received iSCSI errors in /var/log/messags file on Linux

[netappmagic]

I have two NetApp iscsi luns on a Linux server, and via 1Gb Ethernet, with tcp/ip Frame = 9000. but, periodically receiving the following, please help me out. Appreciate your help.

Apr 20 08:09:32 server-name kernel: connection2:0: detected conn error (1011)
Apr 20 08:09:33 server-name iscsid: Kernel reported iSCSI connection 2:0 error (1011 - ISCSI_ERR_CONN_FAILED: iSCSI connection failed) state (3)
Apr 20 08:09:38 server-name iscsid: connection2:0 is operational after recovery (1 attempts)
Apr 20 20:07:10 server-name kernel: connection2:0: ping timeout of 5 secs expired, recv timeout 5, last rx 4401998371, last ping 4402003376, now 4402008384
Apr 20 20:07:10 server-name kernel: connection2:0: detected conn error (1011)
Apr 20 20:07:10 server-name iscsid: Kernel reported iSCSI connection 2:0 error (1011 - ISCSI_ERR_CONN_FAILED: iSCSI connection failed) state (3)
Apr 20 20:07:15 server-name iscsid: connection2:0 is operational after recovery (1 attempts)
Apr 21 01:07:29 server-name kernel: connection2:0: ping timeout of 5 secs expired, recv timeout 5, last rx 4420017295, last ping 4420022296, now 4420027312
Apr 21 01:07:29 server-name kernel: connection2:0: detected conn error (1011)
Apr 21 01:07:29 server-name iscsid: Kernel reported iSCSI connection 2:0 error (1011 - ISCSI_ERR_CONN_FAILED: iSCSI connection failed) state (3)
Apr 21 01:07:35 server-name iscsid: connection2:0 is operational after recovery (1 attempts)

[vshevchuk]

The message you are seeing while not good is not out of the norm.  iscsi is doing it's job and disconnecting after a predetermined time out.  Hopefully you have multipathing configured so as to prevent a disruption in data service.

There is a whole chain of devices involved here: host: ip stack, adapter, network switch, NetApp  network adapter etc.)  So now you get to do the fun stuff and try to nail down what/where is causing the disconnect. Since there are no adapter errors in the messages file it is probably not your host adapter, but no guarantee there.  Host logs should be reviewed further.  Definitely need to look through the rest of the chain.

Good luck