Skip to content

Conversation

@bryanlb
Copy link
Contributor

@bryanlb bryanlb commented Aug 30, 2023

Summary

We currently do not block on cache initialization, which can cause downstream consumers to read incomplete/empty cache results. This now enforces the cache is completely hydrated before we allow any calls that rely on the cache to complete.

Resolves #651, closes #652

@bryanlb bryanlb force-pushed the bburkholder/zk-cache-init branch from 8f52625 to b33c13d Compare August 31, 2023 17:47
@bryanlb bryanlb marked this pull request as ready for review August 31, 2023 17:58
@bryanlb bryanlb merged commit 8d8024c into master Aug 31, 2023
@bryanlb bryanlb deleted the bburkholder/zk-cache-init branch August 31, 2023 18:33
bryanlb added a commit that referenced this pull request Sep 1, 2023
bryanlb added a commit that referenced this pull request Sep 1, 2023
bryanlb added a commit that referenced this pull request Sep 1, 2023
bryanlb added a commit that referenced this pull request Sep 5, 2023
bryanlb added a commit that referenced this pull request Sep 5, 2023
bryanlb added a commit that referenced this pull request Sep 5, 2023
* Revert "Revert "Add latch for ZK hydration initialization (#655)" (#664)"

This reverts commit 451c305.

* Change distributed query snapshot initialization

Reduces churn due to ZK updates, separate executor to prevent potential deadlock

---------

Co-authored-by: Bryan Burkholder <bryanlb@users.noreply.github.com>
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.

Indexer couldn't find all previous metadata snapshots

3 participants