update triton to 21.08 (allowing for content-endocded payloads) #3589
Conversation
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
|
/test integration |
|
/test notebooks |
|
/test integration |
|
Tests passed in the CI!
|
|
This is really good - overall lgtm, I am wondering whether there's a way to document this so it doesn't catch us off-guard if we're someone later on looks at the changes and doesn't know why they are needed. Maybe just adding a comment with a link to this PR for context - thinking it would be mainly this line https://github.com/SeldonIO/seldon-core/pull/3589/files#diff-7b20b52f5ddc2d09eaf37f69a2a8efde655f48cc1ea0d36cd9df39af9d31bcf5R67 |
add in-code comment
|
Added comment, linked PR and additional documentation and also extended PR description. |
|
Nice one @RafalSkolasinski - lgtm! |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: axsaucedo 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 |
…onIO#3589) * update triton to 21.08 * experiment with content-encoding * executor: escape html only when no encoding is present * undo force no compression * remove forgotten fmt.Println * Update client.go add in-code comment * typo fix * fix whitespace
What this PR does / why we need it:
This updates Triton image from
20.08to21.08. As new version of Triton Server respectsAccept-Encodingheader that by default from pythonrequestsreads 'Accept-Encoding': 'gzip, deflate' the response may now bygzipencoded.For this to be correctly the
Content-Encodingheader must be present.Additionally, this PR adds change to not escape HTML characters in executor if content is encoded.
Which issue(s) this PR fixes:
Fixes #3318
Special notes for your reviewer:
Does this PR introduce a user-facing change?: