_info("Node %s is in maintenance-mode", this_node->details->uname);
1194  this_node->details->maintenance = TRUE;
1195  }
1196 
1197  resource_discovery_enabled = pe_node_attribute_raw(this_node, XML_NODE_ATTR_RSC_DISCOVERY);
1198  if (resource_discovery_enabled && !crm_is_true(resource_discovery_enabled)) {
1199  crm_warn("ignoring %s attribute on node %s, disabling resource discovery is not allowed on cluster nodes",
1200  XML_NODE_ATTR_RSC_DISCOVERY, this_node->details->uname);
1201  }
1202 
1203  crm_trace("determining node state");
1204  determine_online_status(state, this_node, data_set);
1205 
1206  if (!pcmk_is_set(data_set->flags, pe_flag_have_quorum)
1207  && this_node->details->online
1208  && (data_set->no_quorum_policy == no_quorum_suicide)) {
1209  /* Everything else should flow from this automatically
1210  * (at least until the scheduler becomes able to migrate off
1211  * healthy resources)
1212  */
1213  pe_fence_node(data_set, this_node, "cluster does not have quorum", FALSE);
1214  }
1215  }
1216  }
1217 
1218 
1219  while(unpack_node_loop(status, FALSE, data_set)) {
1220  crm_trace("Start another loop");
1221  }
1222 
1223  // Now catch any nodes we didn't see
1224  unpack_node_loop(status,
1225  pcmk_is_set(data_set->flags, pe_flag_stonith_enabled),
1226  data_set);
1227 
1228  /* Now that we know where resources are, we can schedule stops of containers
1229  * with failed bundle connections
1230  */
1231  if (data_set->stop_needed != NULL) {
1232  for (GList *item = data_set->stop_needed; item; item = item->next) {
1233  pe_resource_t *container = item->data;
1234  pe_node_t *node = pe__current_node(container);
1235 
1236  if (node) {
1237  stop_action(container, node, FALSE);
1238  }
1239  }
1240  g_list_free(data_set->stop_needed);
1241  data_set->stop_needed = NULL;
1242  }
1243 
1244  for (GListPtr gIter = data_set->nodes; gIter != NULL; gIter = gIter->next) {
1245  pe_node_t *this_node = gIter->data;
1246 
1247  if (this_node == NULL) {
1248  continue;
1249  } else if (!pe__is_guest_or_remote_node(this_node)) {
1250  continue;
1251  } else if(this_node->details->unpacked) {
1252  continue;
1253  }
1254  determine_remote_online_status(data_set, this_node);
1255  }
1256 
1257  return TRUE;
1258 }
1259 
1260 static gboolean
1261 determine_online_status_no_fencing(pe_working_set_t * data_set, xmlNode * node_state,
1262  pe_node_t * this_node)
1263 {
1264  gboolean online = FALSE;
1265  const char *join = crm_element_value(node_state, XML_NODE_JOIN_STATE);
1266  const char *is_peer = crm_element_value(node_state, XML_NODE_IS_PEER);
1267  const char *in_cluster = crm_element_value(node_state, XML_NODE_IN_CLUSTER);
1268  const char *exp_state = crm_element_value(node_state, XML_NODE_EXPECTED);
1269 
1270  if (!crm_is_true(in_cluster)) {
1271  crm_trace("Node is down: in_cluster=%s", crm_str(in_cluster));
1272 
1273  } else if (pcmk__str_eq(is_peer, ONLINESTATUS, pcmk__str_casei)) {
1274  if (pcmk__str_eq(join, CRMD_JOINSTATE_MEMBER, pcmk__str_casei)) {
1275  online = TRUE;
1276  } else {
1277  crm_debug("Node is not ready to run resources: %s", join);
1278  }
1279 
1280  } else if (this_node->details->expected_up == FALSE) {
1281  crm_trace("Controller is down: in_cluster=%s", crm_str(in_cluster));
1282  crm_trace("\tis_peer=%s, join=%s, expected=%s",
1283  crm_str(is_peer), crm_str(join), crm_str(exp_state));
1284 
1285  } else {
1286  /* mark it unclean */
1287  pe_fence_node(data_set, this_node, "peer is unexpectedly down", FALSE);
1288  crm_info("\tin_cluster=%s, is_peer=%s, join=%s, expected=%s",
1289  crm_str(in_cluster), crm_str(is_peer), crm_str(join), crm_str(exp_state));
1290  }
1291  return online;
1292 }
1293 
1294 static gboolean
1295 determine_online_status_fencing(pe_working_set_t * data_set, xmlNode * node_state,
1296  pe_node_t * this_node)
1297 {
1298  gboolean online = FALSE;
1299  gboolean do_terminate = FALSE;
1300  bool crmd_online = FALSE;
1301  const char *join = crm_element_value(node_state, XML_NODE_JOIN_STATE);
1302  const char *is_peer = crm_element_value(node_state, XML_NODE_IS_PEER);
1303  const char *in_cluster = crm_element_value(node_state, XML_NODE_IN_CLUSTER);
1304  const char *exp_state = crm_element_value(node_state, XML_NODE_EXPECTED);
1305  const char *terminate = pe_node_attribute_raw(this_node, "terminate");
1306 
1307 /*
1308  - XML_NODE_IN_CLUSTER ::= true|false
1309  - XML_NODE_IS_PEER ::= online|offline
1310  - XML_NODE_JOIN_STATE ::= member|down|pending|banned
1311  - XML_NODE_EXPECTED ::= member|down
1312 */
1313 
1314  if (crm_is_true(terminate)) {
1315  do_terminate = TRUE;
1316 
1317  } else if (terminate != NULL && strlen(terminate) > 0) {
1318  /* could be a time() value */
1319  char t = terminate[0];
1320 
1321  if (t != '0' && isdigit(t)) {
1322  do_terminate = TRUE;
1323  }
1324  }
1325 
1326  crm_trace("%s: in_cluster=%s, is_peer=%s, join=%s, expected=%s, term=%d",
1327  this_node->details->uname, crm_str(in_cluster), crm_str(is_peer),
1328  crm_str(join), crm_str(exp_state), do_terminate);
1329 
1330  online = crm_is_true(in_cluster);
1331  crmd_online = pcmk__str_eq(is_peer, ONLINESTATUS, pcmk__str_casei);
1332  if (exp_state == NULL) {
1333  exp_state = CRMD_JOINSTATE_DOWN;
1334  }
1335 
1336  if (this_node->details->shutdown) {
1337  crm_debug("%s is shutting down", this_node->details->uname);
1338 
1339  /* Slightly different criteria since we can't shut down a dead peer */
1340  online = crmd_online;
1341 
1342  } else if (in_cluster == NULL) {
1343  pe_fence_node(data_set, this_node, "peer has not been seen by the cluster", FALSE);
1344 
1345  } else if (pcmk__str_eq(join, CRMD_JOINSTATE_NACK, pcmk__str_casei)) {
1346  pe_fence_node(data_set, this_node, "peer failed the pacemaker membership criteria", FALSE);
1347 
1348  } else if (do_terminate == FALSE && pcmk__str_eq(exp_state, CRMD_JOINSTATE_DOWN, pcmk__str_casei)) {
1349 
1350  if (crm_is_true(in_cluster) || crmd_online) {
1351  crm_info("- Node %s is not ready to run resources", this_node->details->uname);
1352  this_node->details->standby = TRUE;
1353  this_node->details->pending = TRUE;
1354 
1355  } else {
1356  crm_trace("%s is down or still coming up", this_node->details->uname);
1357  }
1358 
1359  } else if (do_terminate && pcmk__str_eq(join, CRMD_JOINSTATE_DOWN, pcmk__str_casei)
1360  && crm_is_true(in_cluster) == FALSE && !crmd_online) {
1361  crm_info("Node %s was just shot", this_node->details->uname);
1362  online = FALSE;
1363 
1364  } else if (crm_is_true(in_cluster) == FALSE) {
1365  // Consider `priority-fencing-delay` for lost nodes
1366  pe_fence_node(data_set, this_node, "peer is no longer part of the cluster", TRUE);
1367 
1368  } else if (!crmd_online) {
1369  pe_fence_node(data_set, this_node, "peer process is no longer available", FALSE);
1370 
1371  /* Everything is running at this point, now check join state */
1372  } else if (do_terminate) {
1373  pe_fence_node(data_set, this_node, "termination was requested", FALSE);
1374 
1375  } else if (pcmk__str_eq(join, CRMD_JOINSTATE_MEMBER, pcmk__str_casei)) {
1376  crm_info("Node %s is active", this_node->details->uname);
1377 
1378  } else if (pcmk__strcase_any_of(join, CRMD_JOINSTATE_PENDING, CRMD_JOINSTATE_DOWN, NULL)) {
1379  crm_info("Node %s is not ready to run resources", this_node->details->uname);
1380  this_node->details->standby = TRUE;
1381  this_node->details->pending = TRUE;
1382 
1383  } else {
1384  pe_fence_node(data_set, this_node, "peer was in an unknown state", FALSE);
1385  crm_warn("%s: in-cluster=%s, is-peer=%s, join=%s, expected=%s, term=%d, shutdown=%d",
1386  this_node->details->uname, crm_str(in_cluster), crm_str(is_peer),
1387  crm_str(join), crm_str(exp_state), do_terminate, this_node->details->shutdown);
1388  }
1389 
1390  return online;
1391 }
1392 
1393 static void
1394 determine_remote_online_status(pe_working_set_t * data_set, pe_node_t * this_node)
1395 {
1396  pe_resource_t *rsc = this_node->details->remote_rsc;
1397  pe_resource_t *container = NULL;
1398  pe_node_t *host = NULL;
1399 
1400  /* If there is a node state entry for a (former) Pacemaker Remote node
1401  * but no resource creating that node, the node's connection resource will
1402  * be NULL. Consider it an offline remote node in that case.
1403  */
1404  if (rsc == NULL) {
1405  this_node->details->online = FALSE;
1406  goto remote_online_done;
1407  }
1408 
1409  container = rsc->container;
1410 
1411  if (container && pcmk__list_of_1(rsc->running_on)) {
1412  host = rsc->running_on->data;
1413  }
1414 
1415  /* If the resource is currently started, mark it online. */
1416  if (rsc->role == RSC_ROLE_STARTED) {
1417  crm_trace("%s node %s presumed ONLINE because connection resource is started",
1418  (container? "Guest" : "Remote"), this_node->details->id);
1419  this_node->details->online = TRUE;
1420  }
1421 
1422  /* consider this node shutting down if transitioning start->stop */
1423  if (rsc->role == RSC_ROLE_STARTED && rsc->next_role == RSC_ROLE_STOPPED) {
1424  crm_trace("%s node %s shutting down because connection resource is stopping",
1425  (container? "Guest" : "Remote"), this_node->details->id);
1426  this_node->details->shutdown = TRUE;
1427  }
1428 
1429  /* Now check all the failure conditions. */
1430  if(container && pcmk_is_set(container->flags, pe_rsc_failed)) {
1431  crm_trace("Guest node %s UNCLEAN because guest resource failed",
1432  this_node->details->id);
1433  this_node->details->online = FALSE;
1434  this_node->details->remote_requires_reset = TRUE;
1435 
1436  } else if (pcmk_is_set(rsc->flags, pe_rsc_failed)) {
1437  crm_trace("%s node %s OFFLINE because connection resource failed",
1438  (container? "Guest" : "Remote"), this_node->details->id);
1439  this_node->details->online = FALSE;
1440 
1441  } else if (rsc->role == RSC_ROLE_STOPPED
1442  || (container && container->role == RSC_ROLE_STOPPED)) {
1443 
1444  crm_trace("%s node %s OFFLINE because its resource is stopped",
1445  (container? "Guest" : "Remote"), this_node->details->id);
1446  this_node->details->online = FALSE;
1447  this_node->details->remote_requires_reset = FALSE;
1448 
1449  } else if (host && (host->details->online == FALSE)
1450  && host->details->unclean) {
1451  crm_trace("Guest node %s UNCLEAN because host is unclean",
1452  this_node->details->id);
1453  this_node->details->online = FALSE;
1454  this_node->details->remote_requires_reset = TRUE;
1455  }
1456 
1457 remote_online_done:
1458  crm_trace("Remote node %s online=%s",
1459  this_node->details->id, this_node->details->online ? "TRUE" : "FALSE");
1460 }
1461 
1462 static void
1463 determine_online_status(xmlNode * node_state, pe_node_t * this_node, pe_working_set_t * data_set)
1464 {
1465  gboolean online = FALSE;
1466  const char *exp_state = crm_element_value(node_state, XML_NODE_EXPECTED);
1467 
1468  CRM_CHECK(this_node != NULL, return);
1469 
1470  this_node->details->shutdown = FALSE;
1471  this_node->details->expected_up = FALSE;
1472 
1473  if (pe__shutdown_requested(this_node)) {
1474  this_node->details->shutdown = TRUE;
1475 
1476  } else if (pcmk__str_eq(exp_state, CRMD_JOINSTATE_MEMBER, pcmk__str_casei)) {
1477  this_node->details->expected_up = TRUE;
1478  }
1479 
1480  if (this_node->details->type == node_ping) {
1481  this_node->details->unclean = FALSE;
1482  online = FALSE; /* As far as resource management is concerned,
1483  * the node is safely offline.
1484  * Anyone caught abusing this logic will be shot
1485  */
1486 
1487  } else if (!pcmk_is_set(data_set->flags, pe_flag_stonith_enabled)) {
1488  online = determine_online_status_no_fencing(data_set, node_state, this_node);
1489 
1490  } else {
1491  online = determine_online_status_fencing(data_set, node_state, this_node);
1492  }
1493 
1494  if (online) {
1495  this_node->details->online = TRUE;
1496 
1497  } else {
1498  /* remove node from contention */
1499  this_node->fixed = TRUE;
1500  this_node->weight = -INFINITY;
1501  }
1502 
1503  if (online && this_node->details->shutdown) {
1504  /* don't run resources here */
1505  this_node->fixed = TRUE;
1506  this_node->weight = -INFINITY;
1507  }
1508 
1509  if (this_node->details->type == node_ping) {
1510  crm_info("Node %s is not a pacemaker node", this_node->details->uname);
1511 
1512  } else if (this_node->details->unclean) {
1513  pe_proc_warn("Node %s is unclean", this_node->details->uname);
1514 
1515  } else if (this_node->details->online) {
1516  crm_info("Node %s is %s", this_node->details->uname,
1517  this_node->details->shutdown ? "shutting down" :
1518  this_node->details->pending ? "pending" :
1519  this_node->details->standby ? "standby" :
1520  this_node->details->maintenance ? "maintenance" : "online");
1521 
1522  } else {
1523  crm_trace("Node %s is offline", this_node->details->uname);
1524  }
1525 }
1526 
1535 const char *
1536 pe_base_name_end(const char *id)
1537 {
1538  if (!pcmk__str_empty(id)) {
1539  const char *end = id + strlen(id) - 1;
1540 
1541  for (const char *s = end; s > id; --s) {
1542  switch (*s) {
1543  case '0':
1544  case '1':
1545  case '2':
1546  case '3':
1547  case '4':
1548  case '5':
1549  case '6':
1550  case '7':
1551  case '8':
1552  case '9':
1553  break;
1554  case ':':
1555  return (s == end)? s : (s - 1);
1556  default:
1557  return end;
1558  }
1559  }
1560  return end;
1561  }
1562  return NULL;
1563 }
1564 
1575 char *
1576 clone_strip(const char *last_rsc_id)
1577 {
1578  const char *end = pe_base_name_end(last_rsc_id);
1579  char *basename = NULL;
1580 
1581  CRM_ASSERT(end);
1582  basename = strndup(last_rsc_id, end - last_rsc_id + 1);
1583  CRM_ASSERT(basename);
1584  return basename;
1585 }
1586 
1597 char *
1598 clone_zero(const char *last_rsc_id)
1599 {
1600  const char *end = pe_base_name_end(last_rsc_id);
1601  size_t base_name_len = end - last_rsc_id + 1;
1602  char *zero = NULL;
1603 
1604  CRM_ASSERT(end);
1605  zero = calloc(base_name_len + 3, sizeof(char));
1606  CRM_ASSERT(zero);
1607  memcpy(zero, last_rsc_id, base_name_len);
1608  zero[base_name_len] = ':';
1609  zero[base_name_len + 1] = '0';
1610  return zero;
1611 }
1612 
1613 static pe_resource_t *
1614 create_fake_resource(const char *rsc_id, xmlNode * rsc_entry, pe_working_set_t * data_set)
1615 {
1616  pe_resource_t *rsc = NULL;
1617  xmlNode *xml_rsc = create_xml_node(NULL, XML_CIB_TAG_RESOURCE);
1618 
1619  copy_in_properties(xml_rsc, rsc_entry);
1620  crm_xml_add(xml_rsc, XML_ATTR_ID, rsc_id);
1621  crm_log_xml_debug(xml_rsc, "Orphan resource");
1622 
1623  if (!common_unpack(xml_rsc, &rsc, NULL, data_set)) {
1624  return NULL;
1625  }
1626 
1627  if (xml_contains_remote_node(xml_rsc)) {
1628  pe_node_t *node;
1629 
1630  crm_debug("Detected orphaned remote node %s", rsc_id);
1631  node = pe_find_node(data_set->nodes, rsc_id);
1632  if (node == NULL) {
1633  node = pe_create_node(rsc_id, rsc_id, "remote", NULL, data_set);
1634  }
1635  link_rsc2remotenode(data_set, rsc);
1636 
1637  if (node) {
1638  crm_trace("Setting node %s as shutting down due to orphaned connection resource", rsc_id);
1639  node->details->shutdown = TRUE;
1640  }
1641  }
1642 
1643  if (crm_element_value(rsc_entry, XML_RSC_ATTR_CONTAINER)) {
1644  /* This orphaned rsc needs to be mapped to a container. */
1645  crm_trace("Detected orphaned container filler %s", rsc_id);
1646  pe__set_resource_flags(rsc, pe_rsc_orphan_container_filler);
1647  }
1648  pe__set_resource_flags(rsc, pe_rsc_orphan);
1649  data_set->resources = g_list_append(data_set->resources, rsc);
1650  return rsc;
1651 }
1652 
1657 static pe_resource_t *
1658 create_anonymous_orphan(pe_resource_t *parent, const char *rsc_id,
1659  pe_node_t *node, pe_working_set_t *data_set)
1660 {
1661  pe_resource_t *top = pe__create_clone_child(parent, data_set);
1662 
1663  // find_rsc() because we might be a cloned group
1664  pe_resource_t *orphan = top->fns->find_rsc(top, rsc_id, NULL, pe_find_clone);
1665 
1666  pe_rsc_debug(parent, "Created orphan %s for %s: %s on %s",
1667  top->id, parent->id, rsc_id, node->details->uname);
1668  return orphan;
1669 }
1670 
1685 static pe_resource_t *
1686 find_anonymous_clone(pe_working_set_t * data_set, pe_node_t * node, pe_resource_t * parent,
1687  const char *rsc_id)
1688 {
1689  GListPtr rIter = NULL;
1690  pe_resource_t *rsc = NULL;
1691  pe_resource_t *inactive_instance = NULL;
1692  gboolean skip_inactive = FALSE;
1693 
1694  CRM_ASSERT(parent != NULL);
1695  CRM_ASSERT(pe_rsc_is_clone(parent));
1696  CRM_ASSERT(!pcmk_is_set(parent->flags, pe_rsc_unique));
1697 
1698  // Check for active (or partially active, for cloned groups) instance
1699  pe_rsc_trace(parent, "Looking for %s on %s in %s", rsc_id, node->details->uname, parent->id);
1700  for (rIter = parent->children; rsc == NULL && rIter; rIter = rIter->next) {
1701  GListPtr locations = NULL;
1702  pe_resource_t *child = rIter->data;
1703 
1704  /* Check whether this instance is already known to be active or pending
1705  * anywhere, at this stage of unpacking. Because this function is called
1706  * for a resource before the resource's individual operation history
1707  * entries are unpacked, locations will generally not contain the
1708  * desired node.
1709  *
1710  * However, there are three exceptions:
1711  * (1) when child is a cloned group and we have already unpacked the
1712  * history of another member of the group on the same node;
1713  * (2) when we've already unpacked the history of another numbered
1714  * instance on the same node (which can happen if globally-unique
1715  * was flipped from true to false); and
1716  * (3) when we re-run calculations on the same data set as part of a
1717  * simulation.
1718  */
1719  child->fns->location(child, &locations, 2);
1720  if (locations) {
1721  /* We should never associate the same numbered anonymous clone
1722  * instance with multiple nodes, and clone instances can't migrate,
1723  * so there must be only one location, regardless of history.
1724  */
1725  CRM_LOG_ASSERT(locations->next == NULL);
1726 
1727  if (((pe_node_t *)locations->data)->details == node->details) {
1728  /* This child instance is active on the requested node, so check
1729  * for a corresponding configured resource. We use find_rsc()
1730  * instead of child because child may be a cloned group, and we
1731  * need the particular member corresponding to rsc_id.
1732  *
1733  * If the history entry is orphaned, rsc will be NULL.
1734  */
1735  rsc = parent->fns->find_rsc(child, rsc_id, NULL, pe_find_clone);
1736  if (rsc) {
1737  /* If there are multiple instance history entries for an
1738  * anonymous clone in a single node's history (which can
1739  * happen if globally-unique is switched from true to
1740  * false), we want to consider the instances beyond the
1741  * first as orphans, even if there are inactive instance
1742  * numbers available.
1743  */
1744  if (rsc->running_on) {
1745  crm_notice("Active (now-)anonymous clone %s has "
1746  "multiple (orphan) instance histories on %s",
1747  parent->id, node->details->uname);
1748  skip_inactive = TRUE;
1749  rsc = NULL;
1750  } else {
1751  pe_rsc_trace(parent, "Resource %s, active", rsc->id);
1752  }
1753  }
1754  }
1755  g_list_free(locations);
1756 
1757  } else {
1758  pe_rsc_trace(parent, "Resource %s, skip inactive", child->id);
1759  if (!skip_inactive && !inactive_instance
1760  && !pcmk_is_set(child->flags, pe_rsc_block)) {
1761  // Remember one inactive instance in case we don't find active
1762  inactive_instance = parent->fns->find_rsc(child, rsc_id, NULL,
1763  pe_find_clone);
1764 
1765  /* ... but don't use it if it was already associated with a
1766  * pending action on another node
1767  */
1768  if (inactive_instance && inactive_instance->pending_node
1769  && (inactive_instance->pending_node->details != node->details)) {
1770  inactive_instance = NULL;
1771  }
1772  }
1773  }
1774  }
1775 
1776  if ((rsc == NULL) && !skip_inactive && (inactive_instance != NULL)) {
1777  pe_rsc_trace(parent, "Resource %s, empty slot", inactive_instance->id);
1778  rsc = inactive_instance;
1779  }
1780 
1781  /* If the resource has "requires" set to "quorum" or "nothing", and we don't
1782  * have a clone instance for every node, we don't want to consume a valid
1783  * instance number for unclean nodes. Such instances may appear to be active
1784  * according to the history, but should be considered inactive, so we can
1785  * start an instance elsewhere. Treat such instances as orphans.
1786  *
1787  * An exception is instances running on guest nodes -- since guest node
1788  * "fencing" is actually just a resource stop, requires shouldn't apply.
1789  *
1790  * @TODO Ideally, we'd use an inactive instance number if it is not needed
1791  * for any clean instances. However, we don't know that at this point.
1792  */
1793  if ((rsc != NULL) && !pcmk_is_set(rsc->flags, pe_rsc_needs_fencing)
1794  && (!node->details->online || node->details->unclean)
1795  && !pe__is_guest_node(node)
1796  && !pe__is_universal_clone(parent, data_set)) {
1797 
1798  rsc = NULL;
1799  }
1800 
1801  if (rsc == NULL) {
1802  rsc = create_anonymous_orphan(parent, rsc_id, node, data_set);
1803  pe_rsc_trace(parent, "Resource %s, orphan", rsc->id);
1804  }
1805  return rsc;
1806 }
1807 
1808 static pe_resource_t *
1809 unpack_find_resource(pe_working_set_t * data_set, pe_node_t * node, const char *rsc_id,
1810  xmlNode * rsc_entry)
1811 {
1812  pe_resource_t *rsc = NULL;
1813  pe_resource_t *parent = NULL;
1814 
1815  crm_trace("looking for %s", rsc_id);
1816  rsc = pe_find_resource(data_set->resources, rsc_id);
1817 
1818  if (rsc == NULL) {
1819  /* If we didn't find the resource by its name in the operation history,
1820  * check it again as a clone instance. Even when clone-max=0, we create
1821  * a single :0 orphan to match against here.
1822  */
1823  char *clone0_id = clone_zero(rsc_id);
1824  pe_resource_t *clone0 = pe_find_resource(data_set->resources, clone0_id);
1825 
1826  if (clone0 && !pcmk_is_set(clone0->flags, pe_rsc_unique)) {
1827  rsc = clone0;
1828  parent = uber_parent(clone0);
1829  crm_trace("%s found as %s (%s)", rsc_id, clone0_id, parent->id);
1830  } else {
1831  crm_trace("%s is not known as %s either (orphan)",
1832  rsc_id, clone0_id);
1833  }
1834  free(clone0_id);
1835 
1836  } else if (rsc->variant > pe_native) {
1837  crm_trace("Resource history for %s is orphaned because it is no longer primitive",
1838  rsc_id);
1839  return NULL;
1840 
1841  } else {
1842  parent = uber_parent(rsc);
1843  }
1844 
1845  if (pe_rsc_is_anon_clone(parent)) {
1846 
1847  if (pe_rsc_is_bundled(parent)) {
1848  rsc = pe__find_bundle_replica(parent->parent, node);
1849  } else {
1850  char *base = clone_strip(rsc_id);
1851 
1852  rsc = find_anonymous_clone(data_set, node, parent, base);
1853  free(base);
1854  CRM_ASSERT(rsc != NULL);
1855  }
1856  }
1857 
1858  if (rsc && !pcmk__str_eq(rsc_id, rsc->id, pcmk__str_casei)
1859  && !pcmk__str_eq(rsc_id, rsc->clone_name, pcmk__str_casei)) {
1860 
1861  free(rsc->clone_name);
1862  rsc->clone_name = strdup(rsc_id);
1863  pe_rsc_debug(rsc, "Internally renamed %s on %s to %s%s",
1864  rsc_id, node->details->uname, rsc->id,
1865  (pcmk_is_set(rsc->flags, pe_rsc_orphan)? " (ORPHAN)" : ""));
1866  }
1867  return rsc;
1868 }
1869 
1870 static pe_resource_t *
1871 process_orphan_resource(xmlNode * rsc_entry, pe_node_t * node, pe_working_set_t * data_set)
1872 {
1873  pe_resource_t *rsc = NULL;
1874  const char *rsc_id = crm_element_value(rsc_entry, XML_ATTR_ID);
1875 
1876  crm_debug("Detected orphan resource %s on %s", rsc_id, node->details->uname);
1877  rsc = create_fake_resource(rsc_id, rsc_entry, data_set);
1878 
1879  if (!pcmk_is_set(data_set->flags, pe_flag_stop_rsc_orphans)) {
1880  pe__clear_resource_flags(rsc, pe_rsc_managed);
1881 
1882  } else {
1883  CRM_CHECK(rsc != NULL, return NULL);
1884  pe_rsc_trace(rsc, "Added orphan %s", rsc->id);
1885  resource_location(rsc, NULL, -INFINITY, "__orphan_do_not_run__", data_set);
1886  }
1887  return rsc;
1888 }
1889 
1890 static void
1891 process_rsc_state(pe_resource_t * rsc, pe_node_t * node,
1892  enum action_fail_response on_fail,
1893  xmlNode * migrate_op, pe_working_set_t * data_set)
1894 {
1895  pe_node_t *tmpnode = NULL;
1896  char *reason = NULL;
1897 
1898  CRM_ASSERT(rsc);
1899  pe_rsc_trace(rsc, "Resource %s is %s on %s: on_fail=%s",
1900  rsc->id, role2text(rsc->role), node->details->uname, fail2text(on_fail));
1901 
1902  /* process current state */
1903  if (rsc->role != RSC_ROLE_UNKNOWN) {
1904  pe_resource_t *iter = rsc;
1905 
1906  while (iter) {
1907  if (g_hash_table_lookup(iter->known_on, node->details->id) == NULL) {
1908  pe_node_t *n = pe__copy_node(node);
1909 
1910  pe_rsc_trace(rsc, "%s (aka. %s) known on %s", rsc->id, rsc->clone_name,
1911  n->details->uname);
1912  g_hash_table_insert(iter->known_on, (gpointer) n->details->id, n);
1913  }
1914  if (pcmk_is_set(iter->flags, pe_rsc_unique)) {
1915  break;
1916  }
1917  iter = iter->parent;
1918  }
1919  }
1920 
1921  /* If a managed resource is believed to be running, but node is down ... */
1922  if (rsc->role > RSC_ROLE_STOPPED
1923  && node->details->online == FALSE
1924  && node->details->maintenance == FALSE
1925  && pcmk_is_set(rsc->flags, pe_rsc_managed)) {
1926 
1927  gboolean should_fence = FALSE;
1928 
1929  /* If this is a guest node, fence it (regardless of whether fencing is
1930  * enabled, because guest node fencing is done by recovery of the
1931  * container resource rather than by the fencer). Mark the resource
1932  * we're processing as failed. When the guest comes back up, its
1933  * operation history in the CIB will be cleared, freeing the affected
1934  * resource to run again once we are sure we know its state.
1935  */
1936  if (pe__is_guest_node(node)) {
1937  pe__set_resource_flags(rsc, pe_rsc_failed|pe_rsc_stop);
1938  should_fence = TRUE;
1939 
1940  } else if (pcmk_is_set(data_set->flags, pe_flag_stonith_enabled)) {
1941  if (pe__is_remote_node(node) && node->details->remote_rsc
1942  && !pcmk_is_set(node->details->remote_rsc->flags, pe_rsc_failed)) {
1943 
1944  /* Setting unseen means that fencing of the remote node will
1945  * occur only if the connection resource is not going to start
1946  * somewhere. This allows connection resources on a failed
1947  * cluster node to move to another node without requiring the
1948  * remote nodes to be fenced as well.
1949  */
1950  node->details->unseen = TRUE;
1951  reason = crm_strdup_printf("%s is active there (fencing will be"
1952  " revoked if remote connection can "
1953  "be re-established elsewhere)",
1954  rsc->id);
1955  }
1956  should_fence = TRUE;
1957  }
1958 
1959  if (should_fence) {
1960  if (reason == NULL) {
1961  reason = crm_strdup_printf("%s is thought to be active there", rsc->id);
1962  }
1963  pe_fence_node(data_set, node, reason, FALSE);
1964  }
1965  free(reason);
1966  }
1967 
1968  if (node->details->unclean) {
1969  /* No extra processing needed
1970  * Also allows resources to be started again after a node is shot
1971  */
1972  on_fail = action_fail_ignore;
1973  }
1974 
1975  switch (on_fail) {
1976  case action_fail_ignore:
1977  /* nothing to do */
1978  break;
1979 
1980  case action_fail_demote:
1981  pe__set_resource_flags(rsc, pe_rsc_failed);
1982  demote_action(rsc, node, FALSE);
1983  break;
1984 
1985  case action_fail_fence:
1986  /* treat it as if it is still running
1987  * but also mark the node as unclean
1988  */
1989  reason = crm_strdup_printf("%s failed there", rsc->id);
1990  pe_fence_node(data_set, node, reason, FALSE);
1991  free(reason);
1992  break;
1993 
1994  case action_fail_standby:
1995  node->details->standby = TRUE;
1996  node->details->standby_onfail = TRUE;
1997  break;
1998 
1999  case action_fail_block:
2000  /* is_managed == FALSE will prevent any
2001  * actions being sent for the resource
2002  */
2003  pe__clear_resource_flags(rsc, pe_rsc_managed);
2004  pe__set_resource_flags(rsc, pe_rsc_block);
2005  break;
2006 
2007  case action_fail_migrate:
2008  /* make sure it comes up somewhere else
2009  * or not at all
2010  */
2011  resource_location(rsc, node, -INFINITY, "__action_migration_auto__", data_set);
2012  break;
2013 
2014  case action_fail_stop:
2015  rsc->next_role = RSC_ROLE_STOPPED;
2016  break;
2017 
2018  case action_fail_recover:
2019  if (rsc->role != RSC_ROLE_STOPPED && rsc->role != RSC_ROLE_UNKNOWN) {
2020  pe__set_resource_flags(rsc, pe_rsc_failed|pe_rsc_stop);
2021  stop_action(rsc, node, FALSE);
2022  }
2023  break;
2024 
2025  case action_fail_restart_container:
2026  pe__set_resource_flags(rsc, pe_rsc_failed|pe_rsc_stop);
2027  if (rsc->container && pe_rsc_is_bundled(rsc)) {
2028  /* A bundle's remote connection can run on a different node than
2029  * the bundle's container. We don't necessarily know where the
2030  * container is running yet, so remember it and add a stop
2031  * action for it later.
2032  */
2033  data_set->stop_needed = g_list_prepend(data_set->stop_needed,
2034  rsc->container);
2035  } else if (rsc->container) {
2036  stop_action(rsc->container, node, FALSE);
2037  } else if (rsc->role != RSC_ROLE_STOPPED && rsc->role != RSC_ROLE_UNKNOWN) {
2038  stop_action(rsc, node, FALSE);
2039  }
2040  break;
2041 
2042  case action_fail_reset_remote:
2043  pe__set_resource_flags(rsc, pe_rsc_failed|pe_rsc_stop);
2044  if (pcmk_is_set(data_set->flags, pe_flag_stonith_enabled)) {
2045  tmpnode = NULL;
2046  if (rsc->is_remote_node) {
2047  tmpnode = pe_find_node(data_set->nodes, rsc->id);
2048  }
2049  if (tmpnode &&
2050  pe__is_remote_node(tmpnode) &&
2051  tmpnode->details->remote_was_fenced == 0) {
2052 
2053  /* The remote connection resource failed in a way that
2054  * should result in fencing the remote node.
