[LU-16563] LNet: use discovery ni status to set peer ni availability

• Type: Improvement
• Resolution: Fixed
• Priority: Minor
• Fix Version/s: Lustre 2.16.0
• Labels: Multi-Rail lnet
• Assignee: Serguei Smirnov
• Reporter: Serguei Smirnov

Description

Currently when MR peer is being discovered, it replies with the list of its NIs and their status.  Even if NI is "down" due to a "fatal" condition like locally detected "link down", it is listed with "UP" status in the reply, so the recipient can find out that the NI is not reachable only by trying to communicate to it and failing.

Instead, to avoid unnecessary delay in this scenario, NI status can be tracked such that locally recognized "down" state is available to any discovering peer.

Comments

[Peter Jones]

Landed for 2.16

[Gerrit Updater]

"Oleg Drokin <[e-mail]>" merged in patch https://review.whamcloud.com/c/fs/lustre-release/+/50188/
Subject: LU-16563 tests: Check peer NI health after link down
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: e82e57414d324c1065ddbdaef5baab2ec5b42026

[Gerrit Updater]

"Oleg Drokin <[e-mail]>" merged in patch https://review.whamcloud.com/c/fs/lustre-release/+/50027/
Subject: LU-16563 lnet: use discovered ni status to set initial health
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: da230373bd14306cb97fb48748ebce205f09d468

[Gerrit Updater]

"Chris Horn <[e-mail]>" uploaded a new patch: https://review.whamcloud.com/c/fs/lustre-release/+/50188
Subject: LU-16563 tests: Check peer NI health after link down
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: e4b96c5512c3d2d2c5bda49d23cb445b05eb9678

[Gerrit Updater]

"Serguei Smirnov <[e-mail]>" uploaded a new patch: https://review.whamcloud.com/c/fs/lustre-release/+/50027
Subject: LU-16563 lnet: use discovered ni status to set initial health
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: 2244b767ff0920c0ce488b5147e92744ed462c24