RHBA-2010:0280 - Bug Fix Advisory

Topic

Updated rgmanager packages that fix numerous bugs and add several
enhancements are now available.

Description

The rgmanager packages contain the Red Hat Resource Group Manager, which
provides the ability to create and manage high-availability server
applications in the event of system downtime.

This update addresses the following bugs:

• failed virtual machine migrations are handled more correctly. (BZ#499835)
• clustered file systems are no longer unmounted after a configuration

change. (BZ#506094)

• clustat's output is now consistent when given bad input. (BZ#506346)
• the force_unmount flag now works correctly when used with HA-LVM.

(BZ#514040)

• path support for Xen virtual machines works again. (BZ#519786)
• S/Lang processor no longer leaks memory. (BZ#507431)
• rgmanager exits if killed while waiting for fencing. (BZ#508147)
• bonded link handling when using Xen bridged interfaces works. (BZ#518037)
• services no longer get stuck in the 'recovering' state. (BZ#530409)
• node rejoins prior to fencing completion are now handled correctly.

(BZ#527777)

• live migration of frozen VMs is an invalid operation. (BZ#510017)
• tomcast-5.sh now uses su instead of sudo. (BZ#524757)
• multiple event handlers no longer cause problems with central_processing.

(BZ#527239)

• ip.sh honors monitor_link during service start. (BZ#532756)
• ip.sh now attempts to detect IP address collisions. (BZ#526647)
• ip.sh now handles ipv6 addresses with shortened quads correctly.

(BZ#533461)

• SAPInstance and SAPDatabase may now be referenced from the <resources/>

block. (BZ#529052)

• fs.sh no longer warns about bind mounts. (BZ#526286)
• migrating a virtual machine to an offline node returns an error.

(BZ#536157)

• aggressive timeouts causing services to sometimes erroneously enter the

'failed' state have been changed to scale with the cluster's timeout.
(BZ#548133)

• clusvcadm now checks the return code msg_send. (BZ#529929)
• 'Depend' attributes when using central processing are handled correctly.

(BZ#523999)

• a bug in HA-LVM which could cause metadata corruption in some cases has

been fixed. (BZ#557167)

This update also includes the following enhancements:

• more debugging information has been added to /tmp/rgmanager-dump..

(BZ#512052)

• path support emulation has been added for KVM virtual machines.

(BZ#545916)

• vm.sh now provides more meaningful error reports. (BZ#529926)
• file system resources now report errors when status checks fail.

(BZ#562237)

Red Hat Resource Group Manager users are advised to upgrade to these
updated packages, which address these issues.

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
http://kbase.redhat.com/faq/docs/DOC-11259

Affected Products

• Red Hat Enterprise Linux High Availability for x86_64 5 x86_64
• Red Hat Enterprise Linux High Availability for x86_64 5 ppc
• Red Hat Enterprise Linux High Availability for x86_64 5 ia64
• Red Hat Enterprise Linux High Availability for x86_64 5 i386
• Red Hat Enterprise Linux High Availability (for RHEL Server) from RHUI 5 x86_64
• Red Hat Enterprise Linux High Availability (for RHEL Server) from RHUI 5 i386

Fixes

• BZ - 499835 - detect failed virtual machine migrations in a cluster
• BZ - 506094 - CRM #1919318, GFS mount (shared resources) are unmounted when new service is added using them
• BZ - 506346 - clustat -s <service-name> displays redundant row/column headers when the service-name is non-existing.
• BZ - 507431 - S/Lang stack / memory leaks
• BZ - 508147 - rgmanager fails stop after a failed initialization.
• BZ - 510017 - When freezing virtual service still able to live migrate!
• BZ - 514040 - The force unmount option is not working properly in an HA LVM config.
• BZ - 518037 - IP resources don't work with xen-bridged bonded vlan interfaces
• BZ - 519786 - vm.sh ignores path / drops xm --path support
• BZ - 523999 - "depend" attributes not interacting correctly with exclusive prioritization
• BZ - 524757 - Rgmanager uses sudo to start/manage tomcat5 service which fails as no tty is available.
• BZ - 526286 - Cluster FS resource doesn't ignore bind mounts during status check
• BZ - 526647 - Cluster IP resource does not deal with IP address collision possibility
• BZ - 527239 - clurgmgrd with central processing enabled dies
• BZ - 527777 - if fencing is averted, rgmanager blocks on node transition
• BZ - 529052 - SAP resources does not have a primary attribute associated with resource and shared resource are not seen in service
• BZ - 529926 - vm.sh error reporting
• BZ - 529929 - clusvcadm does not check msg_send return code
• BZ - 532756 - VIP parameters 'monitor_link' not apply
• BZ - 533461 - ip.sh does not parse ipv6 addresses correctly if quad contains less than 4 digits
• BZ - 545916 - Add virsh path support emulation to vm.sh
• BZ - 548133 - rgmanager - Failed changing service status

CVEs

(none)

References

(none)