Entity cache: query hashing should take requires and provides into account #4834
Assignees
Comments
6 tasks
Geal
added a commit
that referenced
this issue
Apr 4, 2024
Fix #4834 This extends the schema aware hashing already employed for subgraph queries in entity caching, to be calculated for client queries, and look at that hash in the query planner cache to be able to reuse cached entries across schema reloads. This contains: - an update of the traverse visitor to use an `ExecutableDocument` (necessary to parse field sets in `key` and `requires` argument - parses field sets in `@join__type`'s `key` argument and `@join__field`'s `requires` argument, because they can be affected by schema updates - remove the hack around `_entities` operation - parse subgraph queries using the subgraph schemas extracted from the supergraph schema - update query planner cache warm up to check if the query hash changed
Merged
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
This method of hashing considers directives, but only their presence and arguments. If we want to extend it outside of subgraph queries, it needs to consider how some arguments work, like
@requires, because it may introduce more fields in the query, and those fields could have changedThe text was updated successfully, but these errors were encountered: