chore: Cache files downloaded from cache#10950
Closed
spalladino wants to merge 2 commits intomasterfrom
Closed
Conversation
spalladino
force-pushed
the
palla/cache-cache
branch
from
9869790 to
e450137
Compare
To avoid a roundtrip to S3 to re-download every artifact on cache_download, we store a local file with the file prefix that points to the latest downloaded file version (ie hash), so we don't re-donwload it we already have it. This required changing the arity of the cache_download and cache_upload scripts, so they accept the key and version for each artifact separately. In addition, to avoid issues with the nargo `show-program-hash` being indeterministic, we also store "symlink" files in the cache for each circuit and vk. These are text files named after the entire hash of the circuits folder, that contain the nargo-computed hash of the circuit. This acts as another key for the same artifact.
spalladino
force-pushed
the
palla/cache-cache
branch
from
e450137 to
3646655
Compare
Contributor
Author
|
Downgraded to draft since it needs more testing, and #11030 seems like a better approach. |
Contributor
Author
|
Closing in favor of #11040 |
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
To avoid a roundtrip to S3 to re-download every artifact on cache_download, we store a local file with the file prefix that points to the latest downloaded file version (ie hash), so we don't re-donwload it we already have it.
This required changing the arity of the cache_download and cache_upload scripts, so they accept the key and version for each artifact separately.
In addition, to avoid issues with the nargo
show-program-hashbeing indeterministic, we also store "symlink" files in the cache for each circuit and vk. These are text files named after the entire hash of the circuits folder, that contain the nargo-computed hash of the circuit. This acts as another key for the same artifact.