8000 run longevity test in insane mode · Issue #22016 · risingwavelabs/risingwave · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

run longevity test in insane mode #22016

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Open
stdrc opened this issue May 27, 2025 · 6 comments
Open

run longevity test in insane mode #22016

stdrc opened this issue May 27, 2025 · 6 comments
Labels
A-ci Area: CI workflow and scripts, etc. A-test Area: Testing framework & misc tests not belonging to any specific component. S-need-discussion Status: A discussion is needed before any further step. Typically used for feat/refactor issues. type/enhancement Type: Enhancement for an existing feature.

Comments

@stdrc
Copy link
Member
stdrc commented May 27, 2025

We introduced an "insane" mode in #15802 to test our streaming executors with inconsistent stream data, but failed to adopt it in our longevity test due to some bug. It's still good to have a Ci workflow to run test workload with this mode on.
@stdrc stdrc added type/enhancement Type: Enhancement for an existing feature. A-test Area: Testing framework & misc tests not belonging to any specific component. A-ci Area: CI workflow and scripts, etc. labels May 27, 2025
@github-actions github-actions bot added this to the release-2.5 milestone May 27, 2025
@stdrc stdrc removed this from the release-2.5 milestone May 27, 2025
@kwannoel
Copy link
Contributor
kwannoel commented May 27, 2025 •
edited
Loading

failed to adopt it in our longevity test due to some bug.

Have we resolved the bug

@stdrc
Copy link
Member Author
stdrc commented May 27, 2025

failed to adopt it in our longevity test due to some bug.

Have we resolved the bug

I guess no, it was a bug of insane mode when generating inconsistent workload.

@kwannoel
Copy link
Contributor

Given the inconsistency is tolerated to "best effort" I'm not sure there's a reliable way to do testing with insane mode.
Because if an error arises, perhaps the operator is not able to tolerate the inconsistency, and that's why it panics. And we may not want to tolerate inconsistency in some cases when it adds too much complexity to the executor.

@stdrc stdrc added the S-need-discussion Status: A discussion is needed before any further step. Typically used for feat/refactor issues. label May 27, 2025
@stdrc
Copy link
Member Author
stdrc commented May 27, 2025

And we may not want to tolerate inconsistency in some cases when it adds too much complexity to the executor.

Do you have an example?

@kwannoel
Copy link
Contributor
kwannoel commented May 27, 2025 •
edited
Loading

Example: #19629 (comment).

I removed the tolerating branch when optimizing cache refill for join.

@kwannoel
Copy link
Contributor

It's still good to have a Ci workflow to run test workload with this mode on.

Perhaps you can also share your motivation, on what coverage we want from this test.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
A-ci Area: CI workflow and scripts, etc. A-test Area: Testing framework & misc tests not belonging to any specific component. S-need-discussion Status: A discussion is needed before any further step. Typically used for feat/refactor issues. type/enhancement Type: Enhancement for an existing feature.
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@stdrc @kwannoel
0