2055  */
2056  pe_fence_node(data_set, tmpnode,
2057  "remote connection is unrecoverable", FALSE);
2058  }
2059  }
2060 
2061  /* require the stop action regardless if fencing is occurring or not. */
2062  if (rsc->role > RSC_ROLE_STOPPED) {
2063  stop_action(rsc, node, FALSE);
2064  }
2065 
2066  /* if reconnect delay is in use, prevent the connection from exiting the
2067  * "STOPPED" role until the failure is cleared by the delay timeout. */
2068  if (rsc->remote_reconnect_ms) {
2069  rsc->next_role = RSC_ROLE_STOPPED;
2070  }
2071  break;
2072  }
2073 
2074  /* ensure a remote-node connection failure forces an unclean remote-node
2075  * to be fenced. By setting unseen = FALSE, the remote-node failure will
2076  * result in a fencing operation regardless if we're going to attempt to
2077  * reconnect to the remote-node in this transition or not. */
2078  if (pcmk_is_set(rsc->flags, pe_rsc_failed) && rsc->is_remote_node) {
2079  tmpnode = pe_find_node(data_set->nodes, rsc->id);
2080  if (tmpnode && tmpnode->details->unclean) {
2081  tmpnode->details->unseen = FALSE;
2082  }
2083  }
2084 
2085  if (rsc->role != RSC_ROLE_STOPPED && rsc->role != RSC_ROLE_UNKNOWN) {
2086  if (pcmk_is_set(rsc->flags, pe_rsc_orphan)) {
2087  if (pcmk_is_set(rsc->flags, pe_rsc_managed)) {
2088  pcmk__config_warn("Detected active orphan %s running on %s",
2089  rsc->id, node->details->uname);
2090  } else {
2091  pcmk__config_warn("Resource '%s' must be stopped manually on "
2092  "%s because cluster is configured not to "
2093  "stop active orphans",
2094  rsc->id, node->details->uname);
2095  }
2096  }
2097 
2098  native_add_running(rsc, node, data_set);
2099  switch (on_fail) {
2100  case action_fail_ignore:
2101  break;
2102  case action_fail_demote:
2103  case action_fail_block:
2104  pe__set_resource_flags(rsc, pe_rsc_failed);
2105  break;
2106  default:
2107  pe__set_resource_flags(rsc, pe_rsc_failed|pe_rsc_stop);
2108  break;
2109  }
2110 
2111  } else if (rsc->clone_name && strchr(rsc->clone_name, ':') != NULL) {
2112  /* Only do this for older status sections that included instance numbers
2113  * Otherwise stopped instances will appear as orphans
2114  */
2115  pe_rsc_trace(rsc, "Resetting clone_name %s for %s (stopped)", rsc->clone_name, rsc->id);
2116  free(rsc->clone_name);
2117  rsc->clone_name = NULL;
2118 
2119  } else {
2120  GList *possible_matches = pe__resource_actions(rsc, node, RSC_STOP,
2121  FALSE);
2122  GListPtr gIter = possible_matches;
2123 
2124  for (; gIter != NULL; gIter = gIter->next) {
2125  pe_action_t *stop = (pe_action_t *) gIter->data;
2126 
2127  pe__set_action_flags(stop, pe_action_optional);
2128  }
2129 
2130  g_list_free(possible_matches);
2131  }
2132 }
2133 
2134 /* create active recurring operations as optional */
2135 static void
2136 process_recurring(pe_node_t * node, pe_resource_t * rsc,
2137  int start_index, int stop_index,
2138  GListPtr sorted_op_list, pe_working_set_t * data_set)
2139 {
2140  int counter = -1;
2141  const char *task = NULL;
2142  const char *status = NULL;
2143  GListPtr gIter = sorted_op_list;
2144 
2145  CRM_ASSERT(rsc);
2146  pe_rsc_trace(rsc, "%s: Start index %d, stop index = %d", rsc->id, start_index, stop_index);
2147 
2148  for (; gIter != NULL; gIter = gIter->next) {
2149  xmlNode *rsc_op = (xmlNode *) gIter->data;
2150 
2151  guint interval_ms = 0;
2152  char *key = NULL;
2153  const char *id = ID(rsc_op);
2154 
2155  counter++;
2156 
2157  if (node->details->online == FALSE) {
2158  pe_rsc_trace(rsc, "Skipping %s/%s: node is offline", rsc->id, node->details->uname);
2159  break;
2160 
2161  /* Need to check if there's a monitor for role="Stopped" */
2162  } else if (start_index < stop_index && counter <= stop_index) {
2163  pe_rsc_trace(rsc, "Skipping %s/%s: resource is not active", id, node->details->uname);
2164  continue;
2165 
2166  } else if (counter < start_index) {
2167  pe_rsc_trace(rsc, "Skipping %s/%s: old %d", id, node->details->uname, counter);
2168  continue;
2169  }
2170 
2171  crm_element_value_ms(rsc_op, XML_LRM_ATTR_INTERVAL_MS, &interval_ms);
2172  if (interval_ms == 0) {
2173  pe_rsc_trace(rsc, "Skipping %s/%s: non-recurring", id, node->details->uname);
2174  continue;
2175  }
2176 
2177  status = crm_element_value(rsc_op, XML_LRM_ATTR_OPSTATUS);
2178  if (pcmk__str_eq(status, "-1", pcmk__str_casei)) {
2179  pe_rsc_trace(rsc, "Skipping %s/%s: status", id, node->details->uname);
2180  continue;
2181  }
2182  task = crm_element_value(rsc_op, XML_LRM_ATTR_TASK);
2183  /* create the action */
2184  key = pcmk__op_key(rsc->id, task, interval_ms);
2185  pe_rsc_trace(rsc, "Creating %s/%s", key, node->details->uname);
2186  custom_action(rsc, key, task, node, TRUE, TRUE, data_set);
2187  }
2188 }
2189 
2190 void
2191 calculate_active_ops(GListPtr sorted_op_list, int *start_index, int *stop_index)
2192 {
2193  int counter = -1;
2194  int implied_monitor_start = -1;
2195  int implied_clone_start = -1;
2196  const char *task = NULL;
2197  const char *status = NULL;
2198  GListPtr gIter = sorted_op_list;
2199 
2200  *stop_index = -1;
2201  *start_index = -1;
2202 
2203  for (; gIter != NULL; gIter = gIter->next) {
2204  xmlNode *rsc_op = (xmlNode *) gIter->data;
2205 
2206  counter++;
2207 
2208  task = crm_element_value(rsc_op, XML_LRM_ATTR_TASK);
2209  status = crm_element_value(rsc_op, XML_LRM_ATTR_OPSTATUS);
2210 
2211  if (pcmk__str_eq(task, CRMD_ACTION_STOP, pcmk__str_casei)
2212  && pcmk__str_eq(status, "0", pcmk__str_casei)) {
2213  *stop_index = counter;
2214 
2215  } else if (pcmk__strcase_any_of(task, CRMD_ACTION_START, CRMD_ACTION_MIGRATED, NULL)) {
2216  *start_index = counter;
2217 
2218  } else if ((implied_monitor_start <= *stop_index) && pcmk__str_eq(task, CRMD_ACTION_STATUS, pcmk__str_casei)) {
2219  const char *rc = crm_element_value(rsc_op, XML_LRM_ATTR_RC);
2220 
2221  if (pcmk__strcase_any_of(rc, "0", "8", NULL)) {
2222  implied_monitor_start = counter;
2223  }
2224  } else if (pcmk__strcase_any_of(task, CRMD_ACTION_PROMOTE, CRMD_ACTION_DEMOTE, NULL)) {
2225  implied_clone_start = counter;
2226  }
2227  }
2228 
2229  if (*start_index == -1) {
2230  if (implied_clone_start != -1) {
2231  *start_index = implied_clone_start;
2232  } else if (implied_monitor_start != -1) {
2233  *start_index = implied_monitor_start;
2234  }
2235  }
2236 }
2237 
2238 // If resource history entry has shutdown lock, remember lock node and time
2239 static void
2240 unpack_shutdown_lock(xmlNode *rsc_entry, pe_resource_t *rsc, pe_node_t *node,
2241  pe_working_set_t *data_set)
2242 {
2243  time_t lock_time = 0; // When lock started (i.e. node shutdown time)
2244 
2245  if ((crm_element_value_epoch(rsc_entry, XML_CONFIG_ATTR_SHUTDOWN_LOCK,
2246  &lock_time) == pcmk_ok) && (lock_time != 0)) {
2247 
2248  if ((data_set->shutdown_lock > 0)
2249  && (get_effective_time(data_set)
2250  > (lock_time + data_set->shutdown_lock))) {
2251  pe_rsc_info(rsc, "Shutdown lock for %s on %s expired",
2252  rsc->id, node->details->uname);
2253  pe__clear_resource_history(rsc, node, data_set);
2254  } else {
2255  rsc->lock_node = node;
2256  rsc->lock_time = lock_time;
2257  }
2258  }
2259 }
2260 
2261 static pe_resource_t *
2262 unpack_lrm_rsc_state(pe_node_t * node, xmlNode * rsc_entry, pe_working_set_t * data_set)
2263 {
2264  GListPtr gIter = NULL;
2265  int stop_index = -1;
2266  int start_index = -1;
2267  enum rsc_role_e req_role = RSC_ROLE_UNKNOWN;
2268 
2269  const char *task = NULL;
2270  const char *rsc_id = crm_element_value(rsc_entry, XML_ATTR_ID);
2271 
2272  pe_resource_t *rsc = NULL;
2273  GListPtr op_list = NULL;
2274  GListPtr sorted_op_list = NULL;
2275 
2276  xmlNode *migrate_op = NULL;
2277  xmlNode *rsc_op = NULL;
2278  xmlNode *last_failure = NULL;
2279 
2280  enum action_fail_response on_fail = action_fail_ignore;
2281  enum rsc_role_e saved_role = RSC_ROLE_UNKNOWN;
2282 
2283  crm_trace("[%s] Processing %s on %s",
2284  crm_element_name(rsc_entry), rsc_id, node->details->uname);
2285 
2286  /* extract operations */
2287  op_list = NULL;
2288  sorted_op_list = NULL;
2289 
2290  for (rsc_op = pcmk__xe_first_child(rsc_entry); rsc_op != NULL;
2291  rsc_op = pcmk__xe_next(rsc_op)) {
2292 
2293  if (pcmk__str_eq((const char *)rsc_op->name, XML_LRM_TAG_RSC_OP,
2294  pcmk__str_none)) {
2295  op_list = g_list_prepend(op_list, rsc_op);
2296  }
2297  }
2298 
2299  if (!pcmk_is_set(data_set->flags, pe_flag_shutdown_lock)) {
2300  if (op_list == NULL) {
2301  // If there are no operations, there is nothing to do
2302  return NULL;
2303  }
2304  }
2305 
2306  /* find the resource */
2307  rsc = unpack_find_resource(data_set, node, rsc_id, rsc_entry);
2308  if (rsc == NULL) {
2309  if (op_list == NULL) {
2310  // If there are no operations, there is nothing to do
2311  return NULL;
2312  } else {
2313  rsc = process_orphan_resource(rsc_entry, node, data_set);
2314  }
2315  }
2316  CRM_ASSERT(rsc != NULL);
2317 
2318  // Check whether the resource is "shutdown-locked" to this node
2319  if (pcmk_is_set(data_set->flags, pe_flag_shutdown_lock)) {
2320  unpack_shutdown_lock(rsc_entry, rsc, node, data_set);
2321  }
2322 
2323  /* process operations */
2324  saved_role = rsc->role;
2325  rsc->role = RSC_ROLE_UNKNOWN;
2326  sorted_op_list = g_list_sort(op_list, sort_op_by_callid);
2327 
2328  for (gIter = sorted_op_list; gIter != NULL; gIter = gIter->next) {
2329  xmlNode *rsc_op = (xmlNode *) gIter->data;
2330 
2331  task = crm_element_value(rsc_op, XML_LRM_ATTR_TASK);
2332  if (pcmk__str_eq(task, CRMD_ACTION_MIGRATED, pcmk__str_casei)) {
2333  migrate_op = rsc_op;
2334  }
2335 
2336  unpack_rsc_op(rsc, node, rsc_op, &last_failure, &on_fail, data_set);
2337  }
2338 
2339  /* create active recurring operations as optional */
2340  calculate_active_ops(sorted_op_list, &start_index, &stop_index);
2341  process_recurring(node, rsc, start_index, stop_index, sorted_op_list, data_set);
2342 
2343  /* no need to free the contents */
2344  g_list_free(sorted_op_list);
2345 
2346  process_rsc_state(rsc, node, on_fail, migrate_op, data_set);
2347 
2348  if (get_target_role(rsc, &req_role)) {
2349  if (rsc->next_role == RSC_ROLE_UNKNOWN || req_role < rsc->next_role) {
2350  pe_rsc_debug(rsc, "%s: Overwriting calculated next role %s"
2351  " with requested next role %s",
2352  rsc->id, role2text(rsc->next_role), role2text(req_role));
2353  rsc->next_role = req_role;
2354 
2355  } else if (req_role > rsc->next_role) {
2356  pe_rsc_info(rsc, "%s: Not overwriting calculated next role %s"
2357  " with requested next role %s",
2358  rsc->id, role2text(rsc->next_role), role2text(req_role));
2359  }
2360  }
2361 
2362  if (saved_role > rsc->role) {
2363  rsc->role = saved_role;
2364  }
2365 
2366  return rsc;
2367 }
2368 
2369 static void
2370 handle_orphaned_container_fillers(xmlNode * lrm_rsc_list, pe_working_set_t * data_set)
2371 {
2372  xmlNode *rsc_entry = NULL;
2373  for (rsc_entry = pcmk__xe_first_child(lrm_rsc_list); rsc_entry != NULL;
2374  rsc_entry = pcmk__xe_next(rsc_entry)) {
2375 
2376  pe_resource_t *rsc;
2377  pe_resource_t *container;
2378  const char *rsc_id;
2379  const char *container_id;
2380 
2381  if (!pcmk__str_eq((const char *)rsc_entry->name, XML_LRM_TAG_RESOURCE, pcmk__str_casei)) {
2382  continue;
2383  }
2384 
2385  container_id = crm_element_value(rsc_entry, XML_RSC_ATTR_CONTAINER);
2386  rsc_id = crm_element_value(rsc_entry, XML_ATTR_ID);
2387  if (container_id == NULL || rsc_id == NULL) {
2388  continue;
2389  }
2390 
2391  container = pe_find_resource(data_set->resources, container_id);
2392  if (container == NULL) {
2393  continue;
2394  }
2395 
2396  rsc = pe_find_resource(data_set->resources, rsc_id);
2397  if (rsc == NULL ||
2398  !pcmk_is_set(rsc->flags, pe_rsc_orphan_container_filler) ||
2399  rsc->container != NULL) {
2400  continue;
2401  }
2402 
2403  pe_rsc_trace(rsc, "Mapped container of orphaned resource %s to %s",
2404  rsc->id, container_id);
2405  rsc->container = container;
2406  container->fillers = g_list_append(container->fillers, rsc);
2407  }
2408 }
2409 
2410 static void
2411 unpack_lrm_resources(pe_node_t *node, xmlNode *lrm_rsc_list,
2412  pe_working_set_t *data_set)
2413 {
2414  xmlNode *rsc_entry = NULL;
2415  gboolean found_orphaned_container_filler = FALSE;
2416 
2417  for (rsc_entry = pcmk__xe_first_child(lrm_rsc_list); rsc_entry != NULL;
2418  rsc_entry = pcmk__xe_next(rsc_entry)) {
2419 
2420  if (pcmk__str_eq((const char *)rsc_entry->name, XML_LRM_TAG_RESOURCE, pcmk__str_none)) {
2421  pe_resource_t *rsc = unpack_lrm_rsc_state(node, rsc_entry, data_set);
2422  if (!rsc) {
2423  continue;
2424  }
2425  if (pcmk_is_set(rsc->flags, pe_rsc_orphan_container_filler)) {
2426  found_orphaned_container_filler = TRUE;
2427  }
2428  }
2429  }
2430 
2431  /* now that all the resource state has been unpacked for this node
2432  * we have to go back and map any orphaned container fillers to their
2433  * container resource */
2434  if (found_orphaned_container_filler) {
2435  handle_orphaned_container_fillers(lrm_rsc_list, data_set);
2436  }
2437 }
2438 
2439 static void
2440 set_active(pe_resource_t * rsc)
2441 {
2442  pe_resource_t *top = uber_parent(rsc);
2443 
2444  if (top && pcmk_is_set(top->flags, pe_rsc_promotable)) {
2445  rsc->role = RSC_ROLE_SLAVE;
2446  } else {
2447  rsc->role = RSC_ROLE_STARTED;
2448  }
2449 }
2450 
2451 static void
2452 set_node_score(gpointer key, gpointer value, gpointer user_data)
2453 {
2454  pe_node_t *node = value;
2455  int *score = user_data;
2456 
2457  node->weight = *score;
2458 }
2459 
2460 #define STATUS_PATH_MAX 1024
2461 static xmlNode *
2462 find_lrm_op(const char *resource, const char *op, const char *node, const char *source,
2463  bool success_only, pe_working_set_t *data_set)
2464 {
2465  int offset = 0;
2466  char xpath[STATUS_PATH_MAX];
2467  xmlNode *xml = NULL;
2468 
2469  offset += snprintf(xpath + offset, STATUS_PATH_MAX - offset, "//node_state[@uname='%s']", node);
2470  offset +=
2471  snprintf(xpath + offset, STATUS_PATH_MAX - offset, "//" XML_LRM_TAG_RESOURCE "[@id='%s']",
2472  resource);
2473 
2474  /* Need to check against transition_magic too? */
2475  if (source && pcmk__str_eq(op, CRMD_ACTION_MIGRATE, pcmk__str_casei)) {
2476  offset +=
2477  snprintf(xpath + offset, STATUS_PATH_MAX - offset,
2478  "/" XML_LRM_TAG_RSC_OP "[@operation='%s' and @migrate_target='%s']", op,
2479  source);
2480  } else if (source && pcmk__str_eq(op, CRMD_ACTION_MIGRATED, pcmk__str_casei)) {
2481  offset +=
2482  snprintf(xpath + offset, STATUS_PATH_MAX - offset,
2483  "/" XML_LRM_TAG_RSC_OP "[@operation='%s' and @migrate_source='%s']", op,
2484  source);
2485  } else {
2486  offset +=
2487  snprintf(xpath + offset, STATUS_PATH_MAX - offset,
2488  "/" XML_LRM_TAG_RSC_OP "[@operation='%s']", op);
2489  }
2490 
2491  CRM_LOG_ASSERT(offset > 0);
2492  xml = get_xpath_object(xpath, data_set->input, LOG_DEBUG);
2493 
2494  if (xml && success_only) {
2495  int rc = PCMK_OCF_UNKNOWN_ERROR;
2496  int status = PCMK_LRM_OP_ERROR;
2497 
2498  crm_element_value_int(xml, XML_LRM_ATTR_RC, &rc);
2499  crm_element_value_int(xml, XML_LRM_ATTR_OPSTATUS, &status);
2500  if ((rc != PCMK_OCF_OK) || (status != PCMK_LRM_OP_DONE)) {
2501  return NULL;
2502  }
2503  }
2504  return xml;
2505 }
2506 
2507 static int
2508 pe__call_id(xmlNode *op_xml)
2509 {
2510  int id = 0;
2511 
2512  if (op_xml) {
2513  crm_element_value_int(op_xml, XML_LRM_ATTR_CALLID, &id);
2514  }
2515  return id;
2516 }
2517 
2534 static bool
2535 stop_happened_after(pe_resource_t *rsc, pe_node_t *node, xmlNode *xml_op,
2536  pe_working_set_t *data_set)
2537 {
2538  xmlNode *stop_op = find_lrm_op(rsc->id, CRMD_ACTION_STOP,
2539  node->details->uname, NULL, TRUE, data_set);
2540 
2541  return (stop_op && (pe__call_id(stop_op) > pe__call_id(xml_op)));
2542 }
2543 
2544 static void
2545 unpack_migrate_to_success(pe_resource_t *rsc, pe_node_t *node, xmlNode *xml_op,
2546  pe_working_set_t *data_set)
2547 {
2548  /* A successful migration sequence is:
2549  * migrate_to on source node
2550  * migrate_from on target node
2551  * stop on source node
2552  *
2553  * If a migrate_to is followed by a stop, the entire migration (successful
2554  * or failed) is complete, and we don't care what happened on the target.
2555  *
2556  * If no migrate_from has happened, the migration is considered to be
2557  * "partial". If the migrate_from failed, make sure the resource gets
2558  * stopped on both source and target (if up).
2559  *
2560  * If the migrate_to and migrate_from both succeeded (which also implies the
2561  * resource is no longer running on the source), but there is no stop, the
2562  * migration is considered to be "dangling". Schedule a stop on the source
2563  * in this case.
2564  */
2565  int from_rc = 0;
2566  int from_status = 0;
2567  pe_node_t *target_node = NULL;
2568  pe_node_t *source_node = NULL;
2569  xmlNode *migrate_from = NULL;
2570  const char *source = crm_element_value(xml_op, XML_LRM_ATTR_MIGRATE_SOURCE);
2571  const char *target = crm_element_value(xml_op, XML_LRM_ATTR_MIGRATE_TARGET);
2572 
2573  // Sanity check
2574  CRM_CHECK(source && target && !strcmp(source, node->details->uname), return);
2575 
2576  if (stop_happened_after(rsc, node, xml_op, data_set)) {
2577  return;
2578  }
2579 
2580  // Clones are not allowed to migrate, so role can't be master
2581  rsc->role = RSC_ROLE_STARTED;
2582 
2583  target_node = pe_find_node(data_set->nodes, target);
2584  source_node = pe_find_node(data_set->nodes, source);
2585 
2586  // Check whether there was a migrate_from action on the target
2587  migrate_from = find_lrm_op(rsc->id, CRMD_ACTION_MIGRATED, target,
2588  source, FALSE, data_set);
2589  if (migrate_from) {
2590  crm_element_value_int(migrate_from, XML_LRM_ATTR_RC, &from_rc);
2591  crm_element_value_int(migrate_from, XML_LRM_ATTR_OPSTATUS, &from_status);
2592  pe_rsc_trace(rsc, "%s op on %s exited with status=%d, rc=%d",
2593  ID(migrate_from), target, from_status, from_rc);
2594  }
2595 
2596  if (migrate_from && from_rc == PCMK_OCF_OK
2597  && from_status == PCMK_LRM_OP_DONE) {
2598  /* The migrate_to and migrate_from both succeeded, so mark the migration
2599  * as "dangling". This will be used to schedule a stop action on the
2600  * source without affecting the target.
