GeezerVisonThe World Wide Web, restored to 800 pixels of glory. | |||||||||||||||||||||||||||||
| Home|About|Gaming cabinet|Archives|How it works|Starter sites | |||||||||||||||||||||||||||||
GeezerVison is showing github.com — Integrate upstream changes from develop by carole-lavillonniere · Pull Request #92 · localstack/lambda-runtime-init Conversation
|
| package invoke | ||
|
|
||
| const ( | ||
| RequestIdHeader = "X-Amzn-RequestId" |
There was a problem hiding this comment.
The reason will be displayed to describe this comment to others. Learn more.
FYI: Interesting to see RequestId supported natively via headers now, which differs from our current approach sending it via the LS invoke payload.
Sorry, something went wrong.
There was an error while loading. Please reload this page.
| const ( | ||
| RequestIdHeader = "X-Amzn-RequestId" | ||
|
|
||
| ClientContextHeader = "X-Amz-Client-Context" |
There was a problem hiding this comment.
The reason will be displayed to describe this comment to others. Learn more.
Sorry, something went wrong.
There was an error while loading. Please reload this page.
|
Closing this PR, given the changes are already merged manually and released as v0.1.44-pre, following our guidance https://github.com/localstack/lambda-runtime-init/blob/localstack/README-LOCALSTACK.md#integrate-upstream-changes |
Sorry, something went wrong.
There was an error while loading. Please reload this page.
I created the release 1.44 just to test in localstack-core/pro but the changes are not merged yet, reopening the PR. |
Sorry, something went wrong.
There was an error while loading. Please reload this page.
I pushed this branch so we can visualize all the conflicts that were resolved: a2cf589 My understanding is that these conflicts happened because PR #62 did cherry-picks instead of a merge. We ended up with the same commits but different commit hashes and both sides created the same files independently. |
Sorry, something went wrong.
There was an error while loading. Please reload this page.
|
Thank you very much for the clarification and full merge conflict transparency 👏👏
I'm looking forward to that 🚀 |
Sorry, something went wrong.
There was an error while loading. Please reload this page.
There was a problem hiding this comment.
The reason will be displayed to describe this comment to others. Learn more.
Looks good from my side! Thanks for resolving the conflicts!
Sorry, something went wrong.
There was an error while loading. Please reload this page.
|
Some tests in LocalStack were failing with this version, not sure why. I started a clean merge here #98, it seems to work but there are some file differences. I'll review them. |
Sorry, something went wrong.
There was an error while loading. Please reload this page.
There was an error while loading. Please reload this page.
Brings the fork up to aws/aws-lambda-runtime-interface-emulator@3ccddd7 (Update from upstream - 2026-09-17), covering four upstream syncs: 3ccddd7 Update from upstream - 2026-09-17 (aws#189) c622c06 Update from upstream - 2026-08-25 (aws#187) ddf2bc4 Update from upstream - 2026-07-31 (aws#182) a0f264a chore: sync from internal (rie-private-to-public) (aws#181) The fork had been on develop@391c3f1 since April, and a0f264a is why this matters now: it makes raptor.Server.Shutdown drain in-flight responses via httpServer.Shutdown instead of calling httpServer.Close(), which severed connections whose handler was still writing. That race is what failed the weekly release twice -- as `Post ...: EOF` in TestRie_InvokeFatalError, and as a 10m package timeout in TestRie_InvokeWaitingForInitError, where the invoke goroutine's require.NoError skipped its un-deferred wg.Done() and deadlocked the package. Both tests are green on this merge: 0 failures in 40 runs of the package under -race, against 10/10 hangs before it. Conflicts, all in files we customize: go.mod Kept our go 1.27.1 directive over upstream's 1.26.6. Makefile Kept our GO_VERSION derived from go.mod, and the nojsonv2 GOEXPERIMENT export. check-binaries.yml Took upstream's switch from robinraju/release-downloader release.yml to `gh release download`, and from softprops/action-gh-release to `gh release create`; kept our Renovate action bumps (checkout/setup-python v7). release.yml's trailing context had already moved to upstream's shape, so keeping ours there would not have parsed. integ-tests.yml Version bumps only; kept ours. Note for the next integration: the previous one (#92) landed as a single squashed commit, so its upstream parent was lost and git computed the merge base as the Go 1.24 bump from 2025 -- 48 conflicts, 36 of them spurious add/add on files that were already identical. This was resolved against the true base (391c3f1) using a local `git replace --graft`, giving the 5 real conflicts above. Merging this one with --ff, per README-LOCALSTACK step 7, records the upstream parent and stops the next integration needing that.
There was an error while loading. Please reload this page.
Successfully merging this pull request may close these issues.
There was an error while loading. Please reload this page.