Skip to content
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.

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

Node operations fuzzy improvements #5360

Merged

Conversation

mmaslankaprv
Copy link
Member

Cover letter

Describe in plain language the motivation (bug, feature, etc.) behind the change in this PR and how the included commits address it.

Fixes #ISSUE-NUMBER, Fixes #ISSUE-NUMBER, ...

Release notes

@mmedenjak mmedenjak added the kind/enhance New feature or request label Jul 6, 2022
@mmaslankaprv mmaslankaprv marked this pull request as ready for review July 6, 2022 14:11
Copy link
Contributor

@andrwng andrwng left a comment

Choose a reason for hiding this comment

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

Overall this looks fine, though I wonder if it's meant to be two separate PRs: one that touches the fuzz test, and one that adjusts and tests the behavior of reducing the replication factor?

return true;
}
// update may be finished by a node that was added to replica set
if (!has_local_replicas(_self, previous_replicas)) {
Copy link
Contributor

Choose a reason for hiding this comment

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

nit: just looking at this method, this comment doesn't quite match up with the behavior. Could you add a note that it's expected that one of current_replicas is local?

@andrwng
Copy link
Contributor

andrwng commented Jul 7, 2022

Overall this looks fine, though I wonder if it's meant to be two separate PRs: one that touches the fuzz test, and one that adjusts and tests the behavior of reducing the replication factor?

Ah, I see there's another PR out for it. Sorry for the noise.

The fuzz test changes LGTM.

Fixed dispatching update_finished command when no replicas were added to
the replica set.

Previously when no replicas were added to the replica set there was no
node that would pass the condition allowing it to dispatch update
finished command. This way partition reconfiguration never finished
causing chaos-tests to fail.

Signed-off-by: Michal Maslanka <[email protected]>
Added test validating if we can shrink raft group replica set.

Signed-off-by: Michal Maslanka <[email protected]>
As we are already using admin operations fuzzer to execute admin day-2
operations we may skip creating/deleting topics explicitly

Signed-off-by: Michal Maslanka <[email protected]>
Added `add_no_wait` operation to nodes operations fuzzy test. The add no
wait does not wait for the node to be populated with partitions before
executing next operation, this way we are giving decommissioning a
change to be executed before addition will be finished therefore
triggering move cancellation code path.

Signed-off-by: Michal Maslanka <[email protected]>
Signed-off-by: Michal Maslanka <[email protected]>
@mmaslankaprv mmaslankaprv force-pushed the node-operations-fuzzy-improvements branch from eae5fc9 to 3795ed6 Compare July 7, 2022 08:30
Copy link
Contributor

@andrwng andrwng left a comment

Choose a reason for hiding this comment

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

Seeing as the other PR has been merged, the node op fuzz test improvements LGTM.

@mmaslankaprv mmaslankaprv merged commit f12a18f into redpanda-data:dev Jul 11, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

3 participants