2601  */
2602  pe_rsc_trace(rsc, "Detected dangling migration op: %s on %s", ID(xml_op),
2603  source);
2604  rsc->role = RSC_ROLE_STOPPED;
2605  rsc->dangling_migrations = g_list_prepend(rsc->dangling_migrations, node);
2606 
2607  } else if (migrate_from && (from_status != PCMK_LRM_OP_PENDING)) { // Failed
2608  if (target_node && target_node->details->online) {
2609  pe_rsc_trace(rsc, "Marking active on %s %p %d", target, target_node,
2610  target_node->details->online);
2611  native_add_running(rsc, target_node, data_set);
2612  }
2613 
2614  } else { // Pending, or complete but erased
2615  if (target_node && target_node->details->online) {
2616  pe_rsc_trace(rsc, "Marking active on %s %p %d", target, target_node,
2617  target_node->details->online);
2618 
2619  native_add_running(rsc, target_node, data_set);
2620  if (source_node && source_node->details->online) {
2621  /* This is a partial migration: the migrate_to completed
2622  * successfully on the source, but the migrate_from has not
2623  * completed. Remember the source and target; if the newly
2624  * chosen target remains the same when we schedule actions
2625  * later, we may continue with the migration.
2626  */
2627  rsc->partial_migration_target = target_node;
2628  rsc->partial_migration_source = source_node;
2629  }
2630  } else {
2631  /* Consider it failed here - forces a restart, prevents migration */
2632  pe__set_resource_flags(rsc, pe_rsc_failed|pe_rsc_stop);
2633  pe__clear_resource_flags(rsc, pe_rsc_allow_migrate);
2634  }
2635  }
2636 }
2637 
2638 // Is there an action_name in node_name's rsc history newer than call_id?
2639 static bool
2640 newer_op(pe_resource_t *rsc, const char *action_name, const char *node_name,
2641  int call_id, pe_working_set_t *data_set)
2642 {
2643  xmlNode *action = find_lrm_op(rsc->id, action_name, node_name, NULL, TRUE,
2644  data_set);
2645 
2646  return pe__call_id(action) > call_id;
2647 }
2648 
2649 static void
2650 unpack_migrate_to_failure(pe_resource_t *rsc, pe_node_t *node, xmlNode *xml_op,
2651  pe_working_set_t *data_set)
2652 {
2653  int target_stop_id = 0;
2654  int target_migrate_from_id = 0;
2655  xmlNode *target_stop = NULL;
2656  xmlNode *target_migrate_from = NULL;
2657  const char *source = crm_element_value(xml_op, XML_LRM_ATTR_MIGRATE_SOURCE);
2658  const char *target = crm_element_value(xml_op, XML_LRM_ATTR_MIGRATE_TARGET);
2659 
2660  // Sanity check
2661  CRM_CHECK(source && target && !strcmp(source, node->details->uname), return);
2662 
2663  /* If a migration failed, we have to assume the resource is active. Clones
2664  * are not allowed to migrate, so role can't be master.
2665  */
2666  rsc->role = RSC_ROLE_STARTED;
2667 
2668  // Check for stop on the target
2669  target_stop = find_lrm_op(rsc->id, CRMD_ACTION_STOP, target, NULL,
2670  TRUE, data_set);
2671  target_stop_id = pe__call_id(target_stop);
2672 
2673  // Check for migrate_from on the target
2674  target_migrate_from = find_lrm_op(rsc->id, CRMD_ACTION_MIGRATED, target,
2675  source, TRUE, data_set);
2676  target_migrate_from_id = pe__call_id(target_migrate_from);
2677 
2678  if ((target_stop == NULL) || (target_stop_id < target_migrate_from_id)) {
2679  /* There was no stop on the target, or a stop that happened before a
2680  * migrate_from, so assume the resource is still active on the target
2681  * (if it is up).
2682  */
2683  pe_node_t *target_node = pe_find_node(data_set->nodes, target);
2684 
2685  pe_rsc_trace(rsc, "stop (%d) + migrate_from (%d)",
2686  target_stop_id, target_migrate_from_id);
2687  if (target_node && target_node->details->online) {
2688  native_add_running(rsc, target_node, data_set);
2689  }
2690 
2691  } else if (target_migrate_from == NULL) {
2692  /* We know there was a stop on the target, but there may not have been a
2693  * migrate_from (the stop could have happened before migrate_from was
2694  * scheduled or attempted).
2695  *
2696  * That means this could be a "dangling" migration. But first, check
2697  * whether there is a newer successful stop, start, or migrate_from on
2698  * the source node -- it's possible the failed migration was followed by
2699  * a successful stop, full restart, or migration in the reverse
2700  * direction, in which case we don't want to force a stop.
2701  */
2702  int source_migrate_to_id = pe__call_id(xml_op);
2703 
2704  if (newer_op(rsc, CRMD_ACTION_MIGRATED, source, source_migrate_to_id,
2705  data_set)
2706  || newer_op(rsc, CRMD_ACTION_START, source, source_migrate_to_id,
2707  data_set)
2708  || newer_op(rsc, CRMD_ACTION_STOP, source, source_migrate_to_id,
2709  data_set)) {
2710  return;
2711  }
2712 
2713  // Mark node as having dangling migration so we can force a stop later
2714  rsc->dangling_migrations = g_list_prepend(rsc->dangling_migrations, node);
2715  }
2716 }
2717 
2718 static void
2719 unpack_migrate_from_failure(pe_resource_t *rsc, pe_node_t *node,
2720  xmlNode *xml_op, pe_working_set_t *data_set)
2721 {
2722  xmlNode *source_stop = NULL;
2723  xmlNode *source_migrate_to = NULL;
2724  const char *source = crm_element_value(xml_op, XML_LRM_ATTR_MIGRATE_SOURCE);
2725  const char *target = crm_element_value(xml_op, XML_LRM_ATTR_MIGRATE_TARGET);
2726 
2727  // Sanity check
2728  CRM_CHECK(source && target && !strcmp(target, node->details->uname), return);
2729 
2730  /* If a migration failed, we have to assume the resource is active. Clones
2731  * are not allowed to migrate, so role can't be master.
2732  */
2733  rsc->role = RSC_ROLE_STARTED;
2734 
2735  // Check for a stop on the source
2736  source_stop = find_lrm_op(rsc->id, CRMD_ACTION_STOP, source, NULL,
2737  TRUE, data_set);
2738 
2739  // Check for a migrate_to on the source
2740  source_migrate_to = find_lrm_op(rsc->id, CRMD_ACTION_MIGRATE,
2741  source, target, TRUE, data_set);
2742 
2743  if ((source_stop == NULL)
2744  || (pe__call_id(source_stop) < pe__call_id(source_migrate_to))) {
2745  /* There was no stop on the source, or a stop that happened before
2746  * migrate_to, so assume the resource is still active on the source (if
2747  * it is up).
2748  */
2749  pe_node_t *source_node = pe_find_node(data_set->nodes, source);
2750 
2751  if (source_node && source_node->details->online) {
2752  native_add_running(rsc, source_node, data_set);
2753  }
2754  }
2755 }
2756 
2757 static void
2758 record_failed_op(xmlNode *op, const pe_node_t *node,
2759  const pe_resource_t *rsc, pe_working_set_t *data_set)
2760 {
2761  xmlNode *xIter = NULL;
2762  const char *op_key = crm_element_value(op, XML_LRM_ATTR_TASK_KEY);
2763 
2764  if (node->details->online == FALSE) {
2765  return;
2766  }
2767 
2768  for (xIter = data_set->failed->children; xIter; xIter = xIter->next) {
2769  const char *key = crm_element_value(xIter, XML_LRM_ATTR_TASK_KEY);
2770  const char *uname = crm_element_value(xIter, XML_ATTR_UNAME);
2771 
2772  if(pcmk__str_eq(op_key, key, pcmk__str_casei) && pcmk__str_eq(uname, node->details->uname, pcmk__str_casei)) {
2773  crm_trace("Skipping duplicate entry %s on %s", op_key, node->details->uname);
2774  return;
2775  }
2776  }
2777 
2778  crm_trace("Adding entry %s on %s", op_key, node->details->uname);
2779  crm_xml_add(op, XML_ATTR_UNAME, node->details->uname);
2780  crm_xml_add(op, XML_LRM_ATTR_RSCID, rsc->id);
2781  add_node_copy(data_set->failed, op);
2782 }
2783 
2784 static const char *get_op_key(xmlNode *xml_op)
2785 {
2786  const char *key = crm_element_value(xml_op, XML_LRM_ATTR_TASK_KEY);
2787  if(key == NULL) {
2788  key = ID(xml_op);
2789  }
2790  return key;
2791 }
2792 
2793 static const char *
2794 last_change_str(xmlNode *xml_op)
2795 {
2796  time_t when;
2797  const char *when_s = NULL;
2798 
2799  if (crm_element_value_epoch(xml_op, XML_RSC_OP_LAST_CHANGE,
2800  &when) == pcmk_ok) {
2801  when_s = pcmk__epoch2str(&when);
2802  if (when_s) {
2803  // Skip day of week to make message shorter
2804  when_s = strchr(when_s, ' ');
2805  if (when_s) {
2806  ++when_s;
2807  }
2808  }
2809  }
2810  return ((when_s && *when_s)? when_s : "unknown time");
2811 }
2812 
2825 static int
2826 cmp_on_fail(enum action_fail_response first, enum action_fail_response second)
2827 {
2828  switch (first) {
2829  case action_fail_demote:
2830  switch (second) {
2831  case action_fail_ignore:
2832  return 1;
2833  case action_fail_demote:
2834  return 0;
2835  default:
2836  return -1;
2837  }
2838  break;
2839 
2840  case action_fail_reset_remote:
2841  switch (second) {
2842  case action_fail_ignore:
2843  case action_fail_demote:
2844  case action_fail_recover:
2845  return 1;
2846  case action_fail_reset_remote:
2847  return 0;
2848  default:
2849  return -1;
2850  }
2851  break;
2852 
2853  case action_fail_restart_container:
2854  switch (second) {
2855  case action_fail_ignore:
2856  case action_fail_demote:
2857  case action_fail_recover:
2858  case action_fail_reset_remote:
2859  return 1;
2860  case action_fail_restart_container:
2861  return 0;
2862  default:
2863  return -1;
2864  }
2865  break;
2866 
2867  default:
2868  break;
2869  }
2870  switch (second) {
2871  case action_fail_demote:
2872  return (first == action_fail_ignore)? -1 : 1;
2873 
2874  case action_fail_reset_remote:
2875  switch (first) {
2876  case action_fail_ignore:
2877  case action_fail_demote:
2878  case action_fail_recover:
2879  return -1;
2880  default:
2881  return 1;
2882  }
2883  break;
2884 
2885  case action_fail_restart_container:
2886  switch (first) {
2887  case action_fail_ignore:
2888  case action_fail_demote:
2889  case action_fail_recover:
2890  case action_fail_reset_remote:
2891  return -1;
2892  default:
2893  return 1;
2894  }
2895  break;
2896 
2897  default:
2898  break;
2899  }
2900  return first - second;
2901 }
2902 
2903 static void
2904 unpack_rsc_op_failure(pe_resource_t * rsc, pe_node_t * node, int rc, xmlNode * xml_op, xmlNode ** last_failure,
2905  enum action_fail_response * on_fail, pe_working_set_t * data_set)
2906 {
2907  guint interval_ms = 0;
2908  bool is_probe = false;
2909  pe_action_t *action = NULL;
2910 
2911  const char *key = get_op_key(xml_op);
2912  const char *task = crm_element_value(xml_op, XML_LRM_ATTR_TASK);
2913  const char *exit_reason = crm_element_value(xml_op,
2914  XML_LRM_ATTR_EXIT_REASON);
2915 
2916  CRM_ASSERT(rsc);
2917  CRM_CHECK(task != NULL, return);
2918 
2919  *last_failure = xml_op;
2920 
2921  crm_element_value_ms(xml_op, XML_LRM_ATTR_INTERVAL_MS, &interval_ms);
2922  if ((interval_ms == 0) && !strcmp(task, CRMD_ACTION_STATUS)) {
2923  is_probe = true;
2924  }
2925 
2926  if (exit_reason == NULL) {
2927  exit_reason = "";
2928  }
2929 
2930  if (!pcmk_is_set(data_set->flags, pe_flag_symmetric_cluster)
2931  && (rc == PCMK_OCF_NOT_INSTALLED)) {
2932  crm_trace("Unexpected result (%s%s%s) was recorded for "
2933  "%s of %s on %s at %s " CRM_XS " rc=%d id=%s",
2934  services_ocf_exitcode_str(rc),
2935  (*exit_reason? ": " : ""), exit_reason,
2936  (is_probe? "probe" : task), rsc->id, node->details->uname,
2937  last_change_str(xml_op), rc, ID(xml_op));
2938  } else {
2939  crm_warn("Unexpected result (%s%s%s) was recorded for "
2940  "%s of %s on %s at %s " CRM_XS " rc=%d id=%s",
2941  services_ocf_exitcode_str(rc),
2942  (*exit_reason? ": " : ""), exit_reason,
2943  (is_probe? "probe" : task), rsc->id, node->details->uname,
2944  last_change_str(xml_op), rc, ID(xml_op));
2945 
2946  if (is_probe && (rc != PCMK_OCF_OK)
2947  && (rc != PCMK_OCF_NOT_RUNNING)
2948  && (rc != PCMK_OCF_RUNNING_MASTER)) {
2949 
2950  /* A failed (not just unexpected) probe result could mean the user
2951  * didn't know resources will be probed even where they can't run.
2952  */
2953  crm_notice("If it is not possible for %s to run on %s, see "
2954  "the resource-discovery option for location constraints",
2955  rsc->id, node->details->uname);
2956  }
2957 
2958  record_failed_op(xml_op, node, rsc, data_set);
2959  }
2960 
2961  action = custom_action(rsc, strdup(key), task, NULL, TRUE, FALSE, data_set);
2962  if (cmp_on_fail(*on_fail, action->on_fail) < 0) {
2963  pe_rsc_trace(rsc, "on-fail %s -> %s for %s (%s)", fail2text(*on_fail),
2964  fail2text(action->on_fail), action->uuid, key);
2965  *on_fail = action->on_fail;
2966  }
2967 
2968  if (!strcmp(task, CRMD_ACTION_STOP)) {
2969  resource_location(rsc, node, -INFINITY, "__stop_fail__", data_set);
2970 
2971  } else if (!strcmp(task, CRMD_ACTION_MIGRATE)) {
2972  unpack_migrate_to_failure(rsc, node, xml_op, data_set);
2973 
2974  } else if (!strcmp(task, CRMD_ACTION_MIGRATED)) {
2975  unpack_migrate_from_failure(rsc, node, xml_op, data_set);
2976 
2977  } else if (!strcmp(task, CRMD_ACTION_PROMOTE)) {
2978  rsc->role = RSC_ROLE_MASTER;
2979 
2980  } else if (!strcmp(task, CRMD_ACTION_DEMOTE)) {
2981  if (action->on_fail == action_fail_block) {
2982  rsc->role = RSC_ROLE_MASTER;
2983  rsc->next_role = RSC_ROLE_STOPPED;
2984 
2985  } else if(rc == PCMK_OCF_NOT_RUNNING) {
2986  rsc->role = RSC_ROLE_STOPPED;
2987 
2988  } else {
2989  /* Staying in master role would put the scheduler and controller
2990  * into a loop. Setting slave role is not dangerous because the
2991  * resource will be stopped as part of recovery, and any master
2992  * promotion will be ordered after that stop.
2993  */
2994  rsc->role = RSC_ROLE_SLAVE;
2995  }
2996  }
2997 
2998  if(is_probe && rc == PCMK_OCF_NOT_INSTALLED) {
2999  /* leave stopped */
3000  pe_rsc_trace(rsc, "Leaving %s stopped", rsc->id);
3001  rsc->role = RSC_ROLE_STOPPED;
3002 
3003  } else if (rsc->role < RSC_ROLE_STARTED) {
3004  pe_rsc_trace(rsc, "Setting %s active", rsc->id);
3005  set_active(rsc);
3006  }
3007 
3008  pe_rsc_trace(rsc, "Resource %s: role=%s, unclean=%s, on_fail=%s, fail_role=%s",
3009  rsc->id, role2text(rsc->role),
3010  pcmk__btoa(node->details->unclean),
3011  fail2text(action->on_fail), role2text(action->fail_role));
3012 
3013  if (action->fail_role != RSC_ROLE_STARTED && rsc->next_role < action->fail_role) {
3014  rsc->next_role = action->fail_role;
3015  }
3016 
3017  if (action->fail_role == RSC_ROLE_STOPPED) {
3018  int score = -INFINITY;
3019 
3020  pe_resource_t *fail_rsc = rsc;
3021 
3022  if (fail_rsc->parent) {
3023  pe_resource_t *parent = uber_parent(fail_rsc);
3024 
3025  if (pe_rsc_is_clone(parent)
3026  && !pcmk_is_set(parent->flags, pe_rsc_unique)) {
3027  /* For clone resources, if a child fails on an operation
3028  * with on-fail = stop, all the resources fail. Do this by preventing
3029  * the parent from coming up again. */
3030  fail_rsc = parent;
3031  }
3032  }
3033  crm_notice("%s will not be started under current conditions",
3034  fail_rsc->id);
3035  /* make sure it doesn't come up again */
3036  if (fail_rsc->allowed_nodes != NULL) {
3037  g_hash_table_destroy(fail_rsc->allowed_nodes);
3038  }
3039  fail_rsc->allowed_nodes = pe__node_list2table(data_set->nodes);
3040  g_hash_table_foreach(fail_rsc->allowed_nodes, set_node_score, &score);
3041  }
3042 
3043  pe_free_action(action);
3044 }
3045 
3065 static int
3066 determine_op_status(
3067  pe_resource_t *rsc, int rc, int target_rc, pe_node_t * node, xmlNode * xml_op, enum action_fail_response * on_fail, pe_working_set_t * data_set)
3068 {
3069  guint interval_ms = 0;
3070  bool is_probe = false;
3071  int result = PCMK_LRM_OP_DONE;
3072  const char *key = get_op_key(xml_op);
3073  const char *task = crm_element_value(xml_op, XML_LRM_ATTR_TASK);
3074  const char *exit_reason = crm_element_value(xml_op,
3075  XML_LRM_ATTR_EXIT_REASON);
3076 
3077  CRM_ASSERT(rsc);
3078  CRM_CHECK(task != NULL, return PCMK_LRM_OP_ERROR);
3079 
3080  if (exit_reason == NULL) {
3081  exit_reason = "";
3082  }
3083 
3084  crm_element_value_ms(xml_op, XML_LRM_ATTR_INTERVAL_MS, &interval_ms);
3085  if ((interval_ms == 0) && !strcmp(task, CRMD_ACTION_STATUS)) {
3086  is_probe = true;
3087  task = "probe";
3088  }
3089 
3090  if (target_rc < 0) {
3091  /* Pre-1.0 Pacemaker versions, and Pacemaker 1.1.6 or earlier with
3092  * Heartbeat 2.0.7 or earlier as the cluster layer, did not include the
3093  * target_rc in the transition key, which (along with the similar case
3094  * of a corrupted transition key in the CIB) will be reported to this
3095  * function as -1. Pacemaker 2.0+ does not support rolling upgrades from
3096  * those versions or processing of saved CIB files from those versions,
3097  * so we do not need to care much about this case.
