FAQ
From ClusterLabs
If you have suggestions for possible FAQ entries, please send them to: andrew@beekhof.net
Organizational
Why Can't I Create a Wiki Account?
At the moment I've not sorted out the site security so I've disabled the creation of new users. If you'd like an account, just email us at pacemaker@oss.clusterlabs.org and we can create one manually for you.
Why was the Project Started?
Pacemaker grew out of the Heartbeat project.
See the project history for more details.
Why is the Project Called Pacemaker?
First of all, the reason it's not called the CRM is because of the abundance of terms that are commonly abbreviated to those three letters.
The Pacemaker name came from Kham, a good friend of mine, and was originally used by a Java GUI that I was prototyping in early 2007. Alas other commitments have prevented the GUI from progressing much and, when it came time to choose a name for this project, Lars suggested it was an even better fit for an independent CRM.
The idea stems from the analogy between the role of this software and that of the little device that keeps the human heart pumping.
Pacemaker monitors the cluster and intervenes when necessary to ensure the smooth operation of the services it provides.
There were a number of other names (and acronyms) tossed around, but suffice to say Pacemaker was the best of the lot :-)
What is the Project's Relationship with Heartbeat?
Pacemaker keeps your applications running when they or the machines they're running on fail. However it can't do this without connectivity to the other machines in the cluster - a significant problem in its own right.
Rather than re-implement the wheel, Pacemaker supports existing implimentations such as Heartbeat. Heartbeat provides:
- a mechansigm to reliably send messages between nodes,
- notifications when machines appear and disappear
- a list of machines that are up that is consistent throughout the cluster
Heartbeat was also the first stack supported by the Pacemaker codebase.
What is the Project's Relationship with Corosync?
Pacemaker keeps your applications running when they or the machines they're running on fail. However it can't do this without connectivity to the other machines in the cluster - a significant problem in its own right.
Rather than re-implement the wheel, Pacemaker supports existing implimentations such as Corosync. Corosync provides pacemaker:
- a mechansigm to reliably send messages between nodes,
- notifications when machines appear and disappear
- a list of machines that are up that is consistent throughout the cluster
Corosync was the second stack supported by the Pacemaker codebase.
What is the Project's Relationship with OpenAIS?
This one is tricky.
Originally Corosync and OpenAIS were the same thing. Then they split into two parts... the core messaging and membership capabilities are now called Corosync, and OpenAIS retained the layer containing the implementation of the AIS standard.
Pacemaker itself only needs the Corosync piece in order to function, however some of the applications it can manage (such as OCFS2 and GFS2) require the OpenAIS layer as well.
What Versions of Pacemaker Are Supported?
When seeking assistance, please try to ensure you have one of the versions supported directly by the project. Please refer to the Releases page for further details including the schedule of planned releases.
| Version | Current Release | Tested w/ Glue | Tested w/ Corosync | Tested w/ Heartbeat | First Released | This Release | Supported Until |
|---|---|---|---|---|---|---|---|
| 1.1 | 1.1.2.1 | 1.0.5 | 1.2.1 | 3.0.3 | Jan 15, 2010 | May 17, 2010 | Q4 2010 (approx) |
| 1.0 | 1.0.8 | 1.0.3 | 1.2.1 | 3.0.3 | Oct 9, 2008 | Mar 16, 2010 | Q4 2011 (at least) |
| Version | Last Release | First Released | Last Released |
|---|---|---|---|
| 0.7 | 0.7.3 | June 25, 2008 | Sep 22, 2008 |
| 0.6 | 0.6.7 | Jan 16, 2008 | Dec 15, 2008 |
Technical
How Do I Install Pacemaker?
Installation from source and from pre-built packages is described on the Install page.
Should I Run Pacemaker on Heartbeat or Coroysnc?
In a word, Corosync.
Red Hat and Novell/SUSE are both backing Corosync (think about it, when was the last time they agreed on anything :-)
Development of the Heartbeat project has been effectively halted. Linbit has taken over maintenance but has publicly stated that there will be no new features.
You also get the capability to manage GFS2 and OCFS2 partitions.
Can I Choose which Messaging Layer to use at Run Time?
Yes. The CRM will automatically detect who started it and behave accordingly.
Can I Have a Mixed Heartbeat-OpenAIS Cluster?
No.
Where Can I Get the Source Code?
