MD3220i - Slow Performance

Good Afternoon,

We had an MD3220i with multiple issues, so we selected to build an entire replacement array with identical parts. I configured the replacement array to match the existing array with matching iSCSI settings, etc., but we're experiencing slow performance from our ESXi 5.5 cluster.

As I went though each setting after iSCSI targets in VMware I noticed that I didn't turn on jumbo frames in MDSM for each iSCSI port. I'm guessing this is the issue as it is enabled on the existing (failing) array.

My question: After the final configuration to enable jumbo frames from MDSM, would this require a reboot of the array, my hosts, or both?

Thanks in advance,

Kevin