Multipathd does not handle path failover correctly in NetApp cluster with FCoE fabric
Issue
- When attempting a node takeover, hosts lose all 4 paths to storage lun. The expectation is to lose half of the paths, rather than all four.
- RHEL Server connected via dual FCoE to redundant storage controllers through Cisco NX-OS Switch running firmware version 7.1(3)N1(2)
Environment
- Red Hat Enterprise Linux 6
- Storage Controllers: Intel, LSI, Qlogic
- NX-OS Switch firmware version 7.1(3)N1(2)
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.