we have a two node cluster on AIX (v184.108.40.206) where we use IBM.AgFileSystem for one VG. But we have overlooked the hints in Post-installation regarding shared volume groups. So the VG is not "enhanced concurrent capable". Now our AIX team wants to know before changing the VG which Flag TSAMP uses to mark the VG "concurrent/ Active" on the active site and "concurrent/ Passive" on the backup site or how TSAMP / IBM.StorageRM ensures that only one site could access the VG in comparsion to HACMP /PowerHA if the VG is marked as enhanced concurrent capable?
Thanks a lot.
NOTICE: developerWorks Community will be offline May 29-30, 2015 while we upgrade to the latest version of IBM Connections. For more information, read our upgrade FAQ.
This topic has been locked.
2 replies Latest Post - 2012-04-24T12:19:03Z by xf00657
Pinned topic shared volume groups enhanced concurrent capable
Answered question This question has been answered.
Unanswered question This question has not been answered yet.
Updated on 2012-04-24T12:19:03Z at 2012-04-24T12:19:03Z by xf00657
sedgewick_de 1000004PXP36 PostsACCEPTED ANSWER
Re: shared volume groups enhanced concurrent capable2012-04-20T11:02:19Z in response to xf00657Hi Rüdiger,
although we have answered this question outside of this forum, a public response here is beneficial.
There is a fundamental difference between an "enhanced concurrent capable" volume group and a "enhanced concurrent" volume group. Marking a volume group as "enhanced concurrent capable" skips establishing SCSI reservations for the disks before varying on the volume group. Beyond that no concurrency is involved, so the volume group can be varied on in the traditional non-concurrent mode. In fact, this is the only way cluster heartbeat devices based on volume groups or logical volumes can be handled.
In contrast a true enhanced concurrent volume group can be varied on on multiple nodes in read/write (active) or read/only (passive) mode. The corresponding line commands are varyonvg -c -n -A <my vg>, resp. varyonvg -c -n -P <my vg>, provided the clvm daemon responsible for keeping the LVM related parts of the ODM in sync over all participating nodes has been started. For experimental purposes this can be done while a TSA domain is online with startsrc -s gsclvmd -e "HA_DOMAIN_NAME=ApplDomain HA_GS_SUBSYS=cthags".
Passive mode provides some level of data protection, because it prevents accidentally mounting a filesystem and hence running a filesystem check.