- Latest Stable Series (1.0)
- The source code can be browsed at http://hg.clusterlabs.org/pacemaker/stable-1.0/ or downloaded as a tar-ball
- Previous Stable Series (0.6)
- The source code can be browsed at http://hg.clusterlabs.org/pacemaker/stable-0.6/ or downloaded as a tar-ball
- Latest and greatest
- The source code can be browsed at http://hg.clusterlabs.org/pacemaker/devel/ or downloaded as a tar-ball
- Alternatively, you can get a full copy of the Mercurial repository by executing
hg clone http://hg.clusterlabs.org/pacemaker/devel pacemaker
Where Can I Get Pre-built Packages?
Most users should be able to install Pacemaker directly from their distribution.
Pacemaker currently ships with Fedora (since 12), Red Hat Enterprise Linux (since 6.0 beta1), openSUSE (since 11.0), Debian Unstable, Ubuntu LTS (since 10.4 "Lucid Lynx”) and as a key component of the High Availability Extension for SUSE Linux Enterprise Server 11 (available free of charge to existing SLES10 customers).
It will also be part of Debian Squeeze.
The latest official packages from the project are available for most major rpm-based distributions from: http://www.clusterlabs.org/rpm/
Users of other distributions should refer to our Install page.
What Do the Prefixes in Changelog Mean?
- High, Med, Low: These all indicate how much the end-user/admin should care about the change.
- Dev: These are changes that fix bugs that don't exist in any released version of the project
Examples:
- High - Preventing a segfault, implementing an important new feature or major changes to the behavior of a feature
- Med - Hard to trigger bugs, bugs with workarounds, minor functional changes
- Low - Non-functional changes, formatting or logging changes, changes to test code
How Do I Test My Cluster?
Pacemaker comes with a Cluster Test Suite (CTS for short) which is an integral part of our release testing. Traditionally this had been hard to set up and use however a new tool has been written to simplify the process.
It can be found at: http://hg.clusterlabs.org/pacemaker/stable-1.0/file/tip/cts/cluster_test
Please give it a try and send feedback via the mailing list.
Resource is Too Active
Pacemaker will try and determine what resources are active on a machine when it starts. To do this, it sends what we call a probe which uses the monitor operation of your ResourceAgent.
There are two common reasons for seeing this message:
- Your resource really is active on more than one node
- Check you are _not_ starting it on boot
- Did Pacemaker suffer an internal failure? If so, please check the Help:Contents page and report it
- Your resource doesn't implement the monitor operation correctly
- Make sure your Resource Agent conforms to the OCF-spec by using the ocf-tester script
I Killed a Node but the Cluster Didn't Recover
One of the most common reasons for this is the way quorum is calculated for a 2-node cluster. Unlike Heartbeat, OpenAIS doesn't pretend 2-node clusters always have quorum.
In order to have quorum, more than half of the total number of cluster nodes need to be online. Clearly this is not the case when a node failure occurs in a 2-node cluster.
If you want to allow the remaining node to provide all the cluster services, you need to set the no-quorum-policy to ignore.
crm configure property no-quorum-policy=ignore
This provides the same behavior as Heartbeat, just be sure to set up STONITH to ensure data integrity.
How Do I Upgrade from Older Versions of Heartbeat?
If you plan to continue using the Heartbeat stack (as opposed to Corosync), check out the step-by-step guide on their website.
Features
How Do I Enable the GUI? (Corosync)
First you need to install the pacemaker-pygui package. Then you need to find the following lines in corosync.conf
service {
# Load the Pacemaker Cluster Resource Manager
name: pacemaker
ver: 0
}
and add
use_mgmtd: 1
before the closing bracket.
How Do I Enable the GUI? (Heartbeat)
First you need to install the pacemaker-pygui package. Then you need to add the following lines to ha.cf
apiauth mgmtd uid=root respawn root /usr/lib/heartbeat/mgmtd -v
These used to be implied when crm yes was present but only when heartbeat is built with the built-in mgmtd (which it no longer is).
NOTE: People on 64-bit platforms will probably need to replace lib with lib64
Collocation Sets
The sequential option does not refer to ordering. Instead it tells Pacemaker to create a collocation chain between the members of the set.
Ie.
colocation myset inf: app1 app2 app3 app4
is the equivalent of
colocation myset-1 inf: app2 app1 colocation myset-2 inf: app3 app2 colocation myset-3 inf: app4 app3
Putting them in brackets sets sequential=false and removes the internal constraints. So
colocation myset inf: app1 ( app2 app3 app4 )
is actually the equivalent of
colocation myset-1 inf: app2 app1 colocation myset-2 inf: app3 app1 colocation myset-3 inf: app4 app1
