[LU-11472] LNet Health: Decrement health value on response timeout

• Type: Bug
• Resolution: Fixed
• Priority: Major
• Fix Version/s: Lustre 2.12.0
• Affects Version/s: Lustre 2.12.0
• Labels: lnet-health
• Severity: 3
• Assignee: Amir Shehata (Inactive)
• Reporter: Amir Shehata (Inactive)

Description

When a response times out we want to decrement the health of the immediate next hop peer ni, so we don't use that interface if there are others available.

Comments

[Peter Jones]

Landed for 2.12

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch https://review.whamcloud.com/33308/
Subject: LU-11472 lnet: Decrement health on timeout
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: 139d69141b73d427490f39d3096b2187e979eaea

[Gerrit Updater]

Amir Shehata ([e-mail]) uploaded a new patch: https://review.whamcloud.com/33308
Subject: LU-11472 lnet: Decrement health on timeout
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: 56b5ef7e5c7a1d0c7aca23504acb2b2bc5862199

[Gerrit Updater]

Amir Shehata ([e-mail]) uploaded a new patch: https://review.whamcloud.com/33295
Subject: LU-11472 lnet: Decrement health on timeout
Project: fs/lustre-release
Branch: multi-rail
Current Patch Set: 1
Commit: 9df50755373be42b64f640e664f1e05690f37531

[Amir Shehata (Inactive)]

I have a patch which I'll commit shortly. However, although this is going to work for directly connected. The behavior might be an issue for routing. If the route is servicing multiple connected peer through the same route. If one of the final destinations has a problem and it doesn't respond then the router interface will be dinged. It'll be put on the recovery queue, and recover, but during that period of time the route will be down.