[ClusterLabs] Pacemaker 2.0.4-rc1 now available

Klaus Wenninger kwenning at redhat.com
Thu Apr 30 03:25:37 EDT 2020

Hi all,

I am happy to announce that source code for the first release candidate
for Pacemaker version 2.0.4 is now available at:


Some of the highlights of these release are:

* 'shutdown-locks'
This "feature" is implemented for parity with proprietary
high-availability software, to make it easier for users of that software
to switch to open source.
It is unlikely to be of interest to existing users of Pacemaker.

The scenario where this comes in useful is when an organization has a
large number of clusters with an active/failover model, with a high
cost (time-wise) to failing over, where host software updates and reboots
are performed during scheduled maintenance windows by junior system
administrators who may not have any cluster awareness. In such a case,
it is desirable that high availability is only for resource failures,
and resources should stop and not be recovered elsewhere for clean node
shutdowns. This must happen automatically, without any administrator
intervention, rather than via existing mechanisms such as temporarily
setting target-role.

The solution here is a new cluster option shutdown-lock which can be
set to true to enable this behavior.

* 'priority-fencing-delay'
Addresses general issue with mainly 2-node clusters where we usually
try to make the node running the most valuable set of resources win
a fencing-race.

With 'priority-fencing-delay' the priority of a node is derived
from the resource-priorities of resources it is running.
In a fencing-race the node with the highest priority has a certain
advantage over the others as fencing requests for that node are
executed with an additional (configurable) delay.

* couple of additional levers to further tailor output of tools
Examples are --include= --exclude= --node= as parameters for
crm_mon to include/exclude certain sections of output or
to limit output to data from certain nodes.

As usual, there were many bug fixes and log message improvements as
well. For more details about changes in this release, please see the
change log:


Everyone is encouraged to download, compile and test the new release.
We do many regression tests and simulations, but we can't cover all
possible use cases, so your feedback is important and appreciated.

Many thanks to all contributors of source code to this release,
including Aleksei Burlakov, Chris Lumens, Gao,Yan, Ferenc Wágner,
Hideo Yamauchi, Inoue Kazunori, Jan Pokorný, Ken Gaillot,
Klaus Wenninger, Sebastian Matuschka, Yoshiufuji Hideaki.


More information about the Users mailing list