VG not released

[nabob1]

It seems that the shared volume group between a two node cluster is not released whenever there is a network switch failure, I am suspecting the Failure Cycle or the Failure Detection Rate, does anyone have any ideas on this.

below is the last last two lines in the output of the HACMP.out file:

get_disk_vg_fs[178] grep -qx db001vg00 /tmp/lsvg.out.108546
get_disk_vg_fs[186] get_disk_vg_fs[186] cl_fs2disk -pg dat003vg00

Help - please

 

[PSD]

Hi,

When you say switch failure do you not have switch resilience? i.e. the service and standby connections spread over differing switches.

Is there a network down event logged in hacmp.out? and do you get the node with the failure bringing down the VG and swapping out the service for the boot address? Looking at the last two lines of the log it looks like the standby server is having a problem mouting the filesystems, is the vg varied on as this point on the standby server? Do you get config_too_long errors?

The failure detection rate is not applicable here as the VG has started to move, hence the failure has been detected.

Cheers


PSD
IBM Certified Specialist - AIX V4.3 Systems Support
IBM Certified Specialist - AIX V4 HACMP