Bug 1402152

Summary: RHEL 6 Pacemaker documentation steps for enabling pcsd are incorrect
Product: Red Hat Enterprise Linux 6 Reporter: Sam Yangsao <syangsao>
Component: doc-High_Availability_Add-On_OverviewAssignee: Steven J. Levine <slevine>
Status: CLOSED CURRENTRELEASE QA Contact: ecs-bugs
Severity: high Docs Contact:
Priority: high    
Version: 6.8CC: cfeist, rhel-docs
Target Milestone: rcKeywords: Documentation
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2017-03-22 18:41:34 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:

Description Sam Yangsao 2016-12-06 22:37:11 UTC
### Document URL: 

https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Linux/6/html-single/Configuring_the_Red_Hat_High_Availability_Add-On_with_Pacemaker/index.html#s1-clustercreate-HAAR

### Section Number and Name: 

⁠3.1.1. Starting the pcsd daemon
The following commands start the pcsd service and enable pcsd at system start. These commands should be run on each node in the cluster.

# service pcsd start
# chkconfig pcsd on

### Describe the issue: 

RHEL 6 Pacemaker documentation steps for enabling pcsd are incorrect

This *only* enables pcsd, which is *not* needed.

'pcsd' does not start the pacemaker cluster services on a node in RHEL 6, the 'pacemaker' service is the one that starts it.

### Suggestions for improvement: 

Suggest that customers check to make sure that the 'pacemaker' service is turned on for pacemaker to run properly on a node upon reboot. 

### Additional information: 

# pcs cluster setup --start --name pacemaker_sap sap1,sap1hb sap2,sap2hb --rrpmode udpu
# chkconfig --list |egrep 'pcs|pacemaker|corosync'
corosync       	0:off	1:off	2:off	3:off	4:off	5:off	6:off
pacemaker      	0:off	1:off	2:off	3:off	4:off	5:off	6:off
pcsd           	0:off	1:off	2:off	3:off	4:off	5:off	6:off
rpcsvcgssd     	0:off	1:off	2:off	3:off	4:off	5:off	6:off

# pcs cluster enable --all
sap1: Cluster Enabled
sap2: Cluster Enabled

# chkconfig --list |egrep 'pcs|pacemaker|corosync'
corosync       	0:off	1:off	2:off	3:off	4:off	5:off	6:off
pacemaker      	0:off	1:off	2:on	3:on	4:on	5:on	6:off
pcsd           	0:off	1:off	2:off	3:off	4:off	5:off	6:off
rpcsvcgssd     	0:off	1:off	2:off	3:off	4:off	5:off	6:off

The above enables pacemaker via chkconfig which is needed to run the pacemaker services when the node gets rebooted.  pcsd is not needed and does not start pacemaker services when a node is rebooted.

# rpm -qa |egrep 'pcsd|pacemaker|corosync'
pacemaker-1.1.14-8.el6_8.1.x86_64
corosync-1.4.7-5.el6.x86_64
pacemaker-cluster-libs-1.1.14-8.el6_8.1.x86_64
pacemaker-cli-1.1.14-8.el6_8.1.x86_64
pacemaker-libs-1.1.14-8.el6_8.1.x86_64
corosynclib-1.4.7-5.el6.x86_64

Comment 1 Sam Yangsao 2016-12-06 23:17:46 UTC
OK, so from further testing, we do _need_ both services enabled.

# chkconfig pcsd on
# chkconfig pacemaker on

As, without pcsd the node would not rejoin the cluster properly from verifying this.

I would say, update the docs to make sure that 'pacemaker' service is turned by via chkconfig, even though we enable via the 'pcs cluster enable ...' command, we should also specify how to check that it gets started on reboot and how.

Comment 2 Steven J. Levine 2017-01-26 23:32:52 UTC
This BZ is related:

https://bugzilla.redhat.com/show_bug.cgi?id=1395639

Comment 3 Chris Feist 2017-01-27 00:03:25 UTC
Sam,

You can see if pacemaker is enabled by running 'pcs status' on a running cluster and looking at the 'Daemon Status' section to verify that pacemaker is enabled.

[root@host-600 lib]# pcs status
....  snip ....
Daemon Status:
  cman: active/disabled
  corosync: active/disabled
  pacemaker: active/enabled
  pcsd: active/disabled

However, if the cluster is not running, you would need to do a 'chkconfig pacemaker list' to verify.  I think we can have Steven update the docs to add both the 'pcs status' and 'chkconfig pacemaker list' checks to verify the cluster is enabled.

Comment 4 Steven J. Levine 2017-01-27 22:10:46 UTC
Chris -- premature needinfo (that was meant for another bug). Clearing for the moment.

Comment 5 Steven J. Levine 2017-01-27 22:26:43 UTC
Previously, the subsection on enabling cluster services was part of the "managing cluster nodes" section (as section 3.2.2).  I have moved it to the end of the "cluster creation" section (as section 3.1.4) since it is part of the general creation process. 

I have also added the information about checking to be sure that clsuter services/pacemaker is enabled to the end of that chapter.

The reason this was originally separated out from the general creation process is that not all sites would want to enable the cluster to start up on reboot -- if the cluster/node went down, you might want to be sure that node was fixed for sure before re-joining the cluster.  There was some discussion about whether we should even recomment that users enable pcsd like this in the first place.

But even separated out it does seem to belong in cluster creation.

The section with the new information is here, Section 3.1.4:

http://jenkinscat.gsslab.pnq.redhat.com:8080/job/doc-Red_Hat_Enterprise_Linux-6-Configuring_the_Red_Hat_High_Availability_Add-On_with_Pacemaker%20%28html-single%29/lastSuccessfulBuild/artifact/tmp/en-US/html-single/index.html#s2-enabledisableservices-HAAR

Chris -- could you verify that information?  thanks.

Comment 6 Chris Feist 2017-01-27 22:45:33 UTC
Reviewed those sections, they look good to me.