Mon: Add new health warning for non prime w+1 in blaum-roth EC profiles by Tom-Sollers · Pull Request #66129 · ceph/ceph

Tom-Sollers · 2025-11-05T10:35:55Z

This PR is the second attempt of the original PR, which was merged to early and reverted in #66114

This PR adds a health warning for when a Erasure code profile uses the blaum-roth technique and has a w value such that w+1 is not prime, which can cause data corruption. It also improves the validation for when the user attempts to create a new erasure code profile by throwing a more descriptive warning when an invalid w value is used, and allows the user to override this validation using the --yes-i-really-mean-it flag. In addition to the previous changes this PR also adds a new test to /qa/standalone/erasure-code/test-erasure-code-plugins.sh to check that the health warning is functioning correctly.

Fixes: http://tracker.ceph.com/issues/64419
Signed-off-by: Tom Sollers tom.sollers@ibm.com

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)
- Includes unit test(s)
- Includes integration test(s)
- Includes bug reproducer
- No tests

Show available Jenkins commands

jenkins test classic perf Jenkins Job | Jenkins Job Definition
jenkins test crimson perf Jenkins Job | Jenkins Job Definition
jenkins test signed Jenkins Job | Jenkins Job Definition
jenkins test make check Jenkins Job | Jenkins Job Definition
jenkins test make check arm64 Jenkins Job | Jenkins Job Definition
jenkins test submodules Jenkins Job | Jenkins Job Definition
jenkins test dashboard Jenkins Job | Jenkins Job Definition
jenkins test dashboard cephadm Jenkins Job | Jenkins Job Definition
jenkins test api Jenkins Job | Jenkins Job Definition
jenkins test docs ReadTheDocs | Github Workflow Definition
jenkins test ceph-volume all Jenkins Jobs | Jenkins Jobs Definition
jenkins test windows Jenkins Job | Jenkins Job Definition
jenkins test rook e2e Jenkins Job | Jenkins Job Definition

You must only issue one Jenkins command per-comment. Jenkins does not understand
comments with more than one command.

github-actions · 2025-11-05T10:36:14Z

This pull request can no longer be automatically merged: a rebase is needed and changes have to be manually resolved

anthonyeleven · 2025-11-05T14:23:00Z

doc/rados/operations/health-checks.rst

+An EC pool is using the ``blaum_roth`` technique and ``w + 1`` is not a prime number. 
+This can result in data corruption.
+
+To check the list of Erasure Code Profiles use the command:


anthonyeleven · 2025-11-05T14:23:28Z

doc/rados/operations/health-checks.rst

+
+   ceph osd erasure-code-profile ls
+
+Then to check the ``w`` value for a particular profile use the command:


a command of the following form:

batrick

Did you locate the cause of the QA error in

#65940 (comment)

src/mon/Monitor.h

src/mon/Monitor.cc

doc/rados/operations/health-checks.rst

src/erasure-code/jerasure/ErasureCodeJerasure.cc

src/erasure-code/jerasure/ErasureCodeJerasure.h

src/mon/HealthMonitor.cc

src/mon/OSDMonitor.cc

Tom-Sollers · 2025-11-05T16:38:16Z

Did you locate the cause of the QA error in

#65940 (comment)

Yeah so in HealthMonitor.cc line 1332. I have added a check for if the "technique" attribute exists in the metadata for the EC profile. The error was caused by line 1333 checking the value of "technique" when there wasn't one, this is now fixed.

src/common/prime.cc

Tom-Sollers · 2025-11-12T11:49:32Z

jenkins test make check

Tom-Sollers · 2025-11-12T11:50:27Z

jenkins test make check arm64

Tom-Sollers · 2025-11-12T14:08:07Z

jenkins test

qa/standalone/erasure-code/test-erasure-code-plugins.sh

src/mon/HealthMonitor.cc

Tom-Sollers · 2025-11-13T14:20:28Z

To avoid making the same mistake from the last PR, I have ran a teuthology test running a subset of runs from the RADOS tests suite to make sure there are no new regressions caused by the change.
https://pulpito.ceph.com/tsollers-2025-11-11_14:14:01-rados-main-distro-default-smithi/
I also ran 2 tests that contained the same configuration that broke the previous PR to demonstrate that the issue is fixed.
https://pulpito.ceph.com/tsollers-2025-11-13_09:21:25-rados-main-distro-default-smithi/

batrick

otherwise LGTM

src/common/prime.h

src/common/prime.cc

doc/rados/operations/health-checks.rst

…th a w+1 that is not prime This commit adds a new health warning for when a user has an erasure code profile using the blaum-roth technique which has a w+1 value that is not prime. Fixes: http://tracker.ceph.com/issues/64419 Signed-off-by: Tom Sollers <tom.sollers@ibm.com>

…alth warn This commit adds a new test to test-erasure-code-plugins.sh that tests for the health warning caused by having a erasure-code-profile with the blaum-roth technique and a w+1 value that is not prime. Fixes: http://tracker.ceph.com/issues/64419 Signed-off-by: Tom Sollers <tom.sollers@ibm.com>

Tom-Sollers · 2025-11-18T16:02:22Z

Pushed what should hopefully be the final version of the PR

bill-scales

LGTM. Once Patrick approves we can set the needs-qa flag and get a full run of the rados suite to double check that there are no regressions.

