[LU-6388] llog_reader cannot parse changelog records Created: 19/Mar/15 Updated: 18/Jul/16 Resolved: 16/Sep/15 |
|
| Status: | Resolved |
| Project: | Lustre |
| Component/s: | None |
| Affects Version/s: | Lustre 2.7.0 |
| Fix Version/s: | Lustre 2.8.0 |
| Type: | Bug | Priority: | Minor |
| Reporter: | Olaf Faaland | Assignee: | Emoly Liu |
| Resolution: | Fixed | Votes: | 0 |
| Labels: | llnl, llnlfixready, patch | ||
| Issue Links: |
|
||||
| Severity: | 3 | ||||
| Rank (Obsolete): | 9223372036854775807 | ||||
| Description |
|
llog_reader does not handle records of type CHANGELOG_REC or CHANGELOG_USER_REC. In addition, its output for records in the changelog_catalog (type LLOG_LOGID_MAGIC) is more obscure than it needs to be. The change submitted adds support for those unhandled record types. It also provides the actual pathname to the lustre log containing changelog records, that one would need when mounting an ldiskfs or zfs target via the posix layer for inspection. |
| Comments |
| Comment by Olaf Faaland [ 19/Mar/15 ] |
|
Patch to be submitted to gerrit RSN. |
| Comment by Gerrit Updater [ 20/Mar/15 ] |
|
Olaf Faaland-LLNL (faaland1@llnl.gov) uploaded a new patch: http://review.whamcloud.com/14120 |
| Comment by Olaf Faaland [ 20/Mar/15 ] |
|
Submitted the patch, link above. I'll check with Li Xi re: |
| Comment by Peter Jones [ 20/Mar/15 ] |
|
Emoly Could you please take care of this patch? Thanks Peter |
| Comment by Olaf Faaland [ 05/May/15 ] |
|
Hello Emoly, Please suggest a next step. Lustre-initialization-1 within review-zfs failed because mkfs.lustre failed to initialize the ZFS OSD on shadow-21vm7 (the OSS). Looking at the syslog files for 21vm7 (OSS) and 21vm8 (MDS) I see that yum installed the lustre-osd-zfs-mount and lustre-osd-zfs packages on 21vm8 but not on 21vm7. This patch does not touch the zfs OSD in any way, and local testing via llmount.sh with FSTYPE=zfs is successful. See below from lustre-initialization-1.suite_stdout.shadow-21vm9.log lustre-initialization-1.suite_stdout.shadow-21vm9.log:00:33:15:CMD: shadow-21vm7 mkfs.lustre --mgsnode=shadow-21vm8@tcp --fsname=lustre --ost --index=0 --param=sys.timeout=20 --backfstype=zfs --device-size=8388608 --reformat lustre-ost1/ost1 /dev/lvm-Role_OSS/P1 lustre-initialization-1.suite_stdout.shadow-21vm9.log:00:33:15:shadow-21vm7: mkfs.lustre FATAL: unhandled fs type 5 'zfs' lustre-initialization-1.suite_stdout.shadow-21vm9.log:00:33:15:shadow-21vm7: mkfs.lustre FATAL: unable to prepare backend (22) lustre-initialization-1.suite_stdout.shadow-21vm9.log:00:33:15:shadow-21vm7: mkfs.lustre: exiting with 22 (Invalid argument) So it looks like the script that sets up the test environment failed to generate the correct yum install commands on the oss node. Please let me know how to address this. thanks, |
| Comment by Emoly Liu [ 06/May/15 ] |
|
Olaf, I will re-trigger the test and see if this error happens again. |
| Comment by Olaf Faaland [ 03/Aug/15 ] |
|
Hi Emoly, |
| Comment by Gerrit Updater [ 15/Sep/15 ] |
|
Oleg Drokin (oleg.drokin@intel.com) merged in patch http://review.whamcloud.com/14120/ |
| Comment by Joseph Gmitter (Inactive) [ 16/Sep/15 ] |
|
Landed for 2.8. |
| Comment by D. Marc Stearman (Inactive) [ 19/Feb/16 ] |
|
Olaf, Chris, do we have this patch in our local release? |
| Comment by Olaf Faaland [ 22/Feb/16 ] |
|
Marc, Yes, this patch is in lustre 2.5.5-3chaos (included in toss 2.4-6.ch5.4) and I believe was in earlier releases as well. -Olaf |