My Oracle Support Banner

Solaris Cluster ScalMountPoint Resources Mounting a ZFS Appliance may Fail with "Failed to get information for filer" (Doc ID 2264857.1)

Last updated on MAY 17, 2017

Applies to:

Solaris Cluster - Version 3.3 U2 to 4.3 [Release 3.3 to 4.3]
Oracle Solaris on SPARC (64-bit)
Oracle Solaris on x86-64 (64-bit)

Symptoms

When starting several Solaris Cluster ScalMountPoint resources in parallel that mount a ZFS Appliance via nfs, it may happen that some of these resources fail to start with the following diagnostic:

SC[SUNW.ScalMountPoint:4,<respurcegroup>,<resource>,/usr/cluster/lib/rgm/rrt/scal_mountpoint/scal_mountpoint_prenet_start]: [ID 571838 local7.error] Failed to get information for filer <filer>

This can happen when the ZFS Appliance is busy doing a lot of replication tasks.

When enabling ScalMountPoint resources one by one all resources come up fine with no error. e.g:

enable resource ; wait until it is reported online
enable next resource ; wait until it is reported online
...

 The problem was reproduced on the affected system for debug purpose using a test resource-group as follows:

# clrg create -S -p Nodelist=ora-zc2 -p Desired_primaries=1 -p Maximum_primaries=1 scal-test-rg
# clrs create -g scal-test-rg -t SUNW.ScalMountPoint:4 -p TargetFileSystem=NAS1:/export/ora_EXPORT -p FileSystemType=nas -p MountPointDir=/dat/export ora-export-test-rs
# clrs create -g scal-test-rg -t SUNW.ScalMountPoint:4 -p TargetFileSystem=NAS1:/export/ora_ORAREDO_B -p FileSystemType=nas -p MountPointDir=/dat/redob ora-redob-test-rs
# clrs create -g scal-test-rg -t SUNW.ScalMountPoint:4 -p TargetFileSystem=NAS1:/export/ora_ORAREDO_A -p FileSystemType=nas -p MountPointDir=/dat/redoa ora-redoa-test-rs
# clrs create -g scal-test-rg -t SUNW.ScalMountPoint:4 -p TargetFileSystem=NAS1:/export/ora_ORASYS -p FileSystemType=nas -p MountPointDir=/dat/orasys ora-orasys-test-rs
# clrs create -g scal-test-rg -t SUNW.ScalMountPoint:4 -p TargetFileSystem=NAS1:/export/ora_ORADATA -p FileSystemType=nas -p MountPointDir=/dat/oradata ora-oradata-test-rs
# clrs create -g scal-test-rg -t SUNW.ScalMountPoint:4 -p TargetFileSystem=NAS1:/export/ora_APP -p FileSystemType=nas -p MountPointDir=/app ora-app-test-rs
# clrs create -g scal-test-rg -t SUNW.ScalMountPoint:4 -p TargetFileSystem=NAS2:/export/ora_BCK -p FileSystemType=nas -p MountPointDir=/backup ora-backup-test-rs
# clrg online -M scal-test-rg

On starting this resourcegroup the issue was reproduced immediately:

...
Nov 10 19:23:43 ora-zc2 SC[SUNW.ScalMountPoint:4,scal-test-rg ,ora-redoa-test-rs,/usr/cluster/lib/rgm/rrt/scal_mountpoint/scal_mountpoint_prenet_start]: [ID 571838 local7.error] Failed to get information for filer "NAS1"
Nov 10 19:23:43 ora-zc2 SC[SUNW.ScalMountPoint:4,scal-test-rg ,ora-redoa-test-rs,/usr/cluster/lib/rgm/rrt/scal_mountpoint/scal_mountpoint_prenet_start]: [ID 571838 local7.error] Failed to get information for filer "NAS1"
Nov 10 19:23:43 ora-zc2 SC[SUNW.ScalMountPoint:4,scal-test-rg ,ora-redoa-test-rs,/usr/cluster/lib/rgm/rrt/scal_mountpoint/scal_mountpoint_prenet_start]: [ID 620193 local7.error] Validation failed.
Nov 10 19:23:43 ora-zc2 SC[SUNW.ScalMountPoint:4,scal-test-rg ,ora-redoa-test-rs,/usr/cluster/lib/rgm/rrt/scal_mountpoint/scal_mountpoint_prenet_start]: [ID 620193 local7.error] Validation failed.
...

With the exception of ora-backup-test-rs all other mounts failed:

clrs status -g scal-test-rg
=== Cluster Resources ===
Resource Name             Node Name    State          Status Message
-------------             ---------    -----          --------------
ora-backup-test-rs        ora-zc2      Online         Online
ora-app-test-rs           ora-zc2      Start_failed   Faulted
ora-oradata-test-rs       ora-zc2      Start_failed   Faulted
ora-orasys-test-rs        ora-zc2      Start_failed   Faulted
ora-redoa-test-rs         ora-zc2      Start_failed   Faulted
ora-redob-test-rs         ora-zc2      Start_failed   Faulted
ora-export-test-rs        ora-zc2      Start_failed   Faulted

Disabling and a subsequently re-enabling of all resources at once led to the same results

clrs disable -g scal-test-rg +
clrs enable -g scal-test-rg +

Further tests showed that enabling 2, 3, 4 resources in parallel also led to a failure of at least one of the resources

Cause

To view full details, sign in with your My Oracle Support account.

Don't have a My Oracle Support account? Click to get started!


My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.