3098  */
3099  result = PCMK_LRM_OP_ERROR;
3100  crm_warn("Expected result not found for %s on %s (corrupt or obsolete CIB?)",
3101  key, node->details->uname);
3102 
3103  } else if (target_rc != rc) {
3104  result = PCMK_LRM_OP_ERROR;
3105  pe_rsc_debug(rsc, "%s on %s: expected %d (%s), got %d (%s%s%s)",
3106  key, node->details->uname,
3107  target_rc, services_ocf_exitcode_str(target_rc),
3108  rc, services_ocf_exitcode_str(rc),
3109  (*exit_reason? ": " : ""), exit_reason);
3110  }
3111 
3112  switch (rc) {
3113  case PCMK_OCF_OK:
3114  if (is_probe && (target_rc == PCMK_OCF_NOT_RUNNING)) {
3115  result = PCMK_LRM_OP_DONE;
3116  pe_rsc_info(rsc, "Probe found %s active on %s at %s",
3117  rsc->id, node->details->uname,
3118  last_change_str(xml_op));
3119  }
3120  break;
3121 
3122  case PCMK_OCF_NOT_RUNNING:
3123  if (is_probe || (target_rc == rc)
3124  || !pcmk_is_set(rsc->flags, pe_rsc_managed)) {
3125 
3126  result = PCMK_LRM_OP_DONE;
3127  rsc->role = RSC_ROLE_STOPPED;
3128 
3129  /* clear any previous failure actions */
3130  *on_fail = action_fail_ignore;
3131  rsc->next_role = RSC_ROLE_UNKNOWN;
3132  }
3133  break;
3134 
3135  case PCMK_OCF_RUNNING_MASTER:
3136  if (is_probe && (rc != target_rc)) {
3137  result = PCMK_LRM_OP_DONE;
3138  pe_rsc_info(rsc,
3139  "Probe found %s active and promoted on %s at %s",
3140  rsc->id, node->details->uname,
3141  last_change_str(xml_op));
3142  }
3143  rsc->role = RSC_ROLE_MASTER;
3144  break;
3145 
3146  case PCMK_OCF_DEGRADED_MASTER:
3147  case PCMK_OCF_FAILED_MASTER:
3148  rsc->role = RSC_ROLE_MASTER;
3149  result = PCMK_LRM_OP_ERROR;
3150  break;
3151 
3152  case PCMK_OCF_NOT_CONFIGURED:
3153  result = PCMK_LRM_OP_ERROR_FATAL;
3154  break;
3155 
3156  case PCMK_OCF_UNIMPLEMENT_FEATURE:
3157  if (interval_ms > 0) {
3158  result = PCMK_LRM_OP_NOTSUPPORTED;
3159  break;
3160  }
3161  // fall through
3162  case PCMK_OCF_NOT_INSTALLED:
3163  case PCMK_OCF_INVALID_PARAM:
3164  case PCMK_OCF_INSUFFICIENT_PRIV:
3165  if (!pe_can_fence(data_set, node)
3166  && !strcmp(task, CRMD_ACTION_STOP)) {
3167  /* If a stop fails and we can't fence, there's nothing else we can do */
3168  pe_proc_err("No further recovery can be attempted for %s "
3169  "because %s on %s failed (%s%s%s) at %s "
3170  CRM_XS " rc=%d id=%s", rsc->id, task,
3171  node->details->uname, services_ocf_exitcode_str(rc),
3172  (*exit_reason? ": " : ""), exit_reason,
3173  last_change_str(xml_op), rc, ID(xml_op));
3174  pe__clear_resource_flags(rsc, pe_rsc_managed);
3175  pe__set_resource_flags(rsc, pe_rsc_block);
3176  }
3177  result = PCMK_LRM_OP_ERROR_HARD;
3178  break;
3179 
3180  default:
3181  if (result == PCMK_LRM_OP_DONE) {
3182  crm_info("Treating unknown exit status %d from %s of %s "
3183  "on %s at %s as failure",
3184  rc, task, rsc->id, node->details->uname,
3185  last_change_str(xml_op));
3186  result = PCMK_LRM_OP_ERROR;
3187  }
3188  break;
3189  }
3190  return result;
3191 }
3192 
3193 // return TRUE if start or monitor last failure but parameters changed
3194 static bool
3195 should_clear_for_param_change(xmlNode *xml_op, const char *task,
3196  pe_resource_t *rsc, pe_node_t *node,
3197  pe_working_set_t *data_set)
3198 {
3199  if (!strcmp(task, "start") || !strcmp(task, "monitor")) {
3200 
3201  if (pe__bundle_needs_remote_name(rsc)) {
3202  /* We haven't allocated resources yet, so we can't reliably
3203  * substitute addr parameters for the REMOTE_CONTAINER_HACK.
3204  * When that's needed, defer the check until later.
3205  */
3206  pe__add_param_check(xml_op, rsc, node, pe_check_last_failure,
3207  data_set);
3208 
3209  } else {
3210  op_digest_cache_t *digest_data = NULL;
3211 
3212  digest_data = rsc_action_digest_cmp(rsc, xml_op, node, data_set);
3213  switch (digest_data->rc) {
3214  case RSC_DIGEST_UNKNOWN:
3215  crm_trace("Resource %s history entry %s on %s"
3216  " has no digest to compare",
3217  rsc->id, get_op_key(xml_op), node->details->id);
3218  break;
3219  case RSC_DIGEST_MATCH:
3220  break;
3221  default:
3222  return TRUE;
3223  }
3224  }
3225  }
3226  return FALSE;
3227 }
3228 
3229 // Order action after fencing of remote node, given connection rsc
3230 static void
3231 order_after_remote_fencing(pe_action_t *action, pe_resource_t *remote_conn,
3232  pe_working_set_t *data_set)
3233 {
3234  pe_node_t *remote_node = pe_find_node(data_set->nodes, remote_conn->id);
3235 
3236  if (remote_node) {
3237  pe_action_t *fence = pe_fence_op(remote_node, NULL, TRUE, NULL,
3238  FALSE, data_set);
3239 
3240  order_actions(fence, action, pe_order_implies_then);
3241  }
3242 }
3243 
3244 static bool
3245 should_ignore_failure_timeout(pe_resource_t *rsc, xmlNode *xml_op,
3246  const char *task, guint interval_ms,
3247  bool is_last_failure, pe_working_set_t *data_set)
3248 {
3249  /* Clearing failures of recurring monitors has special concerns. The
3250  * executor reports only changes in the monitor result, so if the
3251  * monitor is still active and still getting the same failure result,
3252  * that will go undetected after the failure is cleared.
3253  *
3254  * Also, the operation history will have the time when the recurring
3255  * monitor result changed to the given code, not the time when the
3256  * result last happened.
3257  *
3258  * @TODO We probably should clear such failures only when the failure
3259  * timeout has passed since the last occurrence of the failed result.
3260  * However we don't record that information. We could maybe approximate
3261  * that by clearing only if there is a more recent successful monitor or
3262  * stop result, but we don't even have that information at this point
3263  * since we are still unpacking the resource's operation history.
3264  *
3265  * This is especially important for remote connection resources with a
3266  * reconnect interval, so in that case, we skip clearing failures
3267  * if the remote node hasn't been fenced.
3268  */
3269  if (rsc->remote_reconnect_ms
3270  && pcmk_is_set(data_set->flags, pe_flag_stonith_enabled)
3271  && (interval_ms != 0) && pcmk__str_eq(task, CRMD_ACTION_STATUS, pcmk__str_casei)) {
3272 
3273  pe_node_t *remote_node = pe_find_node(data_set->nodes, rsc->id);
3274 
3275  if (remote_node && !remote_node->details->remote_was_fenced) {
3276  if (is_last_failure) {
3277  crm_info("Waiting to clear monitor failure for remote node %s"
3278  " until fencing has occurred", rsc->id);
3279  }
3280  return TRUE;
3281  }
3282  }
3283  return FALSE;
3284 }
3285 
3308 static bool
3309 check_operation_expiry(pe_resource_t *rsc, pe_node_t *node, int rc,
3310  xmlNode *xml_op, pe_working_set_t *data_set)
3311 {
3312  bool expired = FALSE;
3313  bool is_last_failure = pcmk__ends_with(ID(xml_op), "_last_failure_0");
3314  time_t last_run = 0;
3315  guint interval_ms = 0;
3316  int unexpired_fail_count = 0;
3317  const char *task = crm_element_value(xml_op, XML_LRM_ATTR_TASK);
3318  const char *clear_reason = NULL;
3319 
3320  crm_element_value_ms(xml_op, XML_LRM_ATTR_INTERVAL_MS, &interval_ms);
3321 
3322  if ((rsc->failure_timeout > 0)
3323  && (crm_element_value_epoch(xml_op, XML_RSC_OP_LAST_CHANGE,
3324  &last_run) == 0)) {
3325 
3326  // Resource has a failure-timeout, and history entry has a timestamp
3327 
3328  time_t now = get_effective_time(data_set);
3329  time_t last_failure = 0;
3330 
3331  // Is this particular operation history older than the failure timeout?
3332  if ((now >= (last_run + rsc->failure_timeout))
3333  && !should_ignore_failure_timeout(rsc, xml_op, task, interval_ms,
3334  is_last_failure, data_set)) {
3335  expired = TRUE;
3336  }
3337 
3338  // Does the resource as a whole have an unexpired fail count?
3339  unexpired_fail_count = pe_get_failcount(node, rsc, &last_failure,
3340  pe_fc_effective, xml_op,
3341  data_set);
3342 
3343  // Update scheduler recheck time according to *last* failure
3344  crm_trace("%s@%lld is %sexpired @%lld with unexpired_failures=%d timeout=%ds"
3345  " last-failure@%lld",
3346  ID(xml_op), (long long) last_run, (expired? "" : "not "),
3347  (long long) now, unexpired_fail_count, rsc->failure_timeout,
3348  (long long) last_failure);
3349  last_failure += rsc->failure_timeout + 1;
3350  if (unexpired_fail_count && (now < last_failure)) {
3351  pe__update_recheck_time(last_failure, data_set);
3352  }
3353  }
3354 
3355  if (expired) {
3356  if (pe_get_failcount(node, rsc, NULL, pe_fc_default, xml_op, data_set)) {
3357 
3358  // There is a fail count ignoring timeout
3359 
3360  if (unexpired_fail_count == 0) {
3361  // There is no fail count considering timeout
3362  clear_reason = "it expired";
3363 
3364  } else {
3365  /* This operation is old, but there is an unexpired fail count.
3366  * In a properly functioning cluster, this should only be
3367  * possible if this operation is not a failure (otherwise the
3368  * fail count should be expired too), so this is really just a
3369  * failsafe.
3370  */
3371  expired = FALSE;
3372  }
3373 
3374  } else if (is_last_failure && rsc->remote_reconnect_ms) {
3375  /* Clear any expired last failure when reconnect interval is set,
3376  * even if there is no fail count.
3377  */
3378  clear_reason = "reconnect interval is set";
3379  }
3380  }
3381 
3382  if (!expired && is_last_failure
3383  && should_clear_for_param_change(xml_op, task, rsc, node, data_set)) {
3384  clear_reason = "resource parameters have changed";
3385  }
3386 
3387  if (clear_reason != NULL) {
3388  // Schedule clearing of the fail count
3389  pe_action_t *clear_op = pe__clear_failcount(rsc, node, clear_reason,
3390  data_set);
3391 
3392  if (pcmk_is_set(data_set->flags, pe_flag_stonith_enabled)
3393  && rsc->remote_reconnect_ms) {
3394  /* If we're clearing a remote connection due to a reconnect
3395  * interval, we want to wait until any scheduled fencing
3396  * completes.
3397  *
3398  * We could limit this to remote_node->details->unclean, but at
3399  * this point, that's always true (it won't be reliable until
3400  * after unpack_node_loop() is done).
3401  */
3402  crm_info("Clearing %s failure will wait until any scheduled "
3403  "fencing of %s completes", task, rsc->id);
3404  order_after_remote_fencing(clear_op, rsc, data_set);
3405  }
3406  }
3407 
3408  if (expired && (interval_ms == 0) && pcmk__str_eq(task, CRMD_ACTION_STATUS, pcmk__str_casei)) {
3409  switch(rc) {
3410  case PCMK_OCF_OK:
3411  case PCMK_OCF_NOT_RUNNING:
3412  case PCMK_OCF_RUNNING_MASTER:
3413  case PCMK_OCF_DEGRADED:
3414  case PCMK_OCF_DEGRADED_MASTER:
3415  // Don't expire probes that return these values
3416  expired = FALSE;
3417  break;
3418  }
3419  }
3420 
3421  return expired;
3422 }
3423 
3424 int pe__target_rc_from_xml(xmlNode *xml_op)
3425 {
3426  int target_rc = 0;
3427  const char *key = crm_element_value(xml_op, XML_ATTR_TRANSITION_KEY);
3428 
3429  if (key == NULL) {
3430  return -1;
3431  }
3432  decode_transition_key(key, NULL, NULL, NULL, &target_rc);
3433  return target_rc;
3434 }
3435 
3436 static enum action_fail_response
3437 get_action_on_fail(pe_resource_t *rsc, const char *key, const char *task, pe_working_set_t * data_set)
3438 {
3439  enum action_fail_response result = action_fail_recover;
3440  pe_action_t *action = custom_action(rsc, strdup(key), task, NULL, TRUE, FALSE, data_set);
3441 
3442  result = action->on_fail;
3443  pe_free_action(action);
3444 
3445  return result;
3446 }
3447 
3448 static void
3449 update_resource_state(pe_resource_t * rsc, pe_node_t * node, xmlNode * xml_op, const char * task, int rc,
3450  xmlNode * last_failure, enum action_fail_response * on_fail, pe_working_set_t * data_set)
3451 {
3452  gboolean clear_past_failure = FALSE;
3453 
3454  CRM_ASSERT(rsc);
3455  CRM_ASSERT(xml_op);
3456 
3457  if (rc == PCMK_OCF_NOT_RUNNING) {
3458  clear_past_failure = TRUE;
3459 
3460  } else if (rc == PCMK_OCF_NOT_INSTALLED) {
3461  rsc->role = RSC_ROLE_STOPPED;
3462 
3463  } else if (pcmk__str_eq(task, CRMD_ACTION_STATUS, pcmk__str_casei)) {
3464  if (last_failure) {
3465  const char *op_key = get_op_key(xml_op);
3466  const char *last_failure_key = get_op_key(last_failure);
3467 
3468  if (pcmk__str_eq(op_key, last_failure_key, pcmk__str_casei)) {
3469  clear_past_failure = TRUE;
3470  }
3471  }
3472 
3473  if (rsc->role < RSC_ROLE_STARTED) {
3474  set_active(rsc);
3475  }
3476 
3477  } else if (pcmk__str_eq(task, CRMD_ACTION_START, pcmk__str_casei)) {
3478  rsc->role = RSC_ROLE_STARTED;
3479  clear_past_failure = TRUE;
3480 
3481  } else if (pcmk__str_eq(task, CRMD_ACTION_STOP, pcmk__str_casei)) {
3482  rsc->role = RSC_ROLE_STOPPED;
3483  clear_past_failure = TRUE;
3484 
3485  } else if (pcmk__str_eq(task, CRMD_ACTION_PROMOTE, pcmk__str_casei)) {
3486  rsc->role = RSC_ROLE_MASTER;
3487  clear_past_failure = TRUE;
3488 
3489  } else if (pcmk__str_eq(task, CRMD_ACTION_DEMOTE, pcmk__str_casei)) {
3490 
3491  if (*on_fail == action_fail_demote) {
3492  // Demote clears an error only if on-fail=demote
3493  clear_past_failure = TRUE;
3494  }
3495  rsc->role = RSC_ROLE_SLAVE;
3496 
3497  } else if (pcmk__str_eq(task, CRMD_ACTION_MIGRATED, pcmk__str_casei)) {
3498  rsc->role = RSC_ROLE_STARTED;
3499  clear_past_failure = TRUE;
3500 
3501  } else if (pcmk__str_eq(task, CRMD_ACTION_MIGRATE, pcmk__str_casei)) {
3502  unpack_migrate_to_success(rsc, node, xml_op, data_set);
3503 
3504  } else if (rsc->role < RSC_ROLE_STARTED) {
3505  pe_rsc_trace(rsc, "%s active on %s", rsc->id, node->details->uname);
3506  set_active(rsc);
3507  }
3508 
3509  /* clear any previous failure actions */
3510  if (clear_past_failure) {
3511  switch (*on_fail) {
3512  case action_fail_stop:
3513  case action_fail_fence:
3514  case action_fail_migrate:
3515  case action_fail_standby:
3516  pe_rsc_trace(rsc, "%s.%s is not cleared by a completed stop",
3517  rsc->id, fail2text(*on_fail));
3518  break;
3519 
3520  case action_fail_block:
3521  case action_fail_ignore:
3522  case action_fail_demote:
3523  case action_fail_recover:
3524  case action_fail_restart_container:
3525  *on_fail = action_fail_ignore;
3526  rsc->next_role = RSC_ROLE_UNKNOWN;
3527  break;
3528  case action_fail_reset_remote:
3529  if (rsc->remote_reconnect_ms == 0) {
3530  /* With no reconnect interval, the connection is allowed to
3531  * start again after the remote node is fenced and
3532  * completely stopped. (With a reconnect interval, we wait
3533  * for the failure to be cleared entirely before attempting
3534  * to reconnect.)
