Skip to content

DAOS-17905 object: fixes excessive waiting time in system xstream dur… - #16740

Merged
daltonbohning merged 5 commits into
release/2.6from
shilongw/DAOS-17905-release2.6
Sep 15, 2025
Merged

daltonbohning merged 5 commits into
release/2.6from
shilongw/DAOS-17905-release2.6

Conversation

@wangshilong

Copy link
Copy Markdown
Contributor

…ing object ULT creation

During migration, the system xstream only wakes up in ds_migrate_query_status() (called by rebuild_tgt_status_check_ult()), where it sleeps for RBLD_CHECK_INTV (2 seconds). This caused the system xstream to remain blocked for 2-second intervals without performing any operations, severely degrading rebuild performance.

Benchmark results demonstrate the improvement:

Without patch: 99,780 objects processed in 120 seconds

With patch: 225,136 objects processed in 11 seconds

The patch achieves approximately 20x performance improvement

Steps for the author:

  • Commit message follows the guidelines.
  • Appropriate Features or Test-tag pragmas were used.
  • Appropriate Functional Test Stages were run.
  • At least two positive code reviews including at least one code owner from each category referenced in the PR.
  • Testing is complete. If necessary, forced-landing label added and a reason added in a comment.

After all prior steps are complete:

  • Gatekeeper requested (daos-gatekeeper added as a reviewer).

…ing object ULT creation

During migration, the system xstream only wakes up in ds_migrate_query_status()
(called by rebuild_tgt_status_check_ult()), where it sleeps for RBLD_CHECK_INTV (2 seconds).
This caused the system xstream to remain blocked for 2-second intervals without performing
any operations, severely degrading rebuild performance.

Benchmark results demonstrate the improvement:

Without patch: 99,780 objects processed in 120 seconds

With patch: 225,136 objects processed in 11 seconds

The patch achieves approximately 20x performance improvement

Signed-off-by: Wang Shilong <shilong.wang@hpe.com>
@wangshilong
wangshilong requested review from a team August 14, 2025 12:44
@github-actions

github-actions Bot commented Aug 14, 2025 •

Copy link
Copy Markdown

Ticket title is 'fixes excessive waiting time in system xstream during object ULT creation'
Status is 'In Review'
Job should run at elevated priority (1)
https://daosio.atlassian.net/browse/DAOS-17905

1 ms hurts performance on high speed network

Signed-off-by: Wang Shilong <shilong.wang@hpe.com>
@wangshilong
wangshilong force-pushed the shilongw/DAOS-17905-release2.6 branch from 835cfc2 to 574e26b Compare August 20, 2025 09:24
@github-actions github-actions Bot added the priority Ticket has high priority (automatically managed) label Sep 8, 2025
@daosbuild3

Copy link
Copy Markdown
Collaborator

@daosbuild3

Copy link
Copy Markdown
Collaborator

Test stage Functional Hardware Large MD on SSD completed with status FAILURE. https://jenkins-3.daos.hpc.amslabs.hpecorp.net//job/daos-stack/job/daos/view/change-requests/job/PR-16740/8/execution/node/1634/log

@daosbuild3

Copy link
Copy Markdown
Collaborator

@daosbuild3

Copy link
Copy Markdown
Collaborator

Test stage Functional Hardware Medium MD on SSD completed with status FAILURE. https://jenkins-3.daos.hpc.amslabs.hpecorp.net//job/daos-stack/job/daos/view/change-requests/job/PR-16740/9/execution/node/530/log

@wangshilong
wangshilong requested a review from a team September 15, 2025 23:31
@daltonbohning
daltonbohning merged commit edc9589 into release/2.6 Sep 15, 2025
26 of 27 checks passed
@daltonbohning
daltonbohning deleted the shilongw/DAOS-17905-release2.6 branch September 15, 2025 23:43
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

priority Ticket has high priority (automatically managed)

Development

Successfully merging this pull request may close these issues.

5 participants