Skip to content

mds: don't add counters in warning for standby-replay MDS#56065

Merged
rishabh-d-dave merged 2 commits intoceph:mainfrom
rishabh-d-dave:mds-counters
May 3, 2024
Merged

mds: don't add counters in warning for standby-replay MDS#56065
rishabh-d-dave merged 2 commits intoceph:mainfrom
rishabh-d-dave:mds-counters

Conversation

@rishabh-d-dave
Copy link
Contributor

@rishabh-d-dave rishabh-d-dave commented Mar 8, 2024

Don't print cache size warnings for standby-replay MDS since it is
simply replaying. Repeating warnings confuses users that cache for 2
MDSs is crossing the limit.

Fixes: https://tracker.ceph.com/issues/63514

Contribution Guidelines

  • To sign and title your commits, please refer to Submitting Patches to Ceph.

  • If you are submitting a fix for a stable branch (e.g. "quincy"), please refer to Submitting Patches to Ceph - Backports for the proper workflow.

  • When filling out the below checklist, you may click boxes directly in the GitHub web UI. When entering or editing the entire PR message in the GitHub web UI editor, you may also select a checklist item by adding an x between the brackets: [x]. Spaces and capitalization matter when checking off items this way.

Checklist

  • Tracker (select at least one)
    • References tracker ticket
    • Very recent bug; references commit where it was introduced
    • New feature (ticket optional)
    • Doc update (no ticket needed)
    • Code cleanup (no ticket needed)
  • Component impact
    • Affects Dashboard, opened tracker ticket
    • Affects Orchestrator, opened tracker ticket
    • No impact that needs to be tracked
  • Documentation (select at least one)
    • Updates relevant documentation
    • No doc update is appropriate
  • Tests (select at least one)
Show available Jenkins commands
  • jenkins retest this please
  • jenkins test classic perf
  • jenkins test crimson perf
  • jenkins test signed
  • jenkins test make check
  • jenkins test make check arm64
  • jenkins test submodules
  • jenkins test dashboard
  • jenkins test dashboard cephadm
  • jenkins test api
  • jenkins test docs
  • jenkins render docs
  • jenkins test ceph-volume all
  • jenkins test ceph-volume tox
  • jenkins test windows
  • jenkins test rook e2e

@rishabh-d-dave rishabh-d-dave requested a review from a team as a code owner March 8, 2024 15:36
@rishabh-d-dave rishabh-d-dave added the cephfs Ceph File System label Mar 8, 2024
@rishabh-d-dave rishabh-d-dave requested a review from a team March 8, 2024 15:36
@rishabh-d-dave rishabh-d-dave removed the request for review from a team March 8, 2024 15:38
@rishabh-d-dave rishabh-d-dave changed the title src/mds: no cache size warning for standby-replay MDS mds: no cache size warning for standby-replay MDS Mar 8, 2024
@rishabh-d-dave
Copy link
Contributor Author

rishabh-d-dave commented Mar 13, 2024

Note to reviewers, the test is 90% ready. Leaving it here for now till we're sure that modifications to src/ are 100% correct.

@rishabh-d-dave
Copy link
Contributor Author

rishabh-d-dave commented Mar 14, 2024

Note to reviewers, the test is 90% ready. Leaving it here for now till we're sure that modifications to src/ are 100% correct.

@vshankar Based on your reviews, I am assuming that I am moving in the right direction, I am proceeding to add finishing touches to this PR.

@rishabh-d-dave rishabh-d-dave force-pushed the mds-counters branch 7 times, most recently from dffdca8 to dea6c6f Compare April 7, 2024 14:08
@rishabh-d-dave
Copy link
Contributor Author

@vshankar PTAL

Don't include inode and stray counters in the health warnings printed
for standby-replay MDSs. Since these counters are present in the health
warnings only due to replay, it can confuse users, and therefore, do not
include them.

Fixes: https://tracker.ceph.com/issues/63514
Signed-off-by: Rishabh Dave <ridave@redhat.com>
@rishabh-d-dave
Copy link
Contributor Author

Unrelated error on make check failure (Agent went offline during the build). Running it again.

@rishabh-d-dave
Copy link
Contributor Author

jenkins make check

@rishabh-d-dave
Copy link
Contributor Author

jenkins test make check

@rishabh-d-dave
Copy link
Contributor Author

@rishabh-d-dave
Copy link
Contributor Author

Minor, non-functional change that'll make traceback more readable.

@rishabh-d-dave rishabh-d-dave changed the title mds: no cache size warning for standby-replay MDS mds: don't add counters in warning for standby-replay MDS Apr 22, 2024
- Add tests to verify that inode and stray counters are not
  replayed/included in the health warnings printed for the
  standby-replay MDS.

- Add "MDS_CACHE_OVERSIZED" health warning to ignorelist to
  failover.yaml.

- Add a helper method to qa.tasks.cephfs.filesystem.Filesystem to get
  MDS name of standby-replay MDS.

