Recently on testing an MSA serving iSCSI hosts after modifying the MTU size to varying amounts to tests throughput of a hyper V host our SNMP monitoring system detected hundreds of Inbound Errors and 10’s of discard.

To resolve this reversal was performed on the SAN to disable Jumbo frames and to change he MTU size on the Switch and the Networks cards, however this errors still displayed on the SAN Eth 0-2 ports and did nothing to resolve this. In the end we had to restart each of the storage controllers ( A and B )  to resolve this