3535  */
3536  *on_fail = action_fail_ignore;
3537  rsc->next_role = RSC_ROLE_UNKNOWN;
3538  }
3539  break;
3540  }
3541  }
3542 }
3543 
3564 static int
3565 remap_monitor_rc(int rc, xmlNode *xml_op, const pe_node_t *node,
3566  const pe_resource_t *rsc, pe_working_set_t *data_set)
3567 {
3568  int remapped_rc = pcmk__effective_rc(rc);
3569 
3570  if (rc != remapped_rc) {
3571  crm_trace("Remapping monitor result %d to %d", rc, remapped_rc);
3572  if (!node->details->shutdown || node->details->online) {
3573  record_failed_op(xml_op, node, rsc, data_set);
3574  }
3575  }
3576  return remapped_rc;
3577 }
3578 
3579 static void
3580 unpack_rsc_op(pe_resource_t *rsc, pe_node_t *node, xmlNode *xml_op,
3581  xmlNode **last_failure, enum action_fail_response *on_fail,
3582  pe_working_set_t *data_set)
3583 {
3584  int rc = 0;
3585  int task_id = 0;
3586  int target_rc = 0;
3587  int status = PCMK_LRM_OP_UNKNOWN;
3588  guint interval_ms = 0;
3589  const char *task = NULL;
3590  const char *task_key = NULL;
3591  const char *exit_reason = NULL;
3592  bool expired = FALSE;
3593  pe_resource_t *parent = rsc;
3594  enum action_fail_response failure_strategy = action_fail_recover;
3595 
3596  CRM_CHECK(rsc && node && xml_op, return);
3597 
3598  target_rc = pe__target_rc_from_xml(xml_op);
3599  task_key = get_op_key(xml_op);
3600  task = crm_element_value(xml_op, XML_LRM_ATTR_TASK);
3601  exit_reason = crm_element_value(xml_op, XML_LRM_ATTR_EXIT_REASON);
3602  if (exit_reason == NULL) {
3603  exit_reason = "";
3604  }
3605 
3606  crm_element_value_int(xml_op, XML_LRM_ATTR_RC, &rc);
3607  crm_element_value_int(xml_op, XML_LRM_ATTR_CALLID, &task_id);
3608  crm_element_value_int(xml_op, XML_LRM_ATTR_OPSTATUS, &status);
3609  crm_element_value_ms(xml_op, XML_LRM_ATTR_INTERVAL_MS, &interval_ms);
3610 
3611  CRM_CHECK(task != NULL, return);
3612  CRM_CHECK(status <= PCMK_LRM_OP_INVALID, return);
3613  CRM_CHECK(status >= PCMK_LRM_OP_PENDING, return);
3614 
3615  if (!strcmp(task, CRMD_ACTION_NOTIFY) ||
3616  !strcmp(task, CRMD_ACTION_METADATA)) {
3617  /* safe to ignore these */
3618  return;
3619  }
3620 
3621  if (!pcmk_is_set(rsc->flags, pe_rsc_unique)) {
3622  parent = uber_parent(rsc);
3623  }
3624 
3625  pe_rsc_trace(rsc, "Unpacking task %s/%s (call_id=%d, status=%d, rc=%d) on %s (role=%s)",
3626  task_key, task, task_id, status, rc, node->details->uname, role2text(rsc->role));
3627 
3628  if (node->details->unclean) {
3629  pe_rsc_trace(rsc, "Node %s (where %s is running) is unclean."
3630  " Further action depends on the value of the stop's on-fail attribute",
3631  node->details->uname, rsc->id);
3632  }
3633 
3634  /* It should be possible to call remap_monitor_rc() first then call
3635  * check_operation_expiry() only if rc != target_rc, because there should
3636  * never be a fail count without at least one unexpected result in the
3637  * resource history. That would be more efficient by avoiding having to call
3638  * check_operation_expiry() for expected results.
3639  *
3640  * However, we do have such configurations in the scheduler regression
3641  * tests, even if it shouldn't be possible with the current code. It's
3642  * probably a good idea anyway, but that would require updating the test
3643  * inputs to something currently possible.
3644  */
3645 
3646  if ((status != PCMK_LRM_OP_NOT_INSTALLED)
3647  && check_operation_expiry(rsc, node, rc, xml_op, data_set)) {
3648  expired = TRUE;
3649  }
3650 
3651  if (!strcmp(task, CRMD_ACTION_STATUS)) {
3652  rc = remap_monitor_rc(rc, xml_op, node, rsc, data_set);
3653  }
3654 
3655  if (expired && (rc != target_rc)) {
3656  const char *magic = crm_element_value(xml_op, XML_ATTR_TRANSITION_MAGIC);
3657 
3658  if (interval_ms == 0) {
3659  crm_notice("Ignoring expired %s failure on %s "
3660  CRM_XS " actual=%d expected=%d magic=%s",
3661  task_key, node->details->uname, rc, target_rc, magic);
3662  goto done;
3663 
3664  } else if(node->details->online && node->details->unclean == FALSE) {
3665  /* Reschedule the recurring monitor. CancelXmlOp() won't work at
3666  * this stage, so as a hacky workaround, forcibly change the restart
3667  * digest so check_action_definition() does what we want later.
3668  *
3669  * @TODO We should skip this if there is a newer successful monitor.
3670  * Also, this causes rescheduling only if the history entry
3671  * has an op-digest (which the expire-non-blocked-failure
3672  * scheduler regression test doesn't, but that may not be a
3673  * realistic scenario in production).
3674  */
3675  crm_notice("Rescheduling %s after failure expired on %s "
3676  CRM_XS " actual=%d expected=%d magic=%s",
3677  task_key, node->details->uname, rc, target_rc, magic);
3678  crm_xml_add(xml_op, XML_LRM_ATTR_RESTART_DIGEST, "calculated-failure-timeout");
3679  goto done;
3680  }
3681  }
3682 
3683  /* If the executor reported an operation status of anything but done or
3684  * error, consider that final. But for done or error, we know better whether
3685  * it should be treated as a failure or not, because we know the expected
3686  * result.
3687  */
3688  if(status == PCMK_LRM_OP_DONE || status == PCMK_LRM_OP_ERROR) {
3689  status = determine_op_status(rsc, rc, target_rc, node, xml_op, on_fail, data_set);
3690  pe_rsc_trace(rsc, "Remapped %s status to %d", task_key, status);
3691  }
3692 
3693  switch (status) {
3694  case PCMK_LRM_OP_CANCELLED:
3695  // Should never happen
3696  pe_err("Resource history contains cancellation '%s' "
3697  "(%s of %s on %s at %s)",
3698  ID(xml_op), task, rsc->id, node->details->uname,
3699  last_change_str(xml_op));
3700  break;
3701 
3702  case PCMK_LRM_OP_PENDING:
3703  if (!strcmp(task, CRMD_ACTION_START)) {
3704  pe__set_resource_flags(rsc, pe_rsc_start_pending);
3705  set_active(rsc);
3706 
3707  } else if (!strcmp(task, CRMD_ACTION_PROMOTE)) {
3708  rsc->role = RSC_ROLE_MASTER;
3709 
3710  } else if (!strcmp(task, CRMD_ACTION_MIGRATE) && node->details->unclean) {
3711  /* If a pending migrate_to action is out on a unclean node,
3712  * we have to force the stop action on the target. */
3713  const char *migrate_target = crm_element_value(xml_op, XML_LRM_ATTR_MIGRATE_TARGET);
3714  pe_node_t *target = pe_find_node(data_set->nodes, migrate_target);
3715  if (target) {
3716  stop_action(rsc, target, FALSE);
3717  }
3718  }
3719 
3720  if (rsc->pending_task == NULL) {
3721  if ((interval_ms != 0) || strcmp(task, CRMD_ACTION_STATUS)) {
3722  rsc->pending_task = strdup(task);
3723  rsc->pending_node = node;
3724  } else {
3725  /* Pending probes are not printed, even if pending
3726  * operations are requested. If someone ever requests that
3727  * behavior, enable the below and the corresponding part of
3728  * native.c:native_pending_task().
3729  */
3730 #if 0
3731  rsc->pending_task = strdup("probe");
3732  rsc->pending_node = node;
3733 #endif
3734  }
3735  }
3736  break;
3737 
3738  case PCMK_LRM_OP_DONE:
3739  pe_rsc_trace(rsc, "%s of %s on %s completed at %s " CRM_XS " id=%s",
3740  task, rsc->id, node->details->uname,
3741  last_change_str(xml_op), ID(xml_op));
3742  update_resource_state(rsc, node, xml_op, task, rc, *last_failure, on_fail, data_set);
3743  break;
3744 
3745  case PCMK_LRM_OP_NOT_INSTALLED:
3746  failure_strategy = get_action_on_fail(rsc, task_key, task, data_set);
3747  if (failure_strategy == action_fail_ignore) {
3748  crm_warn("Cannot ignore failed %s of %s on %s: "
3749  "Resource agent doesn't exist "
3750  CRM_XS " status=%d rc=%d id=%s",
3751  task, rsc->id, node->details->uname, status, rc,
3752  ID(xml_op));
3753  /* Also for printing it as "FAILED" by marking it as pe_rsc_failed later */
3754  *on_fail = action_fail_migrate;
3755  }
3756  resource_location(parent, node, -INFINITY, "hard-error", data_set);
3757  unpack_rsc_op_failure(rsc, node, rc, xml_op, last_failure, on_fail, data_set);
3758  break;
3759 
3760  case PCMK_LRM_OP_NOT_CONNECTED:
3761  if (pe__is_guest_or_remote_node(node)
3762  && pcmk_is_set(node->details->remote_rsc->flags, pe_rsc_managed)) {
3763  /* We should never get into a situation where a managed remote
3764  * connection resource is considered OK but a resource action
3765  * behind the connection gets a "not connected" status. But as a
3766  * fail-safe in case a bug or unusual circumstances do lead to
3767  * that, ensure the remote connection is considered failed.
3768  */
3769  pe__set_resource_flags(node->details->remote_rsc,
3770  pe_rsc_failed|pe_rsc_stop);
3771  }
3772 
3773  // fall through
3774 
3775  case PCMK_LRM_OP_ERROR:
3776  case PCMK_LRM_OP_ERROR_HARD:
3777  case PCMK_LRM_OP_ERROR_FATAL:
3778  case PCMK_LRM_OP_TIMEOUT:
3779  case PCMK_LRM_OP_NOTSUPPORTED:
3780  case PCMK_LRM_OP_INVALID:
3781 
3782  failure_strategy = get_action_on_fail(rsc, task_key, task, data_set);
3783  if ((failure_strategy == action_fail_ignore)
3784  || (failure_strategy == action_fail_restart_container
3785  && !strcmp(task, CRMD_ACTION_STOP))) {
3786 
3787  crm_warn("Pretending failed %s (%s%s%s) of %s on %s at %s "
3788  "succeeded " CRM_XS " rc=%d id=%s",
3789  task, services_ocf_exitcode_str(rc),
3790  (*exit_reason? ": " : ""), exit_reason, rsc->id,
3791  node->details->uname, last_change_str(xml_op), rc,
3792  ID(xml_op));
3793 
3794  update_resource_state(rsc, node, xml_op, task, target_rc, *last_failure, on_fail, data_set);
3795  crm_xml_add(xml_op, XML_ATTR_UNAME, node->details->uname);
3796  pe__set_resource_flags(rsc, pe_rsc_failure_ignored);
3797 
3798  record_failed_op(xml_op, node, rsc, data_set);
3799 
3800  if ((failure_strategy == action_fail_restart_container)
3801  && cmp_on_fail(*on_fail, action_fail_recover) <= 0) {
3802  *on_fail = failure_strategy;
3803  }
3804 
3805  } else {
3806  unpack_rsc_op_failure(rsc, node, rc, xml_op, last_failure, on_fail, data_set);
3807 
3808  if(status == PCMK_LRM_OP_ERROR_HARD) {
3809  do_crm_log(rc != PCMK_OCF_NOT_INSTALLED?LOG_ERR:LOG_NOTICE,
3810  "Preventing %s from restarting on %s because "
3811  "of hard failure (%s%s%s)" CRM_XS " rc=%d id=%s",
3812  parent->id, node->details->uname,
3813  services_ocf_exitcode_str(rc),
3814  (*exit_reason? ": " : ""), exit_reason,
3815  rc, ID(xml_op));
3816  resource_location(parent, node, -INFINITY, "hard-error", data_set);
3817 
3818  } else if(status == PCMK_LRM_OP_ERROR_FATAL) {
3819  crm_err("Preventing %s from restarting anywhere because "
3820  "of fatal failure (%s%s%s) " CRM_XS " rc=%d id=%s",
3821  parent->id, services_ocf_exitcode_str(rc),
3822  (*exit_reason? ": " : ""), exit_reason,
3823  rc, ID(xml_op));
3824  resource_location(parent, NULL, -INFINITY, "fatal-error", data_set);
3825  }
3826  }
3827  break;
3828  }
3829 
3830  done:
3831  pe_rsc_trace(rsc, "Resource %s after %s: role=%s, next=%s",
3832  rsc->id, task, role2text(rsc->role),
3833  role2text(rsc->next_role));
3834 }
3835 
3836 static void
3837 add_node_attrs(xmlNode *xml_obj, pe_node_t *node, bool overwrite,
3838  pe_working_set_t *data_set)
3839 {
3840  const char *cluster_name = NULL;
3841 
3842  pe_rule_eval_data_t rule_data = {
3843  .node_hash = NULL,
3844  .role = RSC_ROLE_UNKNOWN,
3845  .now = data_set->now,
3846  .match_data = NULL,
3847  .rsc_data = NULL,
3848  .op_data = NULL
3849  };
3850 
3851  g_hash_table_insert(node->details->attrs,
3852  strdup(CRM_ATTR_UNAME), strdup(node->details->uname));
3853 
3854  g_hash_table_insert(node->details->attrs, strdup(CRM_ATTR_ID),
3855  strdup(node->details->id));
3856  if (pcmk__str_eq(node->details->id, data_set->dc_uuid, pcmk__str_casei)) {
3857  data_set->dc_node = node;
3858  node->details->is_dc = TRUE;
3859  g_hash_table_insert(node->details->attrs,
3860  strdup(CRM_ATTR_IS_DC), strdup(XML_BOOLEAN_TRUE));
3861  } else {
3862  g_hash_table_insert(node->details->attrs,
3863  strdup(CRM_ATTR_IS_DC), strdup(XML_BOOLEAN_FALSE));
3864  }
3865 
3866  cluster_name = g_hash_table_lookup(data_set->config_hash, "cluster-name");
3867  if (cluster_name) {
3868  g_hash_table_insert(node->details->attrs, strdup(CRM_ATTR_CLUSTER_NAME),
3869  strdup(cluster_name));
3870  }
3871 
3872  pe__unpack_dataset_nvpairs(xml_obj, XML_TAG_ATTR_SETS, &rule_data,
3873  node->details->attrs, NULL, overwrite, data_set);
3874 
3875  if (pe_node_attribute_raw(node, CRM_ATTR_SITE_NAME) == NULL) {
3876  const char *site_name = pe_node_attribute_raw(node, "site-name");
3877 
3878  if (site_name) {
3879  g_hash_table_insert(node->details->attrs,
3880  strdup(CRM_ATTR_SITE_NAME),
3881  strdup(site_name));
3882 
3883  } else if (cluster_name) {
3884  /* Default to cluster-name if unset */
3885  g_hash_table_insert(node->details->attrs,
3886  strdup(CRM_ATTR_SITE_NAME),
3887  strdup(cluster_name));
3888  }
3889  }
3890 }
3891 
3892 static GListPtr
3893 extract_operations(const char *node, const char *rsc, xmlNode * rsc_entry, gboolean active_filter)
3894 {
3895  int counter = -1;
3896  int stop_index = -1;
3897  int start_index = -1;
3898 
3899  xmlNode *rsc_op = NULL;
3900 
3901  GListPtr gIter = NULL;
3902  GListPtr op_list = NULL;
3903  GListPtr sorted_op_list = NULL;
3904 
3905  /* extract operations */
3906  op_list = NULL;
3907  sorted_op_list = NULL;
3908 
3909  for (rsc_op = pcmk__xe_first_child(rsc_entry);
3910  rsc_op != NULL; rsc_op = pcmk__xe_next(rsc_op)) {
3911 
3912  if (pcmk__str_eq((const char *)rsc_op->name, XML_LRM_TAG_RSC_OP,
3913  pcmk__str_none)) {
3914  crm_xml_add(rsc_op, "resource", rsc);
3915  crm_xml_add(rsc_op, XML_ATTR_UNAME, node);
3916  op_list = g_list_prepend(op_list, rsc_op);
3917  }
3918  }
3919 
3920  if (op_list == NULL) {
3921  /* if there are no operations, there is nothing to do */
3922  return NULL;
3923  }
3924 
3925  sorted_op_list = g_list_sort(op_list, sort_op_by_callid);
3926 
3927  /* create active recurring operations as optional */
3928  if (active_filter == FALSE) {
3929  return sorted_op_list;
3930  }
3931 
3932  op_list = NULL;
3933 
3934  calculate_active_ops(sorted_op_list, &start_index, &stop_index);
3935 
3936  for (gIter = sorted_op_list; gIter != NULL; gIter = gIter->next) {
3937  xmlNode *rsc_op = (xmlNode *) gIter->data;
3938 
3939  counter++;
3940 
3941  if (start_index < stop_index) {
3942  crm_trace("Skipping %s: not active", ID(rsc_entry));
3943  break;
3944 
3945  } else if (counter < start_index) {
3946  crm_trace("Skipping %s: old", ID(rsc_op));
3947  continue;
3948  }
3949  op_list = g_list_append(op_list, rsc_op);
3950  }
3951 
3952  g_list_free(sorted_op_list);
3953  return op_list;
3954 }
3955 
3956 GListPtr
3957 find_operations(const char *rsc, const char *node, gboolean active_filter,
3958  pe_working_set_t * data_set)
3959 {
3960  GListPtr output = NULL;
3961  GListPtr intermediate = NULL;
3962 
3963  xmlNode *tmp = NULL;
3964  xmlNode *status = find_xml_node(data_set->input, XML_CIB_TAG_STATUS, TRUE);
3965 
3966  pe_node_t *this_node = NULL;
3967 
3968  xmlNode *node_state = NULL;
3969 
3970  for (node_state = pcmk__xe_first_child(status); node_state != NULL;
3971  node_state = pcmk__xe_next(node_state)) {
3972 
3973  if (pcmk__str_eq((const char *)node_state->name, XML_CIB_TAG_STATE, pcmk__str_none)) {
3974  const char *uname = crm_element_value(node_state, XML_ATTR_UNAME);
3975 
3976  if (node != NULL && !pcmk__str_eq(uname, node, pcmk__str_casei)) {
3977  continue;
3978  }
3979 
3980  this_node = pe_find_node(data_set->nodes, uname);
3981  if(this_node == NULL) {
3982  CRM_LOG_ASSERT(this_node != NULL);
3983  continue;
3984 
3985  } else if (pe__is_guest_or_remote_node(this_node)) {
3986  determine_remote_online_status(data_set, this_node);
3987 
3988  } else {
3989  determine_online_status(node_state, this_node, data_set);
3990  }
3991 
3992  if (this_node->details->online
3993  || pcmk_is_set(data_set->flags, pe_flag_stonith_enabled)) {
3994  /* offline nodes run no resources...
3995  * unless stonith is enabled in which case we need to
3996  * make sure rsc start events happen after the stonith
3997  */
3998  xmlNode *lrm_rsc = NULL;
3999 
4000  tmp = find_xml_node(node_state, XML_CIB_TAG_LRM, FALSE);
4001  tmp = find_xml_node(tmp, XML_LRM_TAG_RESOURCES, FALSE);
4002 
4003  for (lrm_rsc = pcmk__xe_first_child(tmp); lrm_rsc != NULL;
4004  lrm_rsc = pcmk__xe_next(lrm_rsc)) {
4005 
4006  if (pcmk__str_eq((const char *)lrm_rsc->name,
4007  XML_LRM_TAG_RESOURCE, pcmk__str_none)) {
4008 
4009  const char *rsc_id = crm_element_value(lrm_rsc, XML_ATTR_ID);
4010 
4011  if (rsc != NULL && !pcmk__str_eq(rsc_id, rsc, pcmk__str_casei)) {
4012  continue;
4013  }
4014 
4015  intermediate = extract_operations(uname, rsc_id, lrm_rsc, active_filter);
4016  output = g_list_concat(output, intermediate);
4017  }
4018  }
4019  }
4020  }
4021  }
4022 
4023  return output;
4024 }
pe_working_set_s::tags
GHashTable * tags
Definition: pe_types.h:171
services.h
Services API.
