[LU-9810] Melanox OFED 4.1 support

• Type: Improvement
• Resolution: Fixed
• Priority: Major
• Fix Version/s: Lustre 2.11.0, Lustre 2.10.7
• Affects Version/s: Lustre 2.10.0
• Assignee: Amir Shehata (Inactive)
• Reporter: Alexey Lyashkov

Issue Links

• is related to: LU-9990 MDS fails to mount due to (client.c:96:ptlrpc_uuid_to_connection()) cannot find peer MGC10.37.248.196@o2ib1 _0!
• is related to: LU-9983 LBUG llog_osd.c:327:llog_osd_declare_write_rec() - all DNE MDS

Comments

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch https://review.whamcloud.com/34322/
Subject: LU-9810 lnet: fix build with M-OFED 4.1
Project: fs/lustre-release
Branch: b2_10
Current Patch Set:
Commit: b4c93d99c633003d90f478d999805c76ccd744f1

[Gerrit Updater]

Minh Diep ([e-mail]) uploaded a new patch: https://review.whamcloud.com/34322
Subject: LU-9810 lnet: fix build with M-OFED 4.1
Project: fs/lustre-release
Branch: b2_10
Current Patch Set: 1
Commit: fabd62d9773df6e56b5d0426cb46d1d7e3204d82

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch https://review.whamcloud.com/33974/
Subject: LU-9810 lnd: use less CQ entries for each connection
Project: fs/lustre-release
Branch: b2_10
Current Patch Set:
Commit: 31e16f27ccb18d7c2eb5169f33b1ac55823cc90b

[Gerrit Updater]

Minh Diep ([e-mail]) uploaded a new patch: https://review.whamcloud.com/33974
Subject: LU-9810 lnd: use less CQ entries for each connection
Project: fs/lustre-release
Branch: b2_10
Current Patch Set: 1
Commit: 1045cb96f4e3c36e3003be9f8797a6b80e982bea

[Peter Jones]

I believe that no further work is outstanding on this ticket

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch https://review.whamcloud.com/28279/
Subject: LU-9810 lnd: use less CQ entries for each connection
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: 052f76bf708414b3a127aa9602b4a69415c1cb2f

[Gerrit Updater]

sorry this was meant to be for LU-9990

Amir Shehata ([e-mail]) uploaded a new patch: https://review.whamcloud.com/29333
Subject: LU-9810 lnet: add backwards compatibility for YAML config
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: e72cdc1373dfb930eccbc5d9afba215d8368b331

[James A Simmons]

>what HW you use for testing? if it MLX5, these patches do nothing for you. MLX5 uses a FastReg only model, >while MLX4 support a both Fast and FMR.

The FDR card I have uses the mlx4 driver.

[James A Simmons]

With the prefer Fast Reg patch I'm seeing the follow error using older mlx4 FDR hardware with RHEL7.4.

10000000:01000000:5.0:1505423143.375531:0:6932:0:(mgc_request.c:1205:mgc_target_register()) register lustre-MDT0000
00000800:00000100:1.0:1505423143.375818:0:3469:0:(o2iblnd_cb.c:3464:kiblnd_complete()) FastReg failed: 6
00000800:00000100:1.0:1505423143.375821:0:3469:0:(o2iblnd_cb.c:3475:kiblnd_complete()) RDMA (tx: ffffc90006c03728) failed: 5
00000800:00000100:1.0:1505423143.375834:0:3469:0:(o2iblnd_cb.c:967:kiblnd_tx_complete()) Tx -> 10.37.248.196@o2ib1 cookie 0x1 sending 1 waiting 0: failed 5
00000800:00000100:1.0:1505423143.375838:0:3469:0:(o2iblnd_cb.c:1919:kiblnd_close_conn_locked()) Closing conn to 10.37.248.196@o2ib1: error -5(waiting)
00000100:00000400:5.0:1505423143.375874:0:6932:0:(client.c:2113:ptlrpc_expire_one_request()) @@@ Request sent has failed due to network error: [sent 1505423143/real 1505423143] req@ffff881011e5c300 x1578550577594400/t0(0) o253->MGC10.37.248.196@o2ib1@10.37.248.196@o2ib1:26/25 lens 4768/4768 e 0 to 1 dl 1505423150 ref 2 fl Rpc:eX/0/ffffffff rc 0/-1

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch https://review.whamcloud.com/28278/
Subject: LU-9810 lnet: prefer Fast Reg
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: 8f0d0f052a516a5dd3e588ced6b49c840584855c