Disk operation errors on active node while executing IBM's "getconf" command on disks from standby node in a two-node cluster on Hitachi Sun-branded 9990V array.
Article ID: 100003296
Updated On:
Description
Error Message
getconf: A system call received a parameter that is not valid.and, in "errpt" output:IDENTIFIER TIMESTAMP T C RESOURCE_NAME DESCRIPTIONB6267342 0609105910 P H hdisk50 DISK OPERATION ERROR...
Cause
The relevant Hitachi ODM definitions fileset compatible with DMP - 5.0.52.3 - was not installed.
Resolution
Since DMP was required for the Hitachi array, AIX's MPIO had to be disabled. The relevant Hitachi ODM definitions fileset found to be compatible with DMP - 5.0.52.3 - was then installed.
As a result, note the additional fields from "lsattr": "PR_key_value" and "reserve_policy"
#lsattr -EHl hdisk7attribute value description user_settablePR_key_value none Reserve Key Trueclr_q no Device CLEARS its Queue on error Truelocation Location Label Truelun_id 0x5000000000000 Logical Unit Number ID Falsemax_transfer 0x40000 Maximum TRANSFER Size Truenode_name 0x50060e8005655540 FC Node Name Falsepvid none Physical Volume ID Falseq_err yes Use QERR bit Falseq_type simple Queue TYPE Truequeue_depth 2 Queue DEPTH Truereassign_to 120 REASSIGN time out Truereserve_policy single_path Reserve Policy True
Once MPIO had been disabled and the above fileset installed, IBM's getconf command then completed without issue.
Applies To
OS : AIX 5300-05
VRTSvxvm : 5.0MP3RP2
Hitachi ASL : VRTS-USP 1.0.0.0 VERITAS Array Support Library TagmaStore-USP
Array: HITACHI Sun-branded HDS9990V
Issue/Introduction
Disk operation errors on active node while executing IBM's getconf command on disks from standby node in a two node cluster on Hitachi Sun-branded 9990V array.
Was this article helpful?
Yes
No
Powered by
