Improve memory usage of the new backend #74740
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
sokra
changed the title
Stats from current PRDefault BuildGeneral
Client Bundles (main, webpack)
Legacy Client Bundles (polyfills)
Client Pages
Client Build Manifests
Rendered Page Sizes
Edge SSR bundle Size
Middleware size
Next Runtimes
build cache
Diff detailsDiff for main-HASH.jsDiff too large to display |
Tests Passed |
mischnic
reviewed
Jan 10, 2025
| InnerStorage::Indexed { map, .. } => map.get(&key.index()), | ||
| } | ||
| fn get_map(&self, ty: CachedDataItemType) -> Option<&CachedDataItemStorage> { | ||
| self.map.iter().find(|m| m.ty() == ty) |
There was a problem hiding this comment.
Technically this is still constant time, but it might be beneficial to not have to always iterate through the list here?
There was a problem hiding this comment.
We only have a maximum of 30 different types.
According to https://github.com/yegor256/micromap#benchmark using a Vec is faster. That's also the idea behind the AutoMap we have.
It's also using less memory
There was a problem hiding this comment.
@sokra As a micro-optimization, if we create the maps for the most common CachedDataItemTypes in InnerStorage::new, we can ensure they always end up at the start of the list.
However, this may naturally tend to happen anyways (the most frequently accessed types are likely to get initialized first).
There was a problem hiding this comment.
As follow-up PR I want to pull "common" types out of this vec and just give them a native field in InnerStorage. That would save about 240 bytes per task which is about 11% of total memory
mischnic
approved these changes
Jan 10, 2025
Merge activity
|
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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.
What?
Improves memory usage by changing the way data is stored in the new backend. Instead of keeping a map
enum CachedDataItemKey->enum CachedDataItemValue, it uses a Vecenum CachedDataItemStoragewhich keeps a single storage for everyenum CachedDataItemType. The storage is strongly types and e. g. stores a mapTaskId->(), which makes it more memory efficient while still not wasting memory for unused types.It also calls
shrink_to_fitafter task completion andshrink_amortizedafter removing items, to reduce the unnecessary capacity.Closes PACK-3703