Fall back from chunked pulls when the chunked metadata is too large #2230
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
Try to introduce at least _some_ structure and order to this huge object. Much more could probably be done. For now, the comments are aspirational, documenting methods that don't yet exist. That will be remedied soon. Only reorders existing fields, should not change behavior. Signed-off-by: Miloslav Trmač <[email protected]>
Instead of collecting all data in individual variables, and only creating a fully-formed chunkedDiffer at the very end, start with a newUninitializedDiffer with some values known from the start, and fill out other parts in chunkedDiffer.initialize... This adds some ambiguity in that we now have partially-filled chunkedDiffer objects, but that's counterbalanced by having to move around much fewer parameters and return values, and there will be more to improve in the future. Signed-off-by: Miloslav Trmač <[email protected]>
This centralizes 3 copies of the code into one, removing a parameter, and in one case an error path. Signed-off-by: Miloslav Trmač <[email protected]>
... to decrease the number of return values from the manifest read functions. Should not change behavior. Signed-off-by: Miloslav Trmač <[email protected]>
That's a logically better place, it pairs the getBlobAt calls with the ErrBadRequest types specific to those call sites. We will, also, add more fallback reasons. Signed-off-by: Miloslav Trmač <[email protected]>
Signed-off-by: Miloslav Trmač <[email protected]>
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: mtrmac The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
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.
This is a short-term fix for containers/podman#24885 : if the chunked metadata is too large, trigger fallback to the traditional non-chunked pull path instead of a hard failure. See the linked issue for previous discussion.
RFC: While working on this, the APIs of various functions started looking increasingly unwieldy, so I started with a refactoring of the way
chunkedDifferis initialized, and the wayread…Manifestreturn data. I think it’s a slight improvement, but it’s not at all an obvious improvement. I’d be happy to drop the initial refactor commits and to turn this into only the fallback logic.This is a short-term mitigation of the existing limits. It seems that there are various other (close and not that close) code paths that can similarly allocate a large amount of memory, with, at least… inconsistent enforcement of limits.
Marked draft because I didn’t test the code yet.
@giuseppe PTAL
See individual commit messages for details.