Error Message

vxvm:vxconfigd: V-5-1-16253 Disk group import of sfcache_defaultdg failed with error 150 - No valid disk found containing disk group

Supported Configurations

• Storage Foundation Standard (Read Only)
• Storage Foundation Enterprise
• Storage Foundation High Availability
• Storage Foundation Cluster File System HA

Supported Operating Systems

• Red Hat Enterprise Linux 5, 6
• SUSE Linux Enterprise Server 11
• Oracle Enterprise Linux 5, 6

Even when the SmartIO functionality is not used and the "sfcache_defaultdg" disk group does not exist, the "sfcache list" command still attempts to import the "sfcache_defaultdg" disk group.

# vxdg list
NAME         STATE           ID
iodrive2     enabled,cds          1392149745.20.smartioserver


# sfcache list
NAME                             TYPE   SIZE       ASSOC-TYPE  STATE      DEVICE

The above command triggers the vxconfigd message in relation to the non-existent "sfcache_defaultdg" diskgroup.
 

# tail -f /var/log/messages

Sep 29 01:48:14 smartioserver vxvm:vxconfigd: V-5-1-16253 Disk group import of sfcache_defaultdg failed with error 150 - No valid disk found containing disk group
Sep 29 01:48:15 smartioserver vxvm:vxconfigd: V-5-1-16253 Disk group import of sfcache_defaultdg failed with error 150 - No valid disk found containing disk group


The sfcache.log file records the import event trying to import the "sfcache_defaultdg" disk group.


# tail -18 /etc/vx/log/sfcache.log

#########################################################
Mon Sep 29 01:48:14 PDT 2014
sfcache list
#########################################################
/usr/sbin/vxdg -q list
/usr/sbin/vxprint -o alldgs -q -v -e 'v_cachearea_vm=on'
/usr/sbin/vxdg import sfcache_defaultdg
VxVM vxdg ERROR V-5-1-10978 Disk group sfcache_defaultdg: import failed:
No valid disk found containing disk group
/usr/sbin/vxdg list
/usr/sbin/vxdg list
/usr/sbin/vxprint -o alldgs -q -v -e 'v_cachearea_fs=on'
/usr/sbin/vxdg import sfcache_defaultdg
VxVM vxdg ERROR V-5-1-10978 Disk group sfcache_defaultdg: import failed:
No valid disk found containing disk group
/usr/sbin/vxdg list
/usr/sbin/vxdg list
 

NOTE: Without the code fix, the "sfcache list" internally tries to import the default cachedg "sfcache_defaultdg", if it does not exist or is deported.

 

VOM

VOM also executes the "sfcache list" command at 30 minute intervals to monitor SmartIO resources. As the "sfcache_defaultdg" does not exist, the vxconfigd disk group error messages are reported.
 
Sample files:

 

 

/vom/var/opt/VRTSsfmh/scheduler.conf

#
# xprtl scheduler configuration
#
ping_sfm_vom1.srv.hcvlny.cv.net kickoff|function 5 - - - - execute_ping sfm://vom1.srv.hcvlny.cv.net:5634/
discovery_S10-HOST kickoff 1440 - - - - /opt/VRTSsfmh/bin/perl /opt/VRTSsfmh/bin/mh_driver.pl --family HOST --hidden --id discovery_S10-HOST
discovery_S20-SF : 30 - - - - /opt/VRTSsfmh/bin/perl /opt/VRTSsfmh/bin/mh_driver.pl --family SF --hidden --id discovery_S20-SF  <<<<< Runs every 30 mins!
discovery_S22-NATIVE : 360 - - - - /opt/VRTSsfmh/bin/perl /opt/VRTSsfmh/bin/mh_driver.pl --family NATIVE --hidden --id discovery_S22-NATIVE
discovery_S25-DB : 360 - - - - /opt/VRTSsfmh/bin/perl /opt/VRTSsfmh/bin/mh_driver.pl --family DB --hidden --id discovery_S25-DB
discovery_S30-VCS : 60 - - - - /opt/VRTSsfmh/bin/perl /opt/VRTSsfmh/bin/mh_driver.pl --family VCS --hidden --id discovery_S30-VCS
discovery_S50-NR : 5 - - - - /opt/VRTSsfmh/bin/perl /opt/VRTSsfmh/bin/mh_driver.pl --family NR --hidden --id discovery_S50-NR
discovery_S60-LDR : 1440 - - - - /opt/VRTSsfmh/bin/perl /opt/VRTSsfmh/bin/mh_driver.pl --family LDR --hidden --id discovery_S60-LDR
discovery_S200-VVRBW : 60 - - - - /opt/VRTSsfmh/bin/perl /opt/VRTSsfmh/bin/mh_driver.pl --family VVRBW --hidden --id discovery_S200-VVRBW
discovery_S300-PCV_NOTIFY : 30 - - - - /opt/VRTSsfmh/bin/perl /opt/VRTSsfmh/bin/mh_driver.pl --family PCV_NOTIFY --hidden --id discovery_S300-PCV_NOTIFY
discovery_S85-KVM : 120 - - - - /opt/VRTSsfmh/bin/perl /opt/VRTSsfmh/bin/mh_driver.pl --family KVM --hidden --id discovery_S85-KVM
discovery_S150-PERF : */5 * * * * /opt/VRTSsfmh/bin/perl /opt/VRTSsfmh/bin/mh_driver.pl --family PERF --hidden --id discovery_S150-PERF
sfmh-discovery kickoff 1440 - - - - /opt/VRTSsfmh/adm/vxvmdiscovery-ctrl.sh start
stats-sweeper : 1 1 1,15 * * /opt/VRTSsfmh/bin/perl /opt/VRTSsfmh/adm/stats-sweeper.pl
 

/vom/var/opt/VRTSsfmh/SF/log/mh.log*

 

# grep sfcache *
mh.log:2014-09-02 02:16:26 [debug] 19843 S60Cache: Invoking sfcache list cli
mh.log:2014-09-02 02:16:27 [debug] 19843 S60Cache: Invoking vxprint -g sfcache_defaultdg cli
mh.log:2014-09-02 02:16:27 [debug] 19843 S60Cache: Invoking cli sfcache list -l
mh.log:2014-09-02 02:46:19 [debug] 28829 S60Cache: Invoking sfcache list cli
mh.log:2014-09-02 02:46:19 [debug] 28829 S60Cache: Invoking vxprint -g sfcache_defaultdg cli
mh.log:2014-09-02 02:46:19 [debug] 28829 S60Cache: Invoking cli sfcache list -l
mh.log:2014-09-02 03:16:30 [debug] 5754 S60Cache: Invoking sfcache list cli
mh.log:2014-09-02 03:16:30 [debug] 5754 S60Cache: Invoking vxprint -g sfcache_defaultdg cli
mh.log:2014-09-02 03:16:30 [debug] 5754 S60Cache: Invoking cli sfcache list –l  

NOTE: Impacted VOM perl scripts - S10vm.pm and S60Cache.pm.

To fix this issue, the "sfcache_defaultdg" disk group will no longer be imported by the "sfcache list" command.

The "sfcache_defaultdg" disk group import will be handled by vxattachd daemon.

Whenever any disk attach event occurrs, it is checked if the disk is associated with a deported disk group. If the associated disk group is in a deported state and the disk group matches with the cache disk group entry in the /etc/vx/volboot file or default cache dg (sfcache_defaultdg), then that disk group will be imported by vxattachd.


Sample volboot file
 

#  more /etc/vx/volboot
volboot 3.1 0.6 130
hostid smartioserver
hostguid {8ab7087a-47ae-11e4-ae61-23566e0cc001}
request_threads 2
Command_Shipping 1
vm_cacheareadg sfcache_defaultdg
end
###############################################################
###############################################################
###############################################################
###############################################################
###############################################################
############################

 

Please contact Veritas Support if you require VxVM hot-fix 6.1.1.001 for Linux platforms. The fix is not included in the public patch 6.1.1.100.
To support RedHat 6.6 deployments, VxVM hot-fix 6.1.1.101 is now required.

The VxVM private hot-fix will be made available in the next public patch for 6.1.1.x on Linux platforms only.

Storage Foundation (SF) 6.2 now includes the fix in the base release and includes the fix for other platforms.

 

Applies To

The "sfcache_list" CLI command attempts to import the "sfcache_defaultdg" disk group even when SmartIO functionality has NOT been configured. This results in unwanted vxconfigd messages being generated. The 6.1 release of Storage Foundation High Availability introduces SmartIO which enables customers to increase application performance and reduce storage costs while maintaining availability of data and applications. SmartIO takes advantage of internal Solid State Storage (SSD) devices to provide an intelligent, distributed caching layer to serve more reads and writes from inside the host, effectively decoupling front-end IOPs (Input/Output Operations Per Second, pronounced eye-ops) from back-end capacity.

For more details, please refer to: https://vox.veritas.com/t5/Information-Availability-Blog/SmartIO-in-Storage-Foundation-HA-6-1/ba-p/779685

ETrack: 3561391