Bug 1461557

Summary:	[GANESHA] Document the steps for NFS-Ganesha to support the rolling upgrade
Product:	[Red Hat Storage] Red Hat Gluster Storage	Reporter:	Manisha Saini <msaini>
Component:	Documentation	Assignee:	Laura Bailey <lbailey>
Status:	CLOSED CURRENTRELEASE	QA Contact:	Manisha Saini <msaini>
Severity:	unspecified	Docs Contact:
Priority:	high
Version:	rhgs-3.3	CC:	kkeithle, msaini, rhinduja, rhs-bugs, sankarshan, skoduri, storage-doc, storage-qa-internal
Target Milestone:	---
Target Release:	RHGS 3.3.0
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2017-09-21 04:28:58 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:
Embargoed:
Bug Depends On:
Bug Blocks:	1417157

Description Manisha Saini 2017-06-14 19:07:54 UTC

Document URL: 
https://access.redhat.com/documentation/en-us/red_hat_gluster_storage/3.2/html/installation_guide/ch07s02#NFS_Ganesha_Update

Section Number and Name: 
7.2.2 -Instead of offline upgrade,add the support for rolling upgrade for NFS-Ganesha

Describe the issue: 
Till 3.2 only offline Upgrade was supported.
Need to document the rolling upgrade steps for NFS-Ganesha.

Suggestions for improvement: 

Additional information: 

1st attempt-

Perform following steps one by one on each node-

1.Set the ganesha cluster on standby node on 1st Node

#pcs cluster standby

2.Stop the pcs cluster on 1st node
#pcs cluster stop

3.Disable pcs cluster

#pcs cluster disable

4.Check status of pcs cluster.It should not be running
#pcs status

5.Kill gluster services

#service glusterd stop
#pkill glusterfs
#pkill glusterfsd
#pgrep gluster

6.Perform yum update to update the ganesha from 3.2 to 3.3

#yum update

7.Reboot the 1st node
#reboot

8.When the node comes up,startthe gluster service
#service glusterd start

9.Verify the shared_storage is mounted after the node comes up

10.Start the ganesha service
#service nfs-ganesha start

11.Enable the pcs cluster
#pcs cluster enable

12.Start the cluster
#pcs cluster start

13.Release the cluster from standby mode
#pcs cluster unstandby


14.Check the pcs status.pcs cluster should be up and running
#pcs status

15.Check the volume is exported after upgrade
#showmount -e

16.Repeat the above steps on all nodes one by one.

Comment 2 Manisha Saini 2017-06-14 19:10:59 UTC

Soumya and Kaleb,

Requesting to please review the steps mentioned for rolling upgrade and provide your inputs.

Comment 3 Kaleb KEITHLEY 2017-06-15 10:34:26 UTC

LGTM. Steps 5, 8, and 10 should prefer RHEL7 systemd commands instead of legacy init.d compat commands:

...
5) #systemctl stop glusterd
...
8) #systemctl start glusterd
...
10) #systemctl start nfs-ganesha

Comment 6 Soumya Koduri 2017-07-04 09:05:11 UTC

(In reply to Kaleb KEITHLEY from comment #3)
> LGTM. Steps 5, 8, and 10 should prefer RHEL7 systemd commands instead of
> legacy init.d compat commands:
> 
> ...
> 5) #systemctl stop glusterd
> ...
> 8) #systemctl start glusterd
> ...
> 10) #systemctl start nfs-ganesha

+1.. Since we support NFS-Ganesha only on RHEL-7 from now on. services should be started or stopped via systemd commands. 

@Manisha,

When are we stopping nfs-ganesha service. I think it should between step(4) and step (5). Could you please confirm?

Comment 7 Manisha Saini 2017-07-04 09:09:55 UTC

(In reply to Soumya Koduri from comment #6)
> (In reply to Kaleb KEITHLEY from comment #3)
> > LGTM. Steps 5, 8, and 10 should prefer RHEL7 systemd commands instead of
> > legacy init.d compat commands:
> > 
> > ...
> > 5) #systemctl stop glusterd
> > ...
> > 8) #systemctl start glusterd
> > ...
> > 10) #systemctl start nfs-ganesha
> 
> +1.. Since we support NFS-Ganesha only on RHEL-7 from now on. services
> should be started or stopped via systemd commands. 
> 
> @Manisha,
> 
> When are we stopping nfs-ganesha service. I think it should between step(4)
> and step (5). Could you please confirm?

Soumya,
When i tested the rolling upgrade,I didn't stopped the ganesha service and it worked fine for me.
Do this step needs to be included?

Comment 8 Soumya Koduri 2017-07-04 09:12:16 UTC

yes.. because we may be updating nfs-ganesha package too and the clean way would be to stop the service before updating the binary.

Comment 15 Manisha Saini 2017-07-26 12:45:36 UTC

Clearing the needinfo.


Laura,

I have few comments,updated the doc.I have requested the team to provide inputs.
Rest look good to me.

Comment 21 Manisha Saini 2017-08-21 18:50:20 UTC

Changes looks good to me. Verified these steps for rolling upgrade of NFS-Ganesha.Moving the bug to verified state.

Comment 22 Laura Bailey 2017-08-29 04:12:36 UTC

Fixed in RHGS 3.3 documentation.