RHBA-2019:0263 - Bug Fix Advisory

Topic

Updated glusterfs packages that fix several bugs are now available for Red Hat Gluster Storage 3.4 Batch Update 3 on Red Hat Enterprise Linux 7.

Description

Red Hat Gluster Storage is software only scale-out storage solution that provides flexible and affordable unstructured data storage. It unifies data storage and infrastructure, increases performance, and improves availability and manageability to meet enterprise-level storage challenges.

This advisory fixes the following bugs:

• Earlier, a race condition between `pcs cluster setup ...` and `systemctl enable pacemaker` sometimes left Pacemaker disabled. As a result, Pacemaker failed to start on one or more nodes after reboot. With this release, the customer can add the '--enable' option to the `pcs cluster setup ...` command to enable and start Pacemaker automatically after reboot. (BZ#1637564)
• Previously, during a lookup failure, the stat structure returned was NULL, dereferencing which lead to the crash. With this fix, there is no defer in the stat structure when lookup failed. (BZ#1664529)
• This update introduces a feature, reader-thread-count, for multi-threaded FUSE reader threads, which imports parallel requests on a FUSE mount and handles them through multiple threads, proffering better I/O performance. (BZ#1651040)
• With this release, while setting up geo-replication, the slave volumes are made read-only by default. The intention is to write to the master and have the slave volumes synced to it. Considering write should not happen from the slave, it is competent to make the slave volumes read-only. (BZ#1643370)
• Earlier, the glusterd service did not regenerate volfiles during initialization when there was a change in the maximum operating version supported. This resulted in bricks running with an older volfile in Red Hat OpenShift Container Storage. With this fix, glusterd will regenerate the volfiles whenever there is a variation in the maximum (max-op-version) supported operating version. (BZ#1651460)
• Earlier, enabling group 'virt' failed on non-replicate volumes due to the existing replicate-specific options in the group. As a result, the options that could optimize the volume for virtual machine storage was not set on distributed volumes. This release introduces a new group 'distributed-virt' to optimize the deployment of distributed volumes in the virtual machine storage. (BZ#1653613)
• With this release, the 'gfid2path' option is enabled by default. This provides the path for the file when the GFID is known thus helping administrators to know which file has split-brain. (BZ#1644318)
• In replica 3 volume, renaming a file while the brick with the 'good copy' of the file is down, results in the removal of the file during self-heal leading to data loss. With this release, lookup of a file fails if there is no good copy (as determined by AFR xattrs) found, dismissing the rename option and limiting the data loss. (BZ#1362129)
• Earlier, Samba hook scripts removed the share section from the smb.conf after restarting a volume. Any additional share specific parameters set within the configuration file was lost. In this update a new parameter, available = no, is added to sections making them unavailable to clients. Any share section with this parameter is not accessible when a volume is stopped. Restarting the volume will remove the parameter, making it accessible again. (BZ#1164778)
• This release introduces a new group option for configuring volume options for Samba-CTDB setup which validates the requirement of volume options ideal for the setup. (BZ#1655385)
• With this release, the Samba hook script has been improved for adding volume share section in the smb.conf file. To export gluster volume as SMB share via Samba, one of the following volume options, user.cifs or user.smb is required. The existing glusterfs volume share sections in smb.conf remains unchanged. (BZ#1541568)

Users of Red Hat Gluster Storage are advised to upgrade to these updated packages, which fixes these bugs.

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 Enterprise Linux Server 7 x86_64
• Red Hat Virtualization 4 for RHEL 7 x86_64
• Red Hat Gluster Storage Server for On-premise 3 for RHEL 7 x86_64

Fixes

• BZ - 1350745 - build: remove absolute paths from glusterfs spec file
• BZ - 1362129 - rename of a file can cause data loss in an replica/arbiter volume configuration
• BZ - 1541568 - [GSS] logic in S30samba-start.sh hook script needs tweaking
• BZ - 1596040 - eagerlock seems to lose effect when the file being written is also being read
• BZ - 1597252 - EC volume getting created without any redundant brick
• BZ - 1599220 - glusterd crashed and core generated at gd_mgmt_v3_unlock_timer_cbk after huge number of volumes were created
• BZ - 1608507 - glusterd segfault - memcpy () at /usr/include/bits/string3.h:51
• BZ - 1619857 - glusterd signal 8 - dict_lookup_common
• BZ - 1631851 - Why do all of the gluster logs contain the exact same messages and so many of them?
• BZ - 1633177 - gluster-NFS crash while expanding volume
• BZ - 1637379 - Portmap entries showing stale brick entries when bricks are down
• BZ - 1637445 - Brick process not coming up after force starting volume due to stale port map entries
• BZ - 1637564 - Customer reports regression of BZ 1441055 with NFS-Ganesha in gluster 3.8.4-54
• BZ - 1639476 - glusterd segmentation fault: glusterd_op_ac_brick_op_failed (event=0x7f44e0e63f40, ctx=0x0) at glusterd-op-sm.c:5606
• BZ - 1639568 - when peer detach is issued, throw a warning to remount volumes using other cluster IPs before proceeding
• BZ - 1643370 - [geo-rep]: RFE - Make slave volume read-only while setting up geo-rep (by default)
• BZ - 1645480 - Files pending heal in Arbiter volume
• BZ - 1648296 - dht_revalidate may not heal attrs on the brick root
• BZ - 1648893 - statedump doesn't contain information about newer mempools
• BZ - 1651040 - Add support for multi-threaded fuse reader threads for better performance
• BZ - 1651460 - glusterd can't regenerate volfiles in container storage upgrade workflow
• BZ - 1652466 - [Glusterd]: Glusterd crash while expanding volumes using heketi
• BZ - 1652537 - bump up default value of server.event-threads
• BZ - 1653224 - do not package crypt.so
• BZ - 1653613 - [Dalton] Optimize for virt store fails with distribute volume type
• BZ - 1654103 - Invalid memory read after freed in dht_rmdir_readdirp_cbk
• BZ - 1654161 - glusterd crashed with seg fault possibly during node reboot while volume creates and deletes were happening
• BZ - 1655385 - [Samba-Enhancement] Need for a single group command for setting up volume options for samba
• BZ - 1655578 - Incorrect usage of local->fd in afr_open_ftruncate_cbk
• BZ - 1656357 - missing explicit package dependencies in spec file
• BZ - 1659439 - Memory leak: dict_t leak in rda_opendir
• BZ - 1662059 - [RHV-RHGS] Fuse mount crashed while deleting a 1 TB image file from RHV
• BZ - 1662828 - Longevity: glusterfsd(brick process) crashed when we do volume creates and deletes
• BZ - 1663232 - profile info command is not displaying information of bricks which are hosted on peers
• BZ - 1664235 - [geo-rep]: Directory renames not synced to slave in Hybrid Crawl
• BZ - 1664529 - [geo-rep]: Multiple crashes seen on the slave during automation run
• BZ - 1667169 - glusterd leaks about 1GB memory per day on single machine of storage pool
• BZ - 1668304 - [RHHI-V] Fuse mount crashed while creating the preallocated image

CVEs

(none)

References

(none)