This county1 host is a HACMP cluster and is a mirror of county2 server and my colleague said HACMP is not active rightnow, And SRDF is running on county2 continuously like a loop process ..we have a script provided by the EMC for SRDF , so in the log file ( /var/symapi/log/ symapi-20071228.log ) I seen the processes of synchronize , split , establish , suspend , synchronize and split Are running like loop process and I run symmir command for county2 device group I found the state of devices is keep on changing like syncinprog, synchronize , split like that from time to time with in span of minutes so is this is a reason for emc_cfgmgr saying hdisks are in undetermined state on county1 ..I assume ¿
server1/root: / > symmir -g county2_dg query Device Group (DG) Name: county2_dg DG's Type : REGULAR DG's Symmetrix ID : XXXXX
If you are attempting to mount the BCV's at the R2 side you should first of all stop your replication script after a complete cycle.
Dynamox is right you should run the cfg mgr when the BCV's are in a split state.
If the emc_cfgmgr is running during the middle of a cycle the device state will be changing, the BCV will become not ready when it is in sync in progress and will change to RW when it is split. This could be confusing for the emc_cfgmgr script if the replication process is running.
Before stopping your script you should also ensure you know how to restart it and what state everything needs to be in before kicking off.
dynamox
9 Legend
•
20.4K Posts
0
December 28th, 2007 12:00
pctech321
1 Rookie
•
92 Posts
0
December 28th, 2007 13:00
server1/root: / > symmir -g county2_dg query
Device Group (DG) Name: county2_dg
DG's Type : REGULAR
DG's Symmetrix ID : XXXXX
Standard Device BCV Device State
-------------------------- ------------------------------------- ------------
Inv. Inv.
Logical Sym Tracks Logical Sym Tracks STD <=> BCV
-------------------------- ------------------------------------- ------------
----------snipped------
DEV023 024C 0 BCV023 0353 * 0 Synchronized
DEV024 024D 0 BCV024 0354 * 0 Synchronized
DEV025 067D 0 BCV025 0313 * 1 SyncInProg
DEV026 067E 0 BCV026 0314 * 0 Synchronized
DEV027 067F 0 BCV027 0315 * 0 Synchronized
DEV028 0749 0 BCV028 02DB * 0 Synchronized
DEV029 074A 0 BCV029 02DC * 0 Synchronized
DEV030 074B 0 BCV030 02DD * 0 Synchronized
-------------------snipped---------
Total ------- -------
Track(s) 0 4
MB(s) 0.0 0.1
Legend:
(*): The paired BCV device is associated with this group.
AFTER few SEC¿S Here is the state of devices
Server1/root: / > symmir -g county2_dg query
Device Group (DG) Name: county2_dg
DG's Type : REGULAR
DG's Symmetrix ID :XXXX
Standard Device BCV Device State
-------------------------- ------------------------------------- ------------
Inv. Inv.
Logical Sym Tracks Logical Sym Tracks STD <=> BCV
-------------------------- ------------------------------------- ------------
DEV023 024C 0 BCV023 0353 * 0 Split
DEV024 024D 0 BCV024 0354 * 0 Split
DEV025 067D 0 BCV025 0313 * 0 Split
DEV026 067E 0 BCV026 0314 * 0 Split
DEV027 067F 0 BCV027 0315 * 0 Split
DEV028 0749 0 BCV028 02DB * 0 Split
DEV029 074A 0 BCV029 02DC * 0 Split
DEV030 074B 0 BCV030 02DD * 0 Split
dynamox
9 Legend
•
20.4K Posts
0
December 28th, 2007 14:00
rawstorage
3 Apprentice
•
423 Posts
0
December 28th, 2007 23:00
Dynamox is right you should run the cfg mgr when the BCV's are in a split state.
If the emc_cfgmgr is running during the middle of a cycle the device state will be changing, the BCV will become not ready when it is in sync in progress and will change to RW when it is split. This could be confusing for the emc_cfgmgr script if the replication process is running.
Before stopping your script you should also ensure you know how to restart it and what state everything needs to be in before kicking off.