Fix performance regression by parallelizing alias checkout #1101
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This is necessary now that setting up aliases requires Git checkout and potentially even Git clone operations. This reduces the "Fetching components..." stage for a cluster with 60 component instances from ~12s to ~4s on my system (clones unchanged at ~3s, alias setup from ~9s to ~1s).
We change
fetch_components
to collect aliases and their backing components and use the sameThreadPoolExecutor
approach that we use for the component checkouts.We aggressively parallelize the alias checkouts for aliases which use a dependency URL that we've already cloned during the component checkout phase, and group aliases with the same not-yet-cloned dependency URL to avoid race conditions.
To make this work, we need to refactor the Component alias handling a bit, so that we have enough information to do the actual
alias_checkout()
in a parallelized loop.Checklist
bug
,enhancement
,documentation
,change
,breaking
,dependency
,internal
as they show up in the changelog