For some IBM storage arrays or virtual arrays like the SVC we see inquiries failing shortly after the arrays have been rebooted.
As a possible side effect we can witness on systems with eh_deadline enabled that the SCSI enters the escalation ladder up to the HBA reset and therefore interrupts connections to other ports / LUNs as a consequence of failed / timed-out commands to the IBM array.
- Red Hat Enterprise Linux (any version) connected to
- IBM storage arrays (it's not clear yet, which arrays show the behavior and which don't).
- So far we know that the problem happens with IBM v7000 storage and the SVC.
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase of over 48,000 articles and solutions.