Receiving error on iscsi device: "connectionX:0: detected conn error" when NIC has jumbo frames enabled using RHEL

Issue

• When running a mkfs on an iscsi device we get the following errors:

daemon warning: 2015-04-14T21:13:39.949363+00:00 hostname multipathd: mpathd: sdd - directio checker reports path is down
daemon notice: 2015-04-14T21:13:39.949422+00:00 hostname multipathd: checker failed path 8:48 in map mpathd
daemon notice: 2015-04-14T21:13:39.949428+00:00 hostname multipathd: mpathd: remaining active paths: 3
kern warning: 2015-04-14T21:13:39.950285+00:00 hostname kernel: device-mapper: multipath: Failing path 8:48.
kern info: 2015-04-14T21:13:39.997200+00:00 hostname kernel: connection2:0: detected conn error (1020)
kern info: 2015-04-14T21:13:39.997222+00:00 hostname kernel: connection3:0: detected conn error (1020)
daemon warning: 2015-04-14T21:13:40.362350+00:00 hostname iscsid: Kernel reported iSCSI connection 2:0 error (1020 - ISCSI_ERR_TCP_CONN_CLOSE: TCP connection closed) state (3)
daemon warning: 2015-04-14T21:13:40.362391+00:00 hostname iscsid: Kernel reported iSCSI connection 3:0 error (1020 - ISCSI_ERR_TCP_CONN_CLOSE: TCP connection closed) state (3)
daemon warning: 2015-04-14T21:13:43.363023+00:00 hostname iscsid: connection2:0 is operational after recovery (1 attempts)

• I get detected conn error messages in the logs when using iSCSI connections when I have jumbo frames enabled, but not when I don't.