chore: increase fileCache hit rate in autobuilds lifecycle #18507
Conversation
This file contains hidden or 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
This was referenced Jun 23, 2025
This was referenced Jun 23, 2025
This stack of pull requests is managed by Graphite. Learn more about stacking. |
Emyrk
commented
Jun 23, 2025
coderd/files/cache.go
Outdated
Comment on lines
78
to
79
| Help: "The total number of file references ever opened in the file cache. The label 'hit' indicates whether the file was already in the cache ('true') or not ('false').", | ||
| }, []string{"hit"}) |
There was a problem hiding this comment.
Added a label for detecting our cache hit rate
4f52076 to
c1d05a0
Compare
d73e843 to
ae40278
Compare
ae40278 to
a4c0df4
Compare
c1d05a0 to
129235b
Compare
a4c0df4 to
a97fd53
Compare
129235b to
82af2e0
Compare
a97fd53 to
9945049
Compare
9945049 to
e5a33c1
Compare
e5a33c1 to
4f6244f
Compare
aslilac
approved these changes
Jun 23, 2025
Comment on lines
+164
to
+171
| // `wsbuilder` needs to load the terraform files for a given template version | ||
| // into memory. If 2 workspaces are using the same template version, they will | ||
| // share the same files in the FileCache. This only happens if the builds happen | ||
| // in parallel. | ||
| // TODO: Actually make sure the cache has the files in the cache for the full | ||
| // set of identical template versions. Then unload the files when the builds | ||
| // are done. Right now, this relies on luck for the 10 goroutine workers to | ||
| // overlap and keep the file reference in the cache alive. |
There was a problem hiding this comment.
yeah, it could be nice if we did more of a "mark and sweep" style release strategy. or maybe generational...
oh no, am I about to build a general purpose garbage collector for this? Oh No
There was a problem hiding this comment.
Yea, I had a similar set of thoughts and decided this was a good enough idea for this release.
There is a better way to do this for sure, and I wonder where else in the codebase we will have similar things going on (prebuilds?). That being a large set of workspaces that have some work to be done in bulk.
4f6244f to
bc38b64
Compare
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.
wsbuilderhits the file cache when running validation. This solution is imperfect, but by first sorting workspaces by their template version id, the cache hit rate should improve.If there is a large number of workspaces, this could save some memory impact of dynamic parameters.
Future Work
A lot of builds have overlapping queries to fetch template information:
tx.GetTemplateByIDGetTemplateVersionByIDWe could also cache these between the go routines running the builds to reduce some db load.