IO Error

[GradyVogt]

We have installed 2 new IBM 3950 servers with a 2 disk mirror and an iSCSI shared storage on a NetApp 2020.

One of the servers will lock up and roll IO error dev 08:02 sector 5024944 accross the screen. When this happens you can ping the server its just dead. The VM's do NOT move over to the runnign server utill you power off the offline server? THEN he will HA them over. We had IBM look at them as I thought it was a controller/disk issue but they said the saw nothing and wont replace anything. IBM has already replaced the mother board for a different issue. I am in the process of patching the server now but really dont believe this is a sofware issue. My question is this an issue with local disk or ISCSI conenction? The other server is just fine during this outage...

[kjb007]

I've heard of this issue before when a path to storage is lost. Do you have one path to storage, or more than one? Are you using software iSCSI or iSCSI hba's? If sw, make sure you have two pNICs allocated to the portgroup that you are using for iSCSI, and if hw iSCSI hba's, make sure both hba's (assuming you have two here) are configured for the LUNs.

-KjB