Use LMDB cache instead of regular FS cache to speed up cache hits #932
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
pastelsky
force-pushed
the
sk/use-lmdb-cache
branch
16 times, most recently
from
4d3d769
to
1ef92e1
Compare
pastelsky
force-pushed
the
sk/use-lmdb-cache
branch
from
1ef92e1
to
068be4b
Compare
11 tasks
|
Pre-requisite for this PR — |
11 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Please check if the PR fulfills these requirements
What kind of change does this PR introduce? (Bug fix, feature, docs update, ...)
Improved performance of babel-loader when files are cached. LMDB is much faster to access than the regular file cache used by babel-loader.
In the large codebase I work with, warm boot times on webpack 4 improved from — 79 seconds to 74 seconds.
What is the current behavior? (You can also link to an open issue here)
babel-loadersaves files for caching built contents.What is the new behavior?
Replaces the use of individual files with lmdb, which improves babel-loader times, when hitting caches in large projects.
LMDB is already used by projects such as Parcel and Elasticsearch's Kibana, as the storage layer for HarperDB and Gatsby for fast memory-mapped cached access.
This has a side-effect of making babel-loader's implementation a bit simpler as well.
Does this PR introduce a breaking change?
Provided Change minimum supported node to 12 #933 is merged first.