RHEA-2018:2610 - Product Enhancement Advisory

Topic

Updated nfs-ganesha packages that fix various bugs are now available for Red Hat Gluster Storage 3.4 on Red Hat Enterprise Linux 7.

Description

NFS-Ganesha is a user space file server for the NFS protocol with support for NFS v3 and v4.0

The nfs-ganesha packages has been upgraded to upstream version 2.5.5, which provides a number of bug fixes and enhancements over the previous version. (#1472942)

This update adds the following enhancements:

• With this release, xreaddirplus option is now turned on by default. When xreaddirplus option is turned off, NFS falls back to standardreaddir instead of readdirp. Turning xreaddirplus off would result in more lookup and stat requests being sent from the client which may impact performance. (BZ#1489378)
• With Red Hat Gluster Storage 3.4, NFS-Ganesha log files have been moved to the /var/log/ganesha subdirectory. (BZ#1516699)
• NFS-Ganesha-2.5.x allows registration of NFSv3 on UDPv6. (BZ#1517790)

This advisory also fixes the following bugs:

• NFS Ganesha internally uses a different file descriptor for every lock request. When a client removes a file with a specific file descriptor, other clients who are trying to access the same file with the same file descriptor receive “No such file or directory” error. With this fix, lock request uses the same file descriptor obtained from an open call and no additional open call in the lock handling path is required. (BZ#1480138)
• To limit the NFS version to 3 and 4.0, NFSv4 blocks were added in ganesha.conf file manually. With this fix, these options are added as default in ganesha.conf file. (BZ#1514615)
• Currently, the default interval (in microseconds) between upcall polls is 10 microseconds. For large numbers of threads, this results in a lot of CPU consumption. With this fix, the default polling interval has been increased to 100 milliseconds (100000 microseconds) and helps reduce CPU consumption. (BZ#1481040)
• NFS-Ganesha crashed on NFS version 3 request if the GUARDED flag was set to true. The GUARDED flag checks for the presence of a duplicate file before performing the create or open call and should fail the request with NFS3ERR_EXIST if a duplicate file exists. With this fix, the GUARDED flag can handle the NFS version 3 create request and does not crash NFS-Ganesha. (BZ#1520751)
• While exporting/unexporting multiple volumes in a loop, NFS-Ganesha process goes into failed state and then restarts on its own. The export id cache entry is also updated by race. With this fix, while unexporting, refcount on entry is taken into account and the race is closed. (BZ#1577325)

Users of NFS-Ganesha with Red Hat Gluster Storage are advised to update to these packages which fixes these bugs and adds the enhancements.

Solution

Before applying this update, make sure all previously released errata
relevant to your system have been applied.

For details on how to apply this update, refer to:

https://access.redhat.com/articles/11258

Affected Products

• Red Hat Gluster Storage Server for On-premise 3 for RHEL 7 x86_64

Fixes

• BZ - 1427131 - [Scale] : I/O hangs on Ganesha mounts during add-brick/remove-brick on DR volumes.
• BZ - 1461507 - [Ganesha] : Ganesha logs flooded with "unable to find reclaimable dupreq LRU entry" messages
• BZ - 1461521 - [Ganesha] : Ganesha crashed while running Bonnie from multiple clients.
• BZ - 1467896 - [GANESHA] Ganesha crashed while running diskfill utility on nfs share mounted on windows client
• BZ - 1472445 - packaging: save ganesha config files in (/var)/run/gluster/shared_storage/nfs-ganesha
• BZ - 1472937 - RHGS-3.4, rebase (lib)ntirpc to 1.5.4
• BZ - 1472942 - RHGS-3.4, rebase nfs-ganesha to 2.5.5
• BZ - 1476559 - [Stress] : Input/Output Error while creating files(using touch) / bonnie++/dd during MTSH.
• BZ - 1476563 - [Stress] : Ganesha v4 mounts timed out during MTSH
• BZ - 1480947 - [Ganesha] : Ganesha crashed during service restarts.
• BZ - 1481040 - ganesha.nfsd processes steadily using high CPU on 2 nodes.
• BZ - 1489378 - [RFE]nfs-ganesha: Support xreaddirplus in FSAL_GLUSTER
• BZ - 1492695 - [Ganesha] : Ganesha crashed while exporting multiple volumes in loop.
• BZ - 1492995 - [Ganesha] : Ganesha crashed while bringing up the Ganesha service on crashed nodes.
• BZ - 1509118 - [Ganesha] : Ganesha crashed while exporting volumes in mdc_up_invalidate().
• BZ - 1514615 - Limit nfs ganesha server to support protocol versions to 3 and 4.0
• BZ - 1516699 - [Ganesha] : ganesha.log should be inside /var/log/ganesha.
• BZ - 1519713 - [Ganesha] : Ganesha nodes crash when I/O is started from clients , vmcore generated.
• BZ - 1520751 - [GSS]nfs-ganesha crashing after enabling ganesha.enable on a volume
• BZ - 1542569 - [Ganesha] : nfs-ganesha installs fail.
• BZ - 1546713 - [Ganesha] Default handling of the nfs version in ganesha.conf to limit the vers to 3 and 4.0
• BZ - 1550952 - service nfs-ganesha enable is consistently failing in RHGS 3.4
• BZ - 1551877 - [Ganesha] : NFS-Ganesha crashed during finds and ls in mdcache_new_entry.
• BZ - 1551878 - [Ganesha] : Ganesha logs are flooded with "Futility Count Exceeded" messages.
• BZ - 1551881 - [Ganesha] : Ganesha crashed during rm -rf from multiple clients.
• BZ - 1552082 - Mount count bug in nitirpc causes mounts to stop after a while
• BZ - 1559801 - [Ganesha] : logrotate cron task skips ganesha logs since /var/log/ganesha is world writable.
• BZ - 1560905 - [Ganesha] : Man page of Ganesha makes no mention of FSAL Gluster.
• BZ - 1562766 - [Ganesha] Ganesha got crashed while extracting Zip file content of disk fill utility on NFS Windows mount
• BZ - 1562774 - [Ganesha] Ganesha crashed while deleting diskfill utility content on windows nfs mount
• BZ - 1564481 - [Ganesha] ls -lrt command is stuck on NFS mount point for more than 1 hour having around 10 lakhs file
• BZ - 1571363 - [Ganesha] Extraction of files is failing on nfs share mounted on windows client
• BZ - 1575398 - [Ganesha] Lots of "NFS3 :CRIT :Likely bug: FSAL did not fill in a standard NFSv3 attribute: missing 1dfce" messages were observed in ganesha.log while deleting files from multi client
• BZ - 1577325 - [Ganesha] Ganesha process entered failed state while exporting/unexporting multiple volumes in loop
• BZ - 1580107 - [Ganesha] linux untar failed with "Permission denied" messages when find's were running in parallel (vers=3)
• BZ - 1608523 - Ganesha] Ganesha crashed in _mdcache_readdir_chunked_ while performing lookup,Rootsquash enable
• BZ - 1613273 - [Ganesha] Inconsistent output of du -sh on each iteration
• BZ - 1613890 - [Ganesha] Wrong file count is showing on Ganesha mount for some clients, post running du -sh command

CVEs

(none)

References

(none)