CRM_CHECK
#define CRM_CHECK(expr, failure_action)
Definition: logging.h:215
uname
char uname[MAX_NAME]
Definition: internal.h:85
pe_working_set_s::nodes
GListPtr nodes
Definition: pe_types.h:148
XML_RSC_OP_LAST_CHANGE
#define XML_RSC_OP_LAST_CHANGE
Definition: msg_xml.h:287
verify_pe_options
void verify_pe_options(GHashTable *options)
Definition: common.c:306
find_xml_node
xmlNode * find_xml_node(xmlNode *cib, const char *node_path, gboolean must_find)
Definition: xml.c:447
pe_working_set_s::no_quorum_policy
enum pe_quorum_policy no_quorum_policy
Definition: pe_types.h:140
RSC_STOP
#define RSC_STOP
Definition: crm.h:199
crm.h
A dumping ground.
pe_native
Definition: pe_types.h:37
pe_order_implies_then
Definition: pe_types.h:470
pcmk__str_none
Definition: strings_internal.h:25
pe__node_list2table
GHashTable * pe__node_list2table(GList *list)
Definition: utils.c:201
crm_notice
#define crm_notice(fmt, args...)
Definition: logging.h:349
pe_resource_s::known_on
GHashTable * known_on
Definition: pe_types.h:354
CRMD_ACTION_MIGRATED
#define CRMD_ACTION_MIGRATED
Definition: crm.h:169
pe_working_set_s::failed
xmlNode * failed
Definition: pe_types.h:156
pe_flag_stop_action_orphans
#define pe_flag_stop_action_orphans
Definition: pe_types.h:103
pe__is_guest_or_remote_node
bool pe__is_guest_or_remote_node(const pe_node_t *node)
Definition: remote.c:41
pe_node_shared_s::attrs
GHashTable * attrs
Definition: pe_types.h:223
pe_rsc_debug
#define pe_rsc_debug(rsc, fmt, args...)
Definition: internal.h:19
pe_node_shared_s::unseen
gboolean unseen
Definition: pe_types.h:207
XML_CONFIG_ATTR_SHUTDOWN_LOCK
#define XML_CONFIG_ATTR_SHUTDOWN_LOCK
Definition: msg_xml.h:356
pe_node_s::fixed
gboolean fixed
Definition: pe_types.h:231
INFINITY
#define INFINITY
Definition: crm.h:95
sort_rsc_priority
gint sort_rsc_priority(gconstpointer a, gconstpointer b)
Definition: utils.c:413
pe__set_action_flags
#define pe__set_action_flags(action, flags_to_set)
Definition: internal.h:59
CRM_ATTR_KIND
#define CRM_ATTR_KIND
Definition: crm.h:112
pe_resource_s::dangling_migrations
GListPtr dangling_migrations
Definition: pe_types.h:365
XML_NODE_IS_FENCED
#define XML_NODE_IS_FENCED
Definition: msg_xml.h:250
XML_ATTR_TRANSITION_MAGIC
#define XML_ATTR_TRANSITION_MAGIC
Definition: msg_xml.h:367
pe_ticket_s::state
GHashTable * state
Definition: pe_types.h:442
CRM_ATTR_IS_DC
#define CRM_ATTR_IS_DC
Definition: crm.h:114
stop_action
#define stop_action(rsc, node, optional)
Definition: internal.h:417
pe__add_param_check
void pe__add_param_check(xmlNode *rsc_op, pe_resource_t *rsc, pe_node_t *node, enum pe_check_parameters, pe_working_set_t *data_set)
Definition: remote.c:220
pe_resource_s::container
pe_resource_t * container
Definition: pe_types.h:367
pcmk__strcase_any_of
bool pcmk__strcase_any_of(const char *s,...) G_GNUC_NULL_TERMINATED
Definition: strings.c:842
pe_resource_s::partial_migration_source
pe_node_t * partial_migration_source
Definition: pe_types.h:352
XML_ATTR_QUORUM_PANIC
#define XML_ATTR_QUORUM_PANIC
Definition: msg_xml.h:84
pe_flag_concurrent_fencing
#define pe_flag_concurrent_fencing
Definition: pe_types.h:100
XML_ATTR_TYPE
#define XML_ATTR_TYPE
Definition: msg_xml.h:99
XML_CONFIG_ATTR_PRIORITY_FENCING_DELAY
#define XML_CONFIG_ATTR_PRIORITY_FENCING_DELAY
Definition: msg_xml.h:358
pe__update_recheck_time
void pe__update_recheck_time(time_t recheck, pe_working_set_t *data_set)
Definition: utils.c:2727
pe_resource_s::role
enum rsc_role_e role
Definition: pe_types.h:357
XML_TAG_UTILIZATION
#define XML_TAG_UTILIZATION
Definition: msg_xml.h:178
node_remote
Definition: pe_types.h:72
XML_RULE_ATTR_SCORE
#define XML_RULE_ATTR_SCORE
Definition: msg_xml.h:303
XML_BOOLEAN_FALSE
#define XML_BOOLEAN_FALSE
Definition: msg_xml.h:108
pe_flag_symmetric_cluster
#define pe_flag_symmetric_cluster
Definition: pe_types.h:94
pe_ticket_s::standby
gboolean standby
Definition: pe_types.h:441
pe_rsc_stop
#define pe_rsc_stop
Definition: pe_types.h:251
pe_working_set_s::priority_fencing_delay
int priority_fencing_delay
Definition: pe_types.h:181
pcmk__config_err
#define pcmk__config_err(fmt...)
Definition: internal.h:107
pe_flag_remove_after_stop
#define pe_flag_remove_after_stop
Definition: pe_types.h:107
pe_resource_s::next_role
enum rsc_role_e next_role
Definition: pe_types.h:358
pe_flag_maintenance_mode
#define pe_flag_maintenance_mode
Definition: pe_types.h:95
pe_action_s
Definition: pe_types.h:387
char2score
int char2score(const char *score)
Definition: utils.c:61
pe_proc_warn
#define pe_proc_warn(fmt...)
Definition: internal.h:33
pe_node_shared_s::remote_rsc
pe_resource_t * remote_rsc
Definition: pe_types.h:219
XML_TAG_TRANSIENT_NODEATTRS
#define XML_TAG_TRANSIENT_NODEATTRS
Definition: msg_xml.h:372
CRMD_ACTION_NOTIFY
#define CRMD_ACTION_NOTIFY
Definition: crm.h:182
unpack_resources
gboolean unpack_resources(xmlNode *xml_resources, pe_working_set_t *data_set)
Definition: unpack.c:774
pe_resource_s::meta
GHashTable * meta
Definition: pe_types.h:360
pe_rsc_unique
#define pe_rsc_unique
Definition: pe_types.h:243
pe_rule_eval_data
Definition: common.h:172
pe_resource_s::fns
resource_object_functions_t * fns
Definition: pe_types.h:319
XML_CIB_TAG_TAG
#define XML_CIB_TAG_TAG
Definition: msg_xml.h:399
XML_LRM_TAG_RESOURCE
#define XML_LRM_TAG_RESOURCE
Definition: msg_xml.h:234
crm_xml_add
const char * crm_xml_add(xmlNode *node, const char *name, const char *value)
Create an XML attribute with specified name and value.
Definition: nvpair.c:317
pe__copy_node
pe_node_t * pe__copy_node(const pe_node_t *this_node)
Definition: utils.c:139
no_quorum_freeze
Definition: pe_types.h:62
CRMD_ACTION_PROMOTE
#define CRMD_ACTION_PROMOTE
Definition: crm.h:177
crm_parse_int
int crm_parse_int(const char *text, const char *default_text)
Parse an integer value from a string.
Definition: strings.c:134
unpack_tags
gboolean unpack_tags(xmlNode *xml_tags, pe_working_set_t *data_set)
Definition: unpack.c:837
pe_resource_s::fillers
GListPtr fillers
Definition: pe_types.h:368
node_member
Definition: pe_types.h:71
pe_node_shared_s::pending
gboolean pending
Definition: pe_types.h:205
pe_working_set_s::resources
GListPtr resources
Definition: pe_types.h:149
XML_NVPAIR_ATTR_NAME
#define XML_NVPAIR_ATTR_NAME
Definition: msg_xml.h:346
XML_NODE_IS_MAINTENANCE
#define XML_NODE_IS_MAINTENANCE
Definition: msg_xml.h:251
pe_tag_s::id
char * id
Definition: pe_types.h:446
sort_op_by_callid
gint sort_op_by_callid(gconstpointer a, gconstpointer b)
Definition: utils.c:1713
XML_NODE_EXPECTED
#define XML_NODE_EXPECTED
Definition: msg_xml.h:246
XML_CIB_TAG_RSC_TEMPLATE
#define XML_CIB_TAG_RSC_TEMPLATE
Definition: msg_xml.h:187
host
AIS_Host host
Definition: internal.h:84
get_effective_time
time_t get_effective_time(pe_working_set_t *data_set)
Definition: utils.c:1830
CRM_LOG_ASSERT
#define CRM_LOG_ASSERT(expr)
Definition: logging.h:199
pe_find_node
pe_node_t * pe_find_node(GListPtr node_list, const char *uname)
Definition: status.c:435
pe_pref
const char * pe_pref(GHashTable *options, const char *name)
Definition: common.c:312
resource_location
void resource_location(pe_resource_t *rsc, pe_node_t *node, int score, const char *tag, pe_working_set_t *data_set)
Definition: utils.c:1669
unpack_remote_nodes
gboolean unpack_remote_nodes(xmlNode *xml_resources, pe_working_set_t *data_set)
Definition: unpack.c:637
crm_parse_interval_spec
guint crm_parse_interval_spec(const char *input)
Parse milliseconds from a Pacemaker interval specification.
Definition: utils.c:307
copy_in_properties
void copy_in_properties(xmlNode *target, xmlNode *src)
Definition: xml.c:522
CRMD_JOINSTATE_NACK
#define CRMD_JOINSTATE_NACK
Definition: crm.h:162
pcmk__str_null_matches
Definition: strings_internal.h:27
XML_CIB_TAG_LRM
#define XML_CIB_TAG_LRM
Definition: msg_xml.h:232
CRM_ATTR_CLUSTER_NAME
#define CRM_ATTR_CLUSTER_NAME
Definition: crm.h:115
pe_resource_s::partial_migration_target
pe_node_t * partial_migration_target
Definition: pe_types.h:351
pe_working_set_s::tickets
GHashTable * tickets
Definition: pe_types.h:143
pe_node_shared_s::remote_was_fenced
gboolean remote_was_fenced
Definition: pe_types.h:214
crm_element_value_int
int crm_element_value_int(const xmlNode *data, const char *name, int *dest)
Retrieve the integer value of an XML attribute.
Definition: nvpair.c:559
pcmk__ends_with
bool pcmk__ends_with(const char *s, const char *match)
Definition: strings.c:504
pe_flag_have_quorum
#define pe_flag_have_quorum
Definition: pe_types.h:93
pe_working_set_s
Definition: pe_types.h:127
unpack_config
gboolean unpack_config(xmlNode *config, pe_working_set_t *data_set)
Definition: unpack.c:198
pe_resource_s::pending_task
char * pending_task
Definition: pe_types.h:333
pe__bundle_needs_remote_name
bool pe__bundle_needs_remote_name(pe_resource_t *rsc)
Definition: bundle.c:957
get_xpath_object
xmlNode * get_xpath_object(const char *xpath, xmlNode *xml_obj, int error_level)
Definition: xpath.c:213
pe_proc_err
#define pe_proc_err(fmt...)
Definition: internal.h:32
pe_node_shared_s::remote_requires_reset
gboolean remote_requires_reset
Definition: pe_types.h:213
action_fail_response
action_fail_response
Definition: common.h:34
strndup
char * strndup(const char *str, size_t len)
pe_working_set_s::dc_uuid
char * dc_uuid
Definition: pe_types.h:132
action
const char * action
Definition: pcmk_fence.c:30
pe__set_resource_flags
#define pe__set_resource_flags(resource, flags_to_set)
Definition: internal.h:47
pe_working_set_s::stonith_timeout
int stonith_timeout
Definition: pe_types.h:139
XML_CIB_TAG_PROPSET
#define XML_CIB_TAG_PROPSET
Definition: msg_xml.h:169
pe_resource_s::lock_node
pe_node_t * lock_node
Definition: pe_types.h:371
pcmk__effective_rc
int pcmk__effective_rc(int rc)
Definition: agents.c:71
pe_flag_stop_everything
#define pe_flag_stop_everything
Definition: pe_types.h:104
pe__unpack_dataset_nvpairs
void pe__unpack_dataset_nvpairs(xmlNode *xml_obj, const char *set_name, pe_rule_eval_data_t *rule_data, GHashTable *hash, const char *always_first, gboolean overwrite, pe_working_set_t *data_set)
Definition: utils.c:2741
XML_LRM_ATTR_RSCID
#define XML_LRM_ATTR_RSCID
Definition: msg_xml.h:276
pe_node_shared_s::remote_maintenance
gboolean remote_maintenance
Definition: pe_types.h:215
CRMD_ACTION_START
#define CRMD_ACTION_START
Definition: crm.h:171
id
uint32_t id
Definition: internal.h:80
demote_action
#define demote_action(rsc, node, optional)
Definition: internal.h:443
pe_node_shared_s::is_dc
gboolean is_dc
Definition: pe_types.h:210
XML_LRM_ATTR_TASK_KEY
#define XML_LRM_ATTR_TASK_KEY
Definition: msg_xml.h:268
pe__is_remote_node
bool pe__is_remote_node(const pe_node_t *node)
Definition: remote.c:25
XML_TAG_ATTR_SETS
#define XML_TAG_ATTR_SETS
Definition: msg_xml.h:170
XML_LRM_ATTR_TASK
#define XML_LRM_ATTR_TASK
Definition: msg_xml.h:267
role2text
const char * role2text(enum rsc_role_e role)
Definition: common.c:463
pe_create_remote_xml
xmlNode * pe_create_remote_xml(xmlNode *parent, const char *uname, const char *container_id, const char *migrateable, const char *is_managed, const char *start_timeout, const char *server, const char *port)
Definition: remote.c:156
resource_object_functions_s::location
pe_node_t *(* location)(const pe_resource_t *, GList **, int)
Definition: pe_types.h:53
CRMD_ACTION_STOP
#define CRMD_ACTION_STOP
Definition: crm.h:174
pe__find_bundle_replica
pe_resource_t * pe__find_bundle_replica(const pe_resource_t *bundle, const pe_node_t *node)
Definition: bundle.c:1417
STATUS_PATH_MAX
#define STATUS_PATH_MAX
Definition: unpack.c:2460
pe_node_s::weight
int weight
Definition: pe_types.h:230
pe_flag_have_remote_nodes
#define pe_flag_have_remote_nodes
Definition: pe_types.h:113
CRMD_JOINSTATE_DOWN
#define CRMD_JOINSTATE_DOWN
Definition: crm.h:159
crm_warn
#define crm_warn(fmt, args...)
Definition: logging.h:348
pe_resource_s::remote_reconnect_ms
guint remote_reconnect_ms
Definition: pe_types.h:332
XML_CONFIG_ATTR_SHUTDOWN_LOCK_LIMIT
#define XML_CONFIG_ATTR_SHUTDOWN_LOCK_LIMIT
Definition: msg_xml.h:357
CRMD_ACTION_DEMOTE
#define CRMD_ACTION_DEMOTE
Definition: crm.h:179
crm_atoi
#define crm_atoi(text, default_text)
Definition: util.h:110
clone_zero
char * clone_zero(const char *last_rsc_id)
Definition: unpack.c:1598
pe_rsc_allow_migrate
#define pe_rsc_allow_migrate
Definition: pe_types.h:261
pe_rsc_orphan_container_filler
#define pe_rsc_orphan_container_filler
Definition: pe_types.h:240
rc
int rc
Definition: pcmk_fence.c:35
crm_element_value_ms
int crm_element_value_ms(const xmlNode *data, const char *name, guint *dest)
Retrieve the millisecond value of an XML attribute.
Definition: nvpair.c:615
pe_resource_s
Definition: pe_types.h:307
pe_rsc_failed
#define pe_rsc_failed
Definition: pe_types.h:255
crm_debug
#define crm_debug(fmt, args...)
Definition: logging.h:352
pe__target_rc_from_xml
int pe__target_rc_from_xml(xmlNode *xml_op)
Definition: unpack.c:3424
uber_parent
pe_resource_t * uber_parent(pe_resource_t *rsc)
Definition: complex.c:799
XML_RSC_ATTR_CONTAINER
#define XML_RSC_ATTR_CONTAINER
Definition: msg_xml.h:212
util.h
Utility functions.
XML_ATTR_ID
#define XML_ATTR_ID
Definition: msg_xml.h:96
crm_element_value
const char * crm_element_value(const xmlNode *data, const char *name)
Retrieve the value of an XML attribute.
Definition: nvpair.c:523
XML_CIB_TAG_RESOURCE
#define XML_CIB_TAG_RESOURCE
Definition: msg_xml.h:181
pe_rsc_is_container
#define pe_rsc_is_container
Definition: pe_types.h:265
XML_BOOLEAN_TRUE
#define XML_BOOLEAN_TRUE
Definition: msg_xml.h:107
XML_CIB_TAG_STATE
#define XML_CIB_TAG_STATE
Definition: msg_xml.h:165
pe_base_name_end
const char * pe_base_name_end(const char *id)
Definition: unpack.c:1536
pe__is_guest_node
bool pe__is_guest_node(const pe_node_t *node)
Definition: remote.c:33
pe_node_shared_s::unpacked
gboolean unpacked
Definition: pe_types.h:216
pe_resource_s::failure_timeout
int failure_timeout
Definition: pe_types.h:330
crm_internal.h
pe_check_last_failure
Definition: pe_types.h:188
pe_find_clone
match only clone instances
Definition: pe_types.h:85
pe_rsc_start_pending
#define pe_rsc_start_pending
Definition: pe_types.h:257
pe_warn_once
#define pe_warn_once(pe_wo_bit, fmt...)
Definition: internal.h:153
ticket_new
pe_ticket_t * ticket_new(const char *ticket_id, pe_working_set_t *data_set)
Definition: utils.c:1957
CRM_ATTR_UNAME
#define CRM_ATTR_UNAME
Definition: crm.h:110
XML_NODE_IS_PEER
#define XML_NODE_IS_PEER
Definition: msg_xml.h:248
pe_tag_s::refs
GListPtr refs
Definition: pe_types.h:447
crm_trace
#define crm_trace(fmt, args...)
Definition: logging.h:353
CRMD_JOINSTATE_MEMBER
#define CRMD_JOINSTATE_MEMBER
Definition: crm.h:161
do_crm_log
#define do_crm_log(level, fmt, args...)
Log a message.
