waiProxy: don't strip Content-Length in response for HTTP/2#45
Merged
snoyberg merged 2 commits intofpco:masterfrom Dec 12, 2023
bjin:fix-cl
Merged
waiProxy: don't strip Content-Length in response for HTTP/2#45snoyberg merged 2 commits intofpco:masterfrom bjin:fix-cl
snoyberg merged 2 commits intofpco:masterfrom
bjin:fix-cl
Conversation
snoyberg
reviewed
Dec 12, 2023
Member
snoyberg
left a comment
There was a problem hiding this comment.
I don't have enough experience with proxying and HTTP/2-3 to approve this. @kazu-yamamoto any thoughts?
|
The explanation by @bjin is correct, I think. |
Contributor
Author
|
I would like to emphasize that this change is crucial for reverse proxying the Docker registry ( |
Member
|
In that case could you add a minor version bump and changelog entry for this? |
HTTP/2 (and newer HTTP/3) don't use chunked encoding so it's fine to preserve 'Content-Length' in proxied response header. Also do the same to HEAD requests and empty-body responses, because in both cases, chunked encoding won't be used.
Contributor
Author
Pushed a new commit bumping minor package version (and Changelog change), also edited commit message to reflect HEAD request changes. |
snoyberg
approved these changes
Dec 12, 2023
bjin
added a commit
to bjin/http-reverse-proxy
that referenced
this pull request
Jun 6, 2024
This is a proper fix to fpco#45. In previous merge request (fpco#45), for HTTP/2 connections, Content-Length header is retained in order to avoid problems with Docker client rejecting chunked encoding responses from Docker registry. However, this approach failed when the upstream response was encoded (e.g. with 'Content-Encoding: gzip' in header). The retained 'Content-Length' header then mismatched the actual content size after decoding. To fix this, we now strip the Content-Length header if Content-Encoding header is present, along with our existing checks for cases where chunked encoding is not used (HTTP/2 and HEAD requests).
bjin
added a commit
to bjin/http-reverse-proxy
that referenced
this pull request
Jun 6, 2024
This is a proper fix to fpco#45. In previous merge request (fpco#45), for HTTP/2 connections, Content-Length header is retained in order to avoid problems with Docker client rejecting chunked encoding responses from Docker registry. However, this approach failed when the upstream response was encoded (e.g. with 'Content-Encoding: gzip' in header). The retained 'Content-Length' header then mismatched the actual content size after decoding. To fix this, we now strip the Content-Length header if Content-Encoding header is present, along with our existing checks for cases where chunked encoding is not used (HTTP/2 and HEAD requests).
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
3 participants
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.
HTTP/2 (and newer HTTP/3) don't use chunked encoding so it's fine to keep Content-Length header in responses.
Also include cases for HEAD requests and empty-body responses, because in both cases, chunked encoding won't be used.