[LU-5791] LFSCK 5: use bottom object for consistency verification

• Type: Bug
• Resolution: Fixed
• Priority: Blocker
• Fix Version/s: Lustre 2.7.0
• Affects Version/s: Lustre 2.7.0
• Severity: 3
• Assignee: nasf (Inactive)
• Reporter: nasf (Inactive)

Description

When the LFSCK engine scans the system, it locate the object via the found FID against the low layer device that sometimes is bottom device (OSD/OSP) and sometimes is LOD. So the object handled by LFSCK sometimes is OSD-object/OSP-objecgt and sometimes is LOD-object. Such non-unity may bring trouble to some LFSCK functions. The desirable case is that all the LFSCK functions only handle bottom objects (OSD-object for local, OSP-object for remote).

Please refer to http://review.whamcloud.com/#/c/11845/6/lustre/osd-ldiskfs/osd_handler.c,cm for more discussion.

Issue Links

• is related to: LU-6138 lfsck_namespace not progressing and lfsck_stop hangs
• is related to: LU-6272 sanity-lfsck test_17: MDS deadlock

Comments

[nasf (Inactive)]

The patch has been landed to master.

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch http://review.whamcloud.com/13392/
Subject: LU-5791 lfsck: use bottom device to locate object
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: f6d744d259fe6bc0b8750127565fb13fb2e66c06

[nasf (Inactive)]

We need this patch to resolve some memory leak issues caused by the LFSCK.

[Di Wang (Inactive)]

Nasf: It is not necessarily on 2.7, and I only need it to be landed before landing DNE2 patches.

[nasf (Inactive)]

To be not block DNE2, this patch needs to be handled with high priority.

Di, do you need this patch to be landed before b2_7 released or not?

[Gerrit Updater]

Fan Yong ([e-mail]) uploaded a new patch: http://review.whamcloud.com/13392
Subject: LU-5791 lfsck: use bottom device to locate object
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: 1bc16a18387e0efda8e61b08d7fdbca28691a075

[Di Wang (Inactive)]

Unfortunately, this has to be fixed first. Because this will conflict with thandle layer patch (http://review.whamcloud.com/#/c/10640/), which is the first patch of DNE2 series. Without this fix, we can not even land any DNE patches, i.e. LFSCK will be broken because of this.

So could you please fix this first, i.e. as your priority ticket. Thank you.