Disk path missing on virtual adapter across DWDM
Applies to
- Metrocluster IP
- AFF A90
- ADVA DWDM
Issue
- Remote disks in the metrocluster show as having only 7 paths available instead of the expected 8
- EMS log shows a iSCSI session disconnect and reconnect
[?] Sat Aug 01 02:01:45 -0500 [node-1: kernel: iscsi.session.stateChanged:notice]: iSCSI session state is changed to Reconnecting for the target iqn.2016-06.com.netapp:123456789-abcdefg-98875aqbfasd (type: dr_auxiliary, address: 1.2.3.4:12345). Reason: no ping reply after 5 seconds. [?] Sun Aug 09 02:01:45 -0500 [node-1: doneq0: scsi.mcc.adt.ioTransportError:error]: mcc_adt[0] - Transport error during execution of command: HA status 0x13: CAM transport status 0x1b : cdb 0x28:115c1a40:0040.
[?] Sat Aug 01 02:01:45 -0500 [node-1: scsi_cmdblk_strthr_admin: scsi.cmd.abortedByHost:error]: Disk device 0m.i2.0L82: Command aborted by host adapter: HA status 0x13: cdb 0x28:115c1a40:0040.
[?] Sat Aug 01 02:01:45 -0500 [node-1: scsi_cmdblk_strthr_admin: scsi.cmd.retrySuccess:debug]: Disk device 0v.i2.3L82: request successful after retry #1/#0: cdb 0x9a:000000000aaa7e55:0001:0003 (8380).
[?] Sat Aug 01 02:02:44 -0500 [node-1: iscsid: iscsi.session.stateChanged:notice]: iSCSI session state is changed to Connected for the target iqn.2016-06.com.netapp:123456789-abcdefg-98875aqbfasd (type: dr_auxiliary, address: 1.2.3.4:12345).
- Disk redundancy failed message follows when a path is lost
[?] Sat Aug 01 02:12:28 -0500 [node-1: svc_queue_thread: callhome.dsk.redun.fault:error]: Call home for DISK REDUNDANCY FAILED
- The lost path is only seen on a single node, this node shows Multi-Path HA instead of Quad-Path HA like the others in
sysconfig -aoutput
