Skip to content

Conversation

@spilchen
Copy link
Contributor

@spilchen spilchen commented Jan 7, 2026

Rename operations (e.g., sequences, tables) check for name conflicts at snapshot time but execute later, allowing concurrent transactions to change the destination state. This can cause expected errors (e.g., DuplicateRelation) to not occur, leading to false workload failures.

The workload now treats such cases as potential errors instead of hard failures.

Example:

TX1: Checks seq_w18_32 exists → TRUE
TX1: Expects error on renaming to seq_w18_32
TX2: Renames seq_w18_32 away → commits
TX1: Rename unexpectedly succeeds
- workload fails: error was expected but not raised

Closes #160465
Release note: none
Epic: none

Rename operations (e.g., sequences, tables) check for name conflicts at
snapshot time but execute later, allowing concurrent transactions to change
the destination state. This can cause expected errors (e.g., DuplicateRelation)
to not occur, leading to false workload failures.

The workload now treats such cases as potential errors instead of hard
failures.

Example:
```
TX1: Checks seq_w18_32 exists → TRUE
TX1: Expects error on renaming to seq_w18_32
TX2: Renames seq_w18_32 away → commits
TX1: Rename unexpectedly succeeds
- workload fails: error was expected but not raised
```

Closes cockroachdb#160465
Release note: none
Epic: none
@spilchen spilchen self-assigned this Jan 7, 2026
@cockroach-teamcity
Copy link
Member

This change is Reviewable

@spilchen spilchen marked this pull request as ready for review January 8, 2026 13:58
@spilchen spilchen requested a review from a team as a code owner January 8, 2026 13:58
@spilchen spilchen requested review from a team, herkolategan and shailendra-patel and removed request for a team January 8, 2026 13:58
Copy link
Collaborator

@fqazi fqazi left a comment

Choose a reason for hiding this comment

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

Thank you for tracking this one down @spilchen . I'll get a follow up issue open, since this is a leased descriptor limitation. I think we can get a retry to happen in this scenario, but it's not the more urgent thing, since its not the most realistic scenario.

:lgtm:

@fqazi made 1 comment.
Reviewable status: :shipit: complete! 1 of 0 LGTMs obtained (waiting on @herkolategan and @shailendra-patel).

Copy link
Collaborator

@fqazi fqazi left a comment

Choose a reason for hiding this comment

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

Opened: #160697

@fqazi made 1 comment.
Reviewable status: :shipit: complete! 1 of 0 LGTMs obtained (waiting on @herkolategan and @shailendra-patel).

@spilchen
Copy link
Contributor Author

spilchen commented Jan 8, 2026

TFTR!

bors r+

@craig
Copy link
Contributor

craig bot commented Jan 8, 2026

@craig craig bot merged commit f2613d7 into cockroachdb:master Jan 8, 2026
36 of 38 checks passed
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.

roachtest: schemachange/random-load failed

3 participants