[LU-11308] LustreError: 2742:0:(events.c:199:client_bulk_callback()) event type 2, status -90, desc ffff8807e3876400

• Type: Bug
• Resolution: Duplicate
• Priority: Critical
• Affects Version/s: Lustre 2.7.0, Lustre 2.11.0
• Environment: client 2.11.0 (mofed4.2.1) ; servers 2.7.3 (mofed 3.3)
• Severity: 2
• Assignee: Amir Shehata (Inactive)
• Reporter: Mahmoud Hanafi

Description

IO on the client hangs with these errors

 1535653023.151966] LustreError: 2740:0:(events.c:199:client_bulk_callback()) event type 2, status -90, desc ffff8807e3876400
[1535653024.151931] LustreError: 2740:0:(events.c:199:client_bulk_callback()) event type 2, status -90, desc ffff8807e3876400
[1535653025.151896] LustreError: 2743:0:(events.c:199:client_bulk_callback()) event type 2, status -90, desc ffff8807e3876400

Attaching client side debug logs.

Issue Links

• duplicates: LU-11064 o2iblnd fast reg gaps case is determined incompletely

Comments

[Peter Jones]

Nasa confirmed that patch was not applied

[Mahmoud Hanafi]

Thank you I will double check the build to make sure we applied the patch correctly.

[Amir Shehata (Inactive)]

Hi Mahmoud,

From the logs attached I see that you haven't applied
https://review.whamcloud.com/#/c/32586/4

o2iblnd_cb.c:811:kiblnd_setup_rd_kiov()) fragnob 3564 < available page 4096: with remaining 138 kiovs

what should be printed out is:

  838 »·······»·······»·······CDEBUG(D_NET, "fragnob %d < available page %d: with"
 839 »·······»·······»·······»·······      " remaining %d kiovs with %d nob left\n",
 840 »·······»·······»·······       fragnob, (int)(kiov->kiov_len - offset),
 841 »·······»·······»·······       nkiov, nob);

I believe that patch should fix the bug you're seeing

[Mahmoud Hanafi]

We initially hit bug in LU-11064. We applied that patch and now we are getting these errors.