From Bugzilla Helper: User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.0; .NET CLR 1.0.3705; .NET CLR 1.1.4322) Description of problem: "multipath -ll" takes longer time for updating the path status, when the utility is executed just after disabling the switch ports to storage. This time increases as the retrycount and timeout values are increased for Qlogic HBA driver. However "multipath -l" comes out immediately. Version-Release number of selected component (if applicable): device-mapper-multipath-0.4.5-6.0.RHEL4 How reproducible: Always Steps to Reproduce: 1.disable the path to storage, by disabling the port of the concerned switch. 2.issue "multipath -ll" on the host. 3. Actual Results: Once the command is executed it hangs for a long time and comes out with the new status of the paths. Additional info:
*** Bug 168799 has been marked as a duplicate of this bug. ***
From my understanding of 'multipath' and Qlogic driver, this may be expected behavior. Unlike 'multipath -l', 'multipath -ll' will actually try to check the health of a path by sending some command down to the disk. When you disable a switch port, the Qlogic driver will have a port or link down timer period that it waits, suspending I/O during that period, to see if the device will come back up. Since 'multipath -l' does not send any I/O through the QLogic driver, you get quick feedback regardless of what you do on the SAN. But multipath -ll and any other I/O sending command, like 'dd', will seem to 'suspend' during this period. I think Qlogic uses 60 second for their link down timer period . So by 'hangs for a long time', do you mean you are you seeing it hang for around 60 seconds? If so, I don't think this is a bug.
Since this has been uncommented on in almost a year, I'm closing it. If someone still thinks there is a bug here, please reopen it.