Bug 1889506

Summary: [GSS] [rhel-7] rpcinfo -p localhost shows only nlockmgr version 4
Product: [Red Hat Storage] Red Hat Gluster Storage Reporter: slenzen
Component: nfs-ganeshaAssignee: Frank Filz <ffilz>
Status: CLOSED ERRATA QA Contact: Upasana <ubansal>
Severity: high Docs Contact:
Priority: high    
Version: rhgs-3.5CC: arjsharm, asakthiv, dang, ffilz, grajoria, kkeithle, mbenjamin, mmuench, pprakash, puebele, rcyriac, rhs-bugs, skoduri, storage-qa-internal, ubansal
Target Milestone: ---Keywords: ZStream
Target Release: RHGS 3.5.z Batch Update 4Flags: puebele: needinfo+
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: libntirpc-1.7.3-4.el7rhgs Doc Type: Bug Fix
Doc Text:
Previously, the local RPCBIND service hung because the connected sockets were not closed properly, causing failures. With this update, the sockets are closed properly when not needed and the system works as expected.
Story Points: ---
Clone Of:
: 1898998 (view as bug list) Environment:
Last Closed: 2021-04-29 06:41:22 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: 1898998    

Comment 4 Frank Filz 2020-10-21 14:25:03 UTC
Are there any errors in the Ganesha log?

Also, what is the client side mount command? And just to make sure, they aren't trying to mount from the same host running Ganesha?

Comment 46 Frank Filz 2020-11-20 21:22:01 UTC
To verify you can run cthon04 lock tests over an NFSv3 mount to the server.

Check the sockets before and after with this:

netstat -eap | grep `pidof ganesha.nfsd`

Crosscheck with looking at the open files before and after:

ls -l /proc/`pidof ganesha.nfsd`/fd

You can correlate the sockets with the fds

We expect to not see than a few udp sockets (rquotad if that is running, nfs, and maybe 2 others). With the bug not fixed, the cthon04 lock tests should result in a bunch of open sockets that may be labeled as statd.

Comment 48 Frank Filz 2020-11-23 16:04:11 UTC
I should have been not so lazy...

To run the lock tests:

[root@dhcp35-93 cthon04]# ./server -lf -o vers=3 -p /arbiter-vol -m /arbiter-vol -N 1 dhcp35-100.lab.eng.blr.redhat.com

Comment 49 Frank Filz 2020-11-23 16:05:10 UTC
It would also be good to verify the socket output before/after cthon04 without the fix.

Comment 54 Frank Filz 2020-12-02 15:02:35 UTC
The without fix definitely shows ports not closed after cthon04 run while with fix it shows the ports closed after the cthon04 run. So looks good.

Comment 68 Daniel Gryniewicz 2021-03-05 14:12:22 UTC
Looks good.

Comment 70 errata-xmlrpc 2021-04-29 06:41:22 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory (nfs-ganesha bug fix and enhancement update), and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHBA-2021:1459