Skip to content

Conversation

@guilload
Copy link
Member

No description provided.

@guilload guilload requested a review from fulmicoton December 11, 2025 16:52
);
}
shards_to_move
shards_to_rebalance
Copy link
Collaborator

Choose a reason for hiding this comment

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

This function returns the list of shards to move... It does not move stuff itself.
This is nice because it means any sophisticated logic downstream will be applied to place the shard on the right server.

HOWEVER
The code downstream does not enforce your new condition.

This means that as a set : this function + the function that attributes your stuff does not necessarily ever converges. I'm thinking in particular of the code that tries to place shards from the same index on the same indexer.

Copy link
Member Author

Choose a reason for hiding this comment

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

This logic only exists in the indexing controller. In the ingest controller, the placement logic is much simpler: we allocate shards to the ingesters with the least number of open shards. Right?

Copy link
Collaborator

Choose a reason for hiding this comment

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

Got it!

@guilload guilload force-pushed the guilload/improve-shard-rebalance-algorithm branch from d38cc07 to c0c4a8d Compare December 11, 2025 20:46
@guilload guilload force-pushed the guilload/improve-shard-rebalance-algorithm branch from c0c4a8d to 6489c80 Compare December 11, 2025 21:39
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants