Following up on Andrey's post: Eugene's initiator log shows 3 ping timeouts:
Oct 22 19:14:51 redmine kernel: ping timeout of 5 secs expired, last rx 793181268, last ping 793186268, now 793191268 Oct 22 19:15:09 redmine kernel: ping timeout of 5 secs expired, last rx 793198770, last ping 793203770, now 793208770 Oct 22 19:15:27 redmine kernel: ping timeout of 5 secs expired, last rx 793216540, last ping 793221540, now 793226540 (that an 18 seconds interval) (The initiator sends a 'NOP-out' as a ping request to the target to verify a connection/session is still active.) Eugene, are the clock's of the initiator and the target synchronized? If not, what is the offset? (It really makes life easier if all the clocks are sync'd using NTP!) But Andrey, does not the 'iscsisnoop.d' show the 'nop-received' AFTER the target has just done another login, implying to me that another problem has already occurred, and the target has stopped answering because the original session has been dropped. Eugene, it certainly would be useful to have some time stamps on the 'iscsisnoop.d'. Also if you look at this page: http://blogs.sun.com/ahl/entry/iscsi_dtrace_provider_and_other .. maybe you can figure how to get some information on the relevant connection from 'args[0]'. It would be good to see all three outputs: 1. iscsi target debug log 2. 'iscsisnoop.d' output, with timestamps 3. Initiator log .. to get the right sequence and just to try and figure out what is cause and effect. More verbosity from the initiator log may well also help if that's possible. Regards Nigel Smith -- This message posted from opensolaris.org _______________________________________________ storage-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/storage-discuss