Naveenaidu · 2025-12-09T03:13:28Z

RADOS Approved: https://tracker.ceph.com/issues/73968#note-3

Tom-Sollers self-assigned this Nov 5, 2025

Tom-Sollers requested review from a team as code owners November 5, 2025 10:35

github-actions bot added core documentation mon tests needs-rebase labels Nov 5, 2025

Tom-Sollers added the needs-qa label Nov 5, 2025

Tom-Sollers marked this pull request as draft November 5, 2025 10:47

github-actions bot removed the needs-rebase label Nov 5, 2025

Tom-Sollers force-pushed the fixing_blaum_roth_for_ec_profiles branch from b96a598 to 9299700 Compare November 5, 2025 13:37

anthonyeleven reviewed Nov 5, 2025

View reviewed changes

batrick requested changes Nov 5, 2025

View reviewed changes

batrick removed the needs-qa label Nov 5, 2025

github-actions bot added the common label Nov 10, 2025

batrick reviewed Nov 10, 2025

View reviewed changes

src/common/prime.cc Show resolved Hide resolved

Tom-Sollers force-pushed the fixing_blaum_roth_for_ec_profiles branch from 15ba199 to d91344b Compare November 11, 2025 11:52

github-actions bot added the build/ops label Nov 11, 2025

bill-scales reviewed Nov 13, 2025

View reviewed changes

qa/standalone/erasure-code/test-erasure-code-plugins.sh Outdated Show resolved Hide resolved

src/mon/HealthMonitor.cc Outdated Show resolved Hide resolved

Tom-Sollers force-pushed the fixing_blaum_roth_for_ec_profiles branch 3 times, most recently from e7609ec to a10026c Compare November 13, 2025 13:57

Tom-Sollers marked this pull request as ready for review November 13, 2025 14:20

Tom-Sollers force-pushed the fixing_blaum_roth_for_ec_profiles branch from a10026c to d71d61f Compare November 13, 2025 16:19

Tom-Sollers requested review from anthonyeleven, batrick and bill-scales November 14, 2025 10:10

batrick requested changes Nov 14, 2025

View reviewed changes

src/common/prime.h Show resolved Hide resolved

src/common/prime.cc Show resolved Hide resolved

src/common/prime.cc Show resolved Hide resolved

anthonyeleven reviewed Nov 14, 2025

View reviewed changes

doc/rados/operations/health-checks.rst Show resolved Hide resolved

Tom-Sollers force-pushed the fixing_blaum_roth_for_ec_profiles branch from d71d61f to 74a8ab3 Compare November 14, 2025 13:53

Tom-Sollers added 2 commits November 18, 2025 16:01

Tom-Sollers force-pushed the fixing_blaum_roth_for_ec_profiles branch from 74a8ab3 to e61450c Compare November 18, 2025 16:01

Tom-Sollers requested review from anthonyeleven and batrick November 18, 2025 16:10

bill-scales approved these changes Nov 18, 2025

View reviewed changes

batrick approved these changes Nov 20, 2025

View reviewed changes

batrick added the needs-qa label Nov 20, 2025

anthonyeleven approved these changes Nov 20, 2025

View reviewed changes

SrinivasaBharath added the wip-bharath11-testing label Nov 20, 2025

rzarzynski approved these changes Dec 15, 2025

View reviewed changes

rzarzynski merged commit 551ac2b into ceph:main Dec 15, 2025
13 checks passed

SrinivasaBharath removed the wip-bharath11-testing label Jan 6, 2026


		ceph osd erasure-code-profile ls

		Then to check the ``w`` value for a particular profile use the command:

Conversation

Tom-Sollers commented Nov 5, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Contribution Guidelines

Checklist

Uh oh!

github-actions bot commented Nov 5, 2025

Uh oh!

anthonyeleven Nov 5, 2025

Choose a reason for hiding this comment

Uh oh!

anthonyeleven Nov 5, 2025

Choose a reason for hiding this comment

Uh oh!

anthonyeleven Nov 5, 2025

Choose a reason for hiding this comment

Uh oh!

batrick left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Tom-Sollers commented Nov 5, 2025

Uh oh!

Uh oh!

Tom-Sollers commented Nov 12, 2025

Uh oh!

Tom-Sollers commented Nov 12, 2025

Uh oh!

Tom-Sollers commented Nov 12, 2025

Uh oh!

Uh oh!

Uh oh!

Tom-Sollers commented Nov 13, 2025

Uh oh!

batrick left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Tom-Sollers commented Nov 18, 2025

Uh oh!

bill-scales left a comment

Choose a reason for hiding this comment

Uh oh!

Naveenaidu commented Dec 9, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

Tom-Sollers commented Nov 5, 2025 •

edited

Loading