A new tool, /etc/vx/diag.d/vxcheckconn is introduced in SFCFSHA 6.0 to report clusterwide connectivity information of shared and local disks. The report could also be obtained from the local cached data maintained by CVM on all nodes.

The options provided by vxcheckconn tool is shown below:

node1# /etc/vx/diag.d/vxcheckconn -H
VxVM vxcheckconn ERROR V-5-1-693 Incorrect usage
Usage:
        ./vxcheckconn [-G][-H][-L][-M][-a][-c][-g ][-o][-v] ...       
        -G      Report global connectivity
        -H      Print detailed help message
        -L      Report local connectivity
        -M      Report disks only with connectivity issues
        -a      Report connectivity for all disks (not just shared only)
        -c      Report cached data
        -g Report connectivity for disks in specified DG.
        -o      Report connectivity for a vxvm plex/vol object.
        -s      Return site connectivity info
        -v      Report verbose data
node1#

*** Two shared disks are configured on a two node CVM cluster

node1# vxdisk -o alldgs list | grep shared
emc_clariion0_1 auto:cdsdisk    emc_clariion0_1  sharedg          online shared
emc_clariion0_2 auto:cdsdisk    emc_clariion0_2  sharedg          online shared
node1#

*** Use -L option to check the connectivity information on local system

node1# /etc/vx/diag.d/vxcheckconn -L -v
daname:         emc_clariion0_1
request:        act=0x2 dflag=0x14000000 kflag=0 seq=0 age=0
connmap:        present
connected:      localhost
notconnected:  

daname:         emc_clariion0_2
request:        act=0x2 dflag=0x14000000 kflag=0 seq=0 age=0
connmap:        present
connected:      localhost
notconnected:
node1#

*** Use -G option to check the clusterwide connectivity information

node1# /etc/vx/diag.d/vxcheckconn -G -v
daname:         emc_clariion0_1
request:        act=0x4 dflag=0x14000000 kflag=0 seq=2 age=1131818009
connmap:        0000000000000003
connected:      node2 node1
notconnected:
   

daname:         emc_clariion0_2
request:        act=0x4 dflag=0x14000000 kflag=0 seq=2 age=1131818010
connmap:        0000000000000003
connected:      node2 node1
notconnected:
node1#

*** Use -M option to get the information on disks that have connectivity issues
*** In below case, none of the shared disks have connectivity issues and hence no ouput

node1# /etc/vx/diag.d/vxcheckconn -M -v
node1#

*** Disabling paths for one dmpnode to get more accurate missing

node1# vxdmpadm getsubpaths dmpnodename=emc_clariion0_2
NAME         STATE[A]   PATH-TYPE[M] CTLR-NAME  ENCLR-TYPE   ENCLR-NAME    ATTRS
================================================================================
sdc          ENABLED(A)  PRIMARY      c3         EMC_CLARiiON  emc_clariion0     -
node1#
  node1# vxdmpadm -f disable path=sdc
node1#

 

node1# vxdisk -o alldgs list |grep shared
emc_clariion0_1 auto:cdsdisk    emc_clariion0_1  sharedg          online shared
emc_clariion0_2 auto:cdsdisk    emc_clariion0_2  sharedg          online shared lfailed
node1#

*** -M option now shows that the disk is not connected on node1

node1# /etc/vx/diag.d/vxcheckconn -M -v
daname:         emc_clariion0_2
request:        act=0x5 dflag=0x14000000 kflag=0x100 seq=3 age=1132638265
connmap:        0000000000000001
connected:      node2
notconnected:   node1
node1#

*** This will also be reflected in -G and -L options

node1# /etc/vx/diag.d/vxcheckconn -G -v
[...]
    daname:         emc_clariion0_2
request:        act=0x4 dflag=0x14000000 kflag=0x100 seq=3 age=1132638265
connmap:        0000000000000001
connected:      node2
notconnected:   node1
node1#

*** -c option could also be used to get this status from cached data

node1# /etc/vx/diag.d/vxcheckconn -cL -v
daname:         emc_clariion0_1
request:        act=0x3 dflag=0x14000000 kflag=0 seq=0 age=0
connmap:        present
connected:      localhost
notconnected:

  daname:         emc_clariion0_2
request:        act=0x3 dflag=0x14000000 kflag=0x100 seq=0 age=0
connmap:        missing
connected:
notconnected:   localhost
node1#

 

Applies To

- All UNIX and Linux platforms

- SFCFSHA 6.0

 

It may be time consuming to identify missing disks on a Cluster Volume Manager (CVM) cluster with large number of disks connected to the nodes. A new tool vxcheckconn located at /etc/vx/diag.d directory is introduced in Storage Foundation Cluster File System High Availability (SFCFSHA) 6.0 that provides report of local and shared disk connectivity information across a CVM cluster.