[LU-12440] Correct lnet_is_health_check when msg status is 0

• Type: Improvement
• Resolution: Fixed
• Priority: Minor
• Fix Version/s: Lustre 2.13.0
• Assignee: Chris Horn
• Reporter: Chris Horn

Description

I don't think this leads to misbehavior, hence this ticket is not a "Bug", but when we're sending to the lolnd neither msg_txpeer nor msg_rxpeer are set. As a result, all sends to lolnd result in a misleading error message in the dklog from lnet_is_health_check():

00000400:00000200:18.0:1560465004.906149:0:28738:0:(lib-move.c:4730:LNetPut()) LNetPut msg ffff881f44e36a00 -> 12345-0@lo
00000400:00000200:18.0:1560465004.923925:0:28738:0:(lib-msg.c:859:lnet_is_health_check()) msg ffff881f44e36a00 failed too early to retry and send

from this code:

        if ((msg->msg_tx_committed && !msg->msg_txpeer) ||
            (msg->msg_rx_committed && !msg->msg_rxpeer)) {
                CDEBUG(D_NET, "msg %p failed too early to retry and send\n",
                       msg);
                return false;
        }

We should only be printing this message if the message actually failed. i.e. the msg->msg_ev.status != 0.

Comments

[Peter Jones]

Landed for 2.13

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch https://review.whamcloud.com/35235/
Subject: LU-12440 lnet: Misleading error from lnet_is_health_check
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: 6caa6ed07df0ec40f8cb0e13441ab4b0baf90297

[Gerrit Updater]

Chris Horn ([e-mail]) uploaded a new patch: https://review.whamcloud.com/35235
Subject: LU-12440 lnet: Misleading error from lnet_is_health_check
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: 8a3d4d666b4480cdbb731cc3e76d35da6d0479a5