Definition: logging.h:156
op_digest_cache_s::rc
enum rsc_digest_cmp_val rc
Definition: internal.h:516
xml_contains_remote_node
bool xml_contains_remote_node(xmlNode *xml)
Definition: remote.c:84
pcmk__str_casei
Definition: strings_internal.h:26
pcmk_is_set
#define pcmk_is_set(g, f)
Convenience alias for pcmk_all_flags_set(), to check single flag.
Definition: util.h:196
add_node_copy
xmlNode * add_node_copy(xmlNode *new_parent, xmlNode *xml_node)
Definition: xml.c:641
pe_flag_startup_fencing
#define pe_flag_startup_fencing
Definition: pe_types.h:108
pe_working_set_s::stonith_action
const char * stonith_action
Definition: pe_types.h:134
pe_node_s::details
struct pe_node_shared_s * details
Definition: pe_types.h:233
pe_resource_s::running_on
GListPtr running_on
Definition: pe_types.h:353
get_target_role
gboolean get_target_role(pe_resource_t *rsc, enum rsc_role_e *role)
Definition: utils.c:1845
crm_log_xml_debug
#define crm_log_xml_debug(xml, text)
Definition: logging.h:360
pe__shutdown_requested
bool pe__shutdown_requested(pe_node_t *node)
Definition: utils.c:2712
pe_rsc_needs_fencing
#define pe_rsc_needs_fencing
Definition: pe_types.h:268
order_actions
gboolean order_actions(pe_action_t *lh_action, pe_action_t *rh_action, enum pe_ordering order)
Definition: utils.c:1883
pe_resource_s::flags
unsigned long long flags
Definition: pe_types.h:335
pe_node_shared_s::uname
const char * uname
Definition: pe_types.h:198
pe_rsc_promotable
#define pe_rsc_promotable
Definition: pe_types.h:245
XML_TAG_META_SETS
#define XML_TAG_META_SETS
Definition: msg_xml.h:171
xml.h
Wrappers for and extensions to libxml2.
pe_working_set_s::config_hash
GHashTable * config_hash
Definition: pe_types.h:142
XML_ATTR_UNAME
#define XML_ATTR_UNAME
Definition: msg_xml.h:118
pe_resource_s::clone_name
char * clone_name
Definition: pe_types.h:309
add_tag_ref
gboolean add_tag_ref(GHashTable *tags, const char *tag_name, const char *obj_ref)
Definition: utils.c:2584
XML_RSC_ATTR_MANAGED
#define XML_RSC_ATTR_MANAGED
Definition: msg_xml.h:202
create_xml_node
xmlNode * create_xml_node(xmlNode *parent, const char *name)
Definition: xml.c:663
pe_resource_s::lock_time
time_t lock_time
Definition: pe_types.h:372
xml_internal.h
pe_ticket_s
Definition: pe_types.h:437
pe_flag_stonith_enabled
#define pe_flag_stonith_enabled
Definition: pe_types.h:97
unpack_status
gboolean unpack_status(xmlNode *status, pe_working_set_t *data_set)
Definition: unpack.c:1129
crm_element_value_epoch
int crm_element_value_epoch(const xmlNode *xml, const char *name, time_t *dest)
Retrieve the seconds-since-epoch value of an XML attribute.
Definition: nvpair.c:650
pe_node_attribute_raw
const char * pe_node_attribute_raw(pe_node_t *node, const char *name)
Definition: common.c:629
pe_ticket_s::last_granted
time_t last_granted
Definition: pe_types.h:440
pe__create_clone_child
pe_resource_t * pe__create_clone_child(pe_resource_t *rsc, pe_working_set_t *data_set)
Definition: clone.c:60
resource_object_functions_s::free
void(* free)(pe_resource_t *)
Definition: pe_types.h:54
XML_LRM_ATTR_MIGRATE_TARGET
#define XML_LRM_ATTR_MIGRATE_TARGET
Definition: msg_xml.h:293
CIB_OPTIONS_FIRST
#define CIB_OPTIONS_FIRST
Definition: msg_xml.h:49
pe_node_shared_s::standby
gboolean standby
Definition: pe_types.h:203
XML_RSC_ATTR_REMOTE_NODE
#define XML_RSC_ATTR_REMOTE_NODE
Definition: msg_xml.h:215
XML_LRM_ATTR_EXIT_REASON
#define XML_LRM_ATTR_EXIT_REASON
Definition: msg_xml.h:285
XML_LRM_ATTR_RESTART_DIGEST
#define XML_LRM_ATTR_RESTART_DIGEST
Definition: msg_xml.h:283
no_quorum_ignore
Definition: pe_types.h:64
pe_node_shared_s::expected_up
gboolean expected_up
Definition: pe_types.h:209
free_xml
void free_xml(xmlNode *child)
Definition: xml.c:790
no_quorum_demote
Definition: pe_types.h:66
pe_node_s
Definition: pe_types.h:229
pe_resource_s::variant
enum pe_obj_types variant
Definition: pe_types.h:317
pe_working_set_s::input
xmlNode * input
Definition: pe_types.h:128
pe_ticket_s::granted
gboolean granted
Definition: pe_types.h:439
XML_CIB_TAG_NODE
#define XML_CIB_TAG_NODE
Definition: msg_xml.h:166
pe_working_set_s::placement_strategy
const char * placement_strategy
Definition: pe_types.h:135
pcmk__epoch2str
const char * pcmk__epoch2str(time_t *when)
Definition: iso8601.c:1715
pe_find_resource
pe_resource_t * pe_find_resource(GListPtr rsc_list, const char *id_rh)
Definition: status.c:383
pe_node_shared_s::id
const char * id
Definition: pe_types.h:197
pcmk__op_key
char * pcmk__op_key(const char *rsc_id, const char *op_type, guint interval_ms)
Generate an operation key (RESOURCE_ACTION_INTERVAL)
Definition: operations.c:44
pe_ticket_s::id
char * id
Definition: pe_types.h:438
pe_working_set_s::shutdown_lock
guint shutdown_lock
Definition: pe_types.h:180
no_quorum_stop
Definition: pe_types.h:63
target
const char * target
Definition: pcmk_fence.c:29
native_add_running
void native_add_running(pe_resource_t *rsc, pe_node_t *node, pe_working_set_t *data_set)
Definition: native.c:83
pe_can_fence
bool pe_can_fence(pe_working_set_t *data_set, pe_node_t *node)
Definition: utils.c:87
pcmk__score_green
int pcmk__score_green
Definition: utils.c:57
msg_xml.h
XML_ATTR_TRANSITION_KEY
#define XML_ATTR_TRANSITION_KEY
Definition: msg_xml.h:368
pe_node_shared_s::rsc_discovery_enabled
gboolean rsc_discovery_enabled
Definition: pe_types.h:212
CRM_XS
#define CRM_XS
Definition: logging.h:54
pe_node_shared_s::running_rsc
GListPtr running_rsc
Definition: pe_types.h:220
pe_working_set_s::dc_node
pe_node_t * dc_node
Definition: pe_types.h:133
pe_fence_node
void pe_fence_node(pe_working_set_t *data_set, pe_node_t *node, const char *reason, bool priority_delay)
Schedule a fence action for a node.
Definition: unpack.c:97
pe_create_node
pe_node_t * pe_create_node(const char *id, const char *uname, const char *type, const char *score, pe_working_set_t *data_set)
Definition: unpack.c:396
pe_node_shared_s
Definition: pe_types.h:196
pe_working_set_s::localhost
const char * localhost
Definition: pe_types.h:170
pe_flag_quick_location
#define pe_flag_quick_location
Definition: pe_types.h:115
pe_resource_s::is_remote_node
gboolean is_remote_node
Definition: pe_types.h:338
pe_resource_s::pending_node
pe_node_t * pending_node
Definition: pe_types.h:370
fail2text
const char * fail2text(enum action_fail_response fail)
Definition: common.c:318
pe_resource_s::children
GListPtr children
Definition: pe_types.h:364
pcmk__score_red
int pcmk__score_red
Definition: utils.c:56
XML_LRM_TAG_RESOURCES
#define XML_LRM_TAG_RESOURCES
Definition: msg_xml.h:233
crm_err
#define crm_err(fmt, args...)
Definition: logging.h:347
XML_CIB_TAG_TICKET_STATE
#define XML_CIB_TAG_TICKET_STATE
Definition: msg_xml.h:396
set_config_flag
#define set_config_flag(data_set, option, flag)
Definition: unpack.c:35
CRM_ASSERT
#define CRM_ASSERT(expr)
Definition: results.h:42
xpath_search
xmlXPathObjectPtr xpath_search(xmlNode *xml_top, const char *path)
Definition: xpath.c:138
pe__clear_failcount
pe_action_t * pe__clear_failcount(pe_resource_t *rsc, pe_node_t *node, const char *reason, pe_working_set_t *data_set)
Schedule a controller operation to clear a fail count.
Definition: failcounts.c:360
pe_rule_eval_data::node_hash
GHashTable * node_hash
Definition: common.h:173
decode_transition_key
gboolean decode_transition_key(const char *key, char **uuid, int *transition_id, int *action_id, int *target_rc)
Parse a transition key into its constituent parts.
Definition: operations.c:317
rsc_action_digest_cmp
op_digest_cache_t * rsc_action_digest_cmp(pe_resource_t *rsc, xmlNode *xml_op, pe_node_t *node, pe_working_set_t *data_set)
Definition: utils.c:2148
XML_ATTR_HAVE_WATCHDOG
#define XML_ATTR_HAVE_WATCHDOG
Definition: msg_xml.h:86
XML_NODE_ATTR_RSC_DISCOVERY
#define XML_NODE_ATTR_RSC_DISCOVERY
Definition: msg_xml.h:349
unpack_nodes
gboolean unpack_nodes(xmlNode *xml_nodes, pe_working_set_t *data_set)
Definition: unpack.c:538
pcmk__score_yellow
int pcmk__score_yellow
Definition: utils.c:58
pe_tag_s
Definition: pe_types.h:445
CRMD_ACTION_METADATA
#define CRMD_ACTION_METADATA
Definition: crm.h:186
XML_LRM_ATTR_INTERVAL_MS
#define XML_LRM_ATTR_INTERVAL_MS
Definition: msg_xml.h:265
XML_LRM_ATTR_CALLID
#define XML_LRM_ATTR_CALLID
Definition: msg_xml.h:279
CRMD_ACTION_MIGRATE
#define CRMD_ACTION_MIGRATE
Definition: crm.h:168
XML_NVPAIR_ATTR_VALUE
#define XML_NVPAIR_ATTR_VALUE
Definition: msg_xml.h:347
crm_str_hash
#define crm_str_hash
Definition: util.h:62
pe_node_shared_s::utilization
GHashTable * utilization
Definition: pe_types.h:224
iso8601_internal.h
pe_node_shared_s::shutdown
gboolean shutdown
Definition: pe_types.h:208
data
char data[0]
Definition: internal.h:90
pe__clear_resource_flags
#define pe__clear_resource_flags(resource, flags_to_clear)
Definition: internal.h:53
crm_str
#define crm_str(x)
Definition: logging.h:373
XML_LRM_ATTR_OPSTATUS
#define XML_LRM_ATTR_OPSTATUS
Definition: msg_xml.h:277
no_quorum_suicide
Definition: pe_types.h:65
pe__set_working_set_flags
#define pe__set_working_set_flags(working_set, flags_to_set)
Definition: internal.h:35
CRMD_JOINSTATE_PENDING
#define CRMD_JOINSTATE_PENDING
Definition: crm.h:160
op_digest_cache_s
Definition: internal.h:515
rsc_role_e
rsc_role_e
Definition: common.h:91
pe_rsc_block
#define pe_rsc_block
Definition: pe_types.h:239
pe_status_private.h
CRM_TRACE_INIT_DATA
CRM_TRACE_INIT_DATA(pe_status)
pe_node_shared_s::maintenance
gboolean maintenance
Definition: pe_types.h:211
XML_LRM_ATTR_RC
#define XML_LRM_ATTR_RC
Definition: msg_xml.h:278
pcmk_ok
#define pcmk_ok
Definition: results.h:67
pe_node_shared_s::digest_cache
GHashTable * digest_cache
cache of calculated resource digests
Definition: pe_types.h:225
pe__is_universal_clone
bool pe__is_universal_clone(pe_resource_t *rsc, pe_working_set_t *data_set)
Definition: clone.c:1170
pe_rsc_failure_ignored
#define pe_rsc_failure_ignored
Definition: pe_types.h:263
XML_NODE_JOIN_STATE
#define XML_NODE_JOIN_STATE
Definition: msg_xml.h:245
destroy_ticket
void destroy_ticket(gpointer data)
Definition: utils.c:1945
XML_CIB_TAG_STATUS
#define XML_CIB_TAG_STATUS
Definition: msg_xml.h:146
XML_CIB_TAG_OBJ_REF
#define XML_CIB_TAG_OBJ_REF
Definition: msg_xml.h:400
common_unpack
gboolean common_unpack(xmlNode *xml_obj, pe_resource_t **rsc, pe_resource_t *parent, pe_working_set_t *data_set)
Definition: complex.c:396
XML_NODE_IN_CLUSTER
#define XML_NODE_IN_CLUSTER
Definition: msg_xml.h:247
crm_is_true
gboolean crm_is_true(const char *s)
Definition: strings.c:392
pe_flag_have_stonith_resource
#define pe_flag_have_stonith_resource
Definition: pe_types.h:98
CRM_ATTR_SITE_NAME
#define CRM_ATTR_SITE_NAME
Definition: crm.h:116
XML_CIB_TAG_GROUP
#define XML_CIB_TAG_GROUP
Definition: msg_xml.h:182
pe_flag_enable_unfencing
#define pe_flag_enable_unfencing
Definition: pe_types.h:99
XML_LRM_TAG_RSC_OP
#define XML_LRM_TAG_RSC_OP
Definition: msg_xml.h:235
pe_rsc_trace
#define pe_rsc_trace(rsc, fmt, args...)
Definition: internal.h:20
find_operations
GListPtr find_operations(const char *rsc, const char *node, gboolean active_filter, pe_working_set_t *data_set)
Definition: unpack.c:3957
pe_wo
uint32_t pe_wo
Definition: unpack.c:68
ID
#define ID(x)
Definition: msg_xml.h:425
pe_working_set_s::flags
unsigned long long flags
Definition: pe_types.h:137
pe_err
#define pe_err(fmt...)
Definition: internal.h:22
calculate_active_ops
void calculate_active_ops(GListPtr sorted_op_list, int *start_index, int *stop_index)
Definition: unpack.c:2191
pe_action_optional
Definition: pe_types.h:282
ONLINESTATUS
#define ONLINESTATUS
Definition: util.h:37
pe__resource_actions
GList * pe__resource_actions(const pe_resource_t *rsc, const pe_node_t *node, const char *task, bool require_node)
Find all actions of given type for a resource.
Definition: utils.c:1621
pe_node_shared_s::standby_onfail
gboolean standby_onfail
Definition: pe_types.h:204
pe__clear_resource_history
pe_action_t * pe__clear_resource_history(pe_resource_t *rsc, pe_node_t *node, pe_working_set_t *data_set)
Definition: utils.c:2787
name
char * name
Definition: pcmk_fence.c:31
XML_LRM_ATTR_MIGRATE_SOURCE
#define XML_LRM_ATTR_MIGRATE_SOURCE
Definition: msg_xml.h:292
resource_object_functions_s::find_rsc
pe_resource_t *(* find_rsc)(pe_resource_t *parent, const char *search, const pe_node_t *node, int flags)
Definition: pe_types.h:45
freeXpathObject
void freeXpathObject(xmlXPathObjectPtr xpathObj)
Definition: xpath.c:38
CRM_ATTR_ID
#define CRM_ATTR_ID
Definition: crm.h:111
sort_node_uname
gint sort_node_uname(gconstpointer a, gconstpointer b)
Definition: utils.c:215
pe_node_shared_s::unclean
gboolean unclean
Definition: pe_types.h:206
crm_strdup_printf
char * crm_strdup_printf(char const *format,...) __attribute__((__format__(__printf__
XPATH_ENABLE_UNFENCING
#define XPATH_ENABLE_UNFENCING
Definition: unpack.c:177
GListPtr
GList * GListPtr
Definition: crm.h:214
pe_flag_start_failure_fatal
#define pe_flag_start_failure_fatal
Definition: pe_types.h:106
pe_node_shared_s::type
enum node_type type
Definition: pe_types.h:199
XML_CIB_TAG_TICKETS
#define XML_CIB_TAG_TICKETS
Definition: msg_xml.h:395
pe_working_set_s::now
crm_time_t * now
Definition: pe_types.h:129
node_ping
Definition: pe_types.h:70
crm_info
#define crm_info(fmt, args...)
Definition: logging.h:350
pe_rsc_managed
#define pe_rsc_managed
Definition: pe_types.h:238
pe_rsc_orphan
#define pe_rsc_orphan
Definition: pe_types.h:237
pe_working_set_s::template_rsc_sets
GHashTable * template_rsc_sets
Definition: pe_types.h:169
pcmk__config_warn
#define pcmk__config_warn(fmt...)
Definition: internal.h:119
pe_find_node_any
pe_node_t * pe_find_node_any(GListPtr node_list, const char *id, const char *uname)
Definition: status.c:407
pe_node_shared_s::online
gboolean online
Definition: pe_types.h:202
clone_strip
char * clone_strip(const char *last_rsc_id)
Definition: unpack.c:1576
pe_get_failcount
int pe_get_failcount(pe_node_t *node, pe_resource_t *rsc, time_t *last_failure, uint32_t flags, xmlNode *xml_op, pe_working_set_t *data_set)
Definition: failcounts.c:251
pe_working_set_s::stop_needed
GList * stop_needed
Definition: pe_types.h:177
pe_resource_s::parent
pe_resource_t * parent
Definition: pe_types.h:315
pe_fence_op
pe_action_t * pe_fence_op(pe_node_t *node, const char *op, bool optional, const char *reason, bool priority_delay, pe_working_set_t *data_set)
Definition: utils.c:2443
pe_free_action
void pe_free_action(pe_action_t *action)
Definition: utils.c:1419
pe_flag_shutdown_lock
#define pe_flag_shutdown_lock
Definition: pe_types.h:109
type
enum crm_ais_msg_types type
Definition: internal.h:83
pe_rsc_info
#define pe_rsc_info(rsc, fmt, args...)
Definition: internal.h:18
pe_resource_s::id
char * id
Definition: pe_types.h:308
pe_resource_s::allowed_nodes
GHashTable * allowed_nodes
Definition: pe_types.h:355
custom_action
pe_action_t * custom_action(pe_resource_t *rsc, char *key, const char *task, pe_node_t *on_node, gboolean optional, gboolean foo, pe_working_set_t *data_set)
Definition: utils.c:465
pe_flag_startup_probes
#define pe_flag_startup_probes
Definition: pe_types.h:111
CRMD_ACTION_STATUS
#define CRMD_ACTION_STATUS
Definition: crm.h:185
pe_flag_stop_rsc_orphans
#define pe_flag_stop_rsc_orphans
Definition: pe_types.h:102
Generated on Wed Dec 2 2020 15:53:07 for pacemaker by   doxygen 1.8.15