Issue

• When I fail all paths to one of my multipath devices for which there is an fs resource in a running cluster service, it takes almost 3 minutes before it picks up the failure

Aug 13 11:57:49 node1 kernel: connection3:0: ping timeout of 5 secs expired, recv timeout 5, last rx 4296127569, last ping 4296132569, now 4296137569
Aug 13 11:57:49 node1 kernel: connection3:0: detected conn error (1011)
Aug 13 11:57:50 node1 iscsid: Kernel reported iSCSI connection 3:0 error (1011 - ISCSI_ERR_CONN_FAILED: iSCSI connection failed) state (3)
Aug 13 11:57:54 node1 kernel: session2: session recovery timed out after 5 secs
Aug 13 11:57:54 node1 kernel: sd 9:0:0:0: [sdd] Unhandled error code
Aug 13 11:57:54 node1 kernel: sd 9:0:0:0: [sdd] Result: hostbyte=DID_TRANSPORT_FAILFAST driverbyte=DRIVER_OK
Aug 13 11:57:54 node1 kernel: sd 9:0:0:0: [sdd] Unhandled error code
[...]
Aug 13 12:00:27 node1 kernel: Aborting journal on device dm-2-8.
Aug 13 12:00:27 node1 kernel: Buffer I/O error on device dm-2, logical block 2012774400
Aug 13 12:00:27 node1 kernel: lost page write due to I/O error on dm-2
Aug 13 12:00:27 node1 kernel: JBD2: I/O error detected when updating journal superblock for dm-2-8.
Aug 13 12:00:32 node1 kernel: EXT4-fs error (device dm-2): ext4_journal_start_sb: Detected aborted journal
Aug 13 12:00:32 node1 kernel: EXT4-fs (dm-2): Remounting filesystem read-only
Aug 13 12:00:32 node1 rgmanager[29042]: [fs] fs:homefs: is_alive: failed write test on [/nfshome]. Return code: 1
Aug 13 12:00:32 node1 rgmanager[29064]: [fs] fs:homefs: Mount point is not accessible!

• Trying to figure out ifwhat can be done in the cluster.conf file to reduce the amount of time it takes the network to detect that the storage has failed; currently takes 150 seconds