Score:0

Non-valid paths in "multipath -ll" output

co flag
Hem

I see paths that are no longer valid along with valid ones. How do I prevent them showing up in the output. I expect four valid paths.

The volume was earlier exported with LUN 13 and later 14 which shows as failed paths in the below output. The volume was properly flushed from the host and later exported to host with LUN 12. I made sure "multipath -ll" output did not have this volume before exporting with LUN 12.

root# multipath -ll
1703248.744611 | sdar: prio = const (setting: emergency fallback - alua failed)
1703248.745270 | sdas: prio = const (setting: emergency fallback - alua failed)
1703248.745918 | sday: prio = const (setting: emergency fallback - alua failed)
1703248.746573 | sdba: prio = const (setting: emergency fallback - alua failed)
1703248.747204 | sdbf: prio = const (setting: emergency fallback - alua failed)
1703248.747835 | sdbg: prio = const (setting: emergency fallback - alua failed)
mpathhl (1234567891bcdefg123456780911234ab) dm-0 ABC,FlashArray
size=7.0G features='0' hwhandler='1 alua' wp=rw
`-+- policy='round-robin 0' prio=50 status=active
  |- 29:0:0:12  sdan  66:112   active ready running
  |- 25:0:0:12  sdap  66:144   active ready running
  |- 30:0:0:12  sdaq  66:160   active ready running
  |- 13:0:0:12  sdaj  66:48    active ready running
  |- 29:0:0:13  sdar  66:176   failed faulty running
  |- 25:0:0:13  sdas  66:192   failed faulty running
  |- 30:0:0:13  sday  67:32    failed faulty running
  |- 29:0:0:14  sdba  67:64    failed faulty running
  |- 25:0:0:14  sdbf  67:144   failed faulty running
  `- 30:0:0:14  sdbg  67:160   failed faulty running
Score:0
co flag
Hem

I found the issue.

  1. The previous attempts of multipath flush was not done cleanly.
  2. Once it was flushed, associated device references needed to be cleaned under /sys/class/scsi_device/. Set the device to offline and then delete it.
I sit in a Tesla and translated this thread with Ai:

mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.