Signed-off-by: Rishabh Dave <ridave@redhat.com>
@rishabh-d-dave
Copy link
Contributor Author

Added MDS_CACHE_OVERSIZED to ignorelist, otherwise test_failover.py finishes running successfully but teuthology job is marked as failed since this warning wasn't part of ignorelist for the job.

@rishabh-d-dave
Copy link
Contributor Author

@vshankar
Copy link
Contributor

Minor, non-functional change that'll make traceback more readable.

Could you please explain a bit more about this? Why wasn't the traceback less readable before this update?

@rishabh-d-dave
Copy link
Contributor Author

@vshankar

Minor, non-functional change that'll make traceback more readable.

Could you please explain a bit more about this? Why wasn't the traceback less readable before this update?

The command arguments were not on the same line as the function call, which made it impossible to look at what command failed by looking at traceback. It looked like following -

self.get_ceph_cmd_stdout(

With that change it will look like following -

self.get_ceph_cmd_stdout('health detail'

It might look minor but while debugging I don't need to remember the line number from traceback and then check QA code to find out what command failed. The command failed is apparent from code snippet in traceback itself which makes debugging much much easier and quicker since a test method may have multiple calls to such methods.

@vshankar
Copy link
Contributor

vshankar commented May 2, 2024

@vshankar

Minor, non-functional change that'll make traceback more readable.

Could you please explain a bit more about this? Why wasn't the traceback less readable before this update?

The command arguments were not on the same line as the function call, which made it impossible to look at what command failed by looking at traceback. It looked like following -

self.get_ceph_cmd_stdout(

With that change it will look like following -

self.get_ceph_cmd_stdout('health detail'

It might look minor but while debugging I don't need to remember the line number from traceback and then check QA code to find out what command failed. The command failed is apparent from code snippet in traceback itself which makes debugging much much easier and quicker since a test method may have multiple calls to such methods.

As discussed, please share how the old and new backtraces look like. @rishabh-d-dave

@rishabh-d-dave
Copy link
Contributor Author

@vshankar
Snippet from Traceback 1 -

        2024-04-12T15:31:30.745 INFO:tasks.cephfs_test_runner:  File "/home/teuthworker/src/github.com_rishabh-d-dave_ceph_fade9250abf1848492bcb8474d2c5b8ce569ede4/qa/tasks/cephfs/test_failover.py", line 747, in test_health_warn_oversize_cache_has_no_counters
        2024-04-12T15:31:30.745 INFO:tasks.cephfs_test_runner:    self.get_ceph_cmd_stdout(

Snippet from Traceback 2 -

        2024-04-12T15:31:30.745 INFO:tasks.cephfs_test_runner:  File "/home/teuthworker/src/github.com_rishabh-d-dave_ceph_fade9250abf1848492bcb8474d2c5b8ce569ede4/qa/tasks/cephfs/test_failover.py", line 747, in test_health_warn_oversize_cache_has_no_counters
        2024-04-12T15:31:30.745 INFO:tasks.cephfs_test_runner:    self.get_ceph_cmd_stdout('health detail'

rishabh-d-dave added a commit to rishabh-d-dave/ceph that referenced this pull request May 3, 2024
* refs/pull/56065/head:
	qa/cephfs: test that counters are not printed for SR MDS
	mds: add no counters in warning for standby-replay MDS

Reviewed-by: Venky Shankar <vshankar@redhat.com>
Reviewed-by: Dhairya Parmar <dparmar@redhat.com>
Copy link
Contributor Author

@rishabh-d-dave rishabh-d-dave left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

QA run was successful - https://tracker.ceph.com/projects/cephfs/wiki/main#3-May-2024.

Testing took more time than expected because there were 25-30 new failures. Most of them caused by a PR in the testing branch but these were resolved on removing that PR.

@vshankar
Copy link
Contributor

vshankar commented May 3, 2024

@vshankar Snippet from Traceback 1 -

        2024-04-12T15:31:30.745 INFO:tasks.cephfs_test_runner:  File "/home/teuthworker/src/github.com_rishabh-d-dave_ceph_fade9250abf1848492bcb8474d2c5b8ce569ede4/qa/tasks/cephfs/test_failover.py", line 747, in test_health_warn_oversize_cache_has_no_counters
        2024-04-12T15:31:30.745 INFO:tasks.cephfs_test_runner:    self.get_ceph_cmd_stdout(

Snippet from Traceback 2 -

        2024-04-12T15:31:30.745 INFO:tasks.cephfs_test_runner:  File "/home/teuthworker/src/github.com_rishabh-d-dave_ceph_fade9250abf1848492bcb8474d2c5b8ce569ede4/qa/tasks/cephfs/test_failover.py", line 747, in test_health_warn_oversize_cache_has_no_counters
        2024-04-12T15:31:30.745 INFO:tasks.cephfs_test_runner:    self.get_ceph_cmd_stdout('health detail'

Looks fine.

@rishabh-d-dave
Copy link
Contributor Author

Linking to QA run summary once more - #56065 (review)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants