Why did the connections on the RHEL 5.8 servers not automatically recover when they did on other servers?
Issue
- We have had a number of RHEL 5.8 SAN attached servers dropped SAN connections all with the same error:
kernel: rport-4:0-2: blocked FC remote port time out: saving binding
kernel: sd 4:0:0:6: Unhandled error code
kernel: sd 4:0:0:6: SCSI error: return code = 0x00010000
kernel: Result: hostbyte=DID_NO_CONNECT driverbyte=DRIVER_OK,SUGGEST_OK
- Why did the connections on the RHEL 5.8 servers not automatically recover when they did on other servers? (Solaris, RHEL 5.5, ...)
- Do we have to manually rescan every time we get an RSCN?
Environment
- Red Hat Enterprise Linux 5.8 with Veritas VxDMP multpath software
- The following servers were unaffected by the event shown above:
- Solaris servers
- Windows Servers
- Red Hat Enterprise Linux 5.5 server
- Red Hat Enterprise Linux 5.8 server without Veritas VxDMP multipath software.
- The following servers were unaffected by the event shown above:
- Redundant Storage Fabric Switches (Cisco MDS 9509).
- Each server has redundant paths to the storage
- Emulex Host-Bus Adapter (HBA)
- lpfc driver shipped with Red Hat Enterprise Linux
- Multiple HBA's with multiple links to each device. Storage is multipathed.
- Hitachi VSP Storage array
- A change was made to the fabric switch which inexplicably caused an RSCN to be sent to the SAN at the time of the issue.
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.
