RHBA-2015:2530 - Bug Fix Advisory

Topic

Updated vdsm packages that fix several bugs and add various enhancements are now
available.

Description

VDSM is a management module that serves as a Red Hat Enterprise Virtualization
Manager agent on Red Hat Enterprise Virtualization Hypervisor or Red Hat
Enterprise Linux hosts.

Changes to the vdsm component:

• Previously, upgrading a data center with compatibility version of 3.0 (with a

V1 master domain) when it's on maintenance to a compatibility version of 3.5
caused the SPM start to fail. The SPM start in the 3.5 host attempted to inquire
the cluster lock and since it is not supported, the SPM failed to start. With
this update, SPM start won't fail even if the cluster lock inquiring is not
supported. (BZ#1269411)

• Previously, VDSM uses a high number of system threads, and caused high

resource usage making the system slow. With this update, a new option,
cpu_affinity, in vdsm.conf was added so users can tune the CPU affinity. The
option is disabled by default.
To enable the new option, edit the vdsm.conf file. Under [vars] section, edit
the 'cpu_affinity' option. It accepts a comma separated whitelist of CPU cores
on which VDSM is allowed to run. The default is "", meaning VDSM can be
scheduled by the operating system to run on any core. Some valid examples
include: "1", "0,1", "0,2,3".
The resource usage of VDSM dramatically improves when the cpu_affinity is
enabled. It is recommended to enable the cpu_affinity option if VDSM uses too
much CPU. It is also recommended to set the affinity to one CPU only and to
avoid cpu #0, because other system task may default to that CPU. (BZ#1265205)

• When VDSM was ported to systemd for Red Hat Enterprise Linux 7, it stopped

generating its pid file which was required for safelease. As a result, when
safelease failed to maintain the SPM lease, VDSM didn't get terminated cleanly
and rebooted the SPM host, killing virtual machines running on the host. With
this update, VDSM passes its pid to safelease via a command line parameter. If
safelease fails to maintain the SPM lease, VDSM is terminated in a clean way
without rebooting the machine, and the SPM role is assigned to another host
without effecting virtual machines running on the SPM host. (BZ#1265177)

• Previously, the multipath device configuration was overwritten by the iSCSI

daemon using the default timeout 120 seconds, and resulted in long delays before
VDSM attempts to try the next path. With this update, using an updated kernel,
the iSCSI daemon now uses the multipath device timeout instead of overwriting it
with a longer timeout time. There is now minimal delay during failover.
(BZ#1273421)

• Previously, when starting a configured but non-active VDSM device, for example

if a device was manually set to down, the VDSM service executed the 'service
network start' command instead of the 'service network restart' command. If the
device was also running DHCP, the dhclient was started twice and led to VDSM
service start failure. With this update, 'service network restart' is actually
run so if DHCP was running on the device, the dhclient it not started a second
time, and the VDSM service starts as normal. (BZ#1267500)

Solution

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

This update is available via the Red Hat Network. Details on how to
use the Red Hat Network to apply this update are available at
https://access.redhat.com/articles/11258

Affected Products

• Red Hat Virtualization 3 for RHEL 7 x86_64
• Red Hat Virtualization 3 for RHEL 6 x86_64

Fixes

• BZ - 1243710 - vdsm log is flooded with JsonRpcServer and stompReactor messages
• BZ - 1261507 - KSM sleep_millisecs bellow 10ms for systems above 16GB of RAM
• BZ - 1263979 - Vdsm should recover ifcfg files in case they are no longer exist and recover all networks on the server
• BZ - 1265177 - regression for EL7: spmprotect always reboot when fencing vdsm on systemd
• BZ - 1265205 - [scale] high vdsm threads overhead
• BZ - 1265907 - zstream clone: Extend of VG does not check if additional devices are already part of it
• BZ - 1265965 - zstream clone: OSError: [Errno 24] Too many open files while running automation tests
• BZ - 1267500 - vdsm fails to start if dhclient is running
• BZ - 1269187 - [vdsm] Domain deactivation doesn't clear domain from the domains to be upgraded list
• BZ - 1269411 - [z-stream clone 3.5.6] [vdsm] SpmStart fails after "offline" upgrade of DC with V1 master domain to 3.5
• BZ - 1273421 - Need to add deps on kernel] vdsm iscsi failover taking too long during controller maintenance
• BZ - 1275637 - Consume fix for "Multipath is not correctly identifying iscsi devices, and misconfiguring them"

CVEs

(none)

References

(none)