Uh oh!
There was an error while loading. Please reload this page.
[v3-3-test] Run lang-SDK K8s system test once in a dedicated job, not 6x per run (#69500) - #69587
Merged
Merged
Conversation
… 6x per run (apache#69500) The Multi-Lang KubernetesExecutor coordinator system test was gated on KubernetesExecutor + use-standard-naming==false, but that condition is true once per kubernetes-combo, so it ran redundantly across all six Python/K8s combos on ci-amd (and again on ci-arm) - adding ~4 minutes of Go/Java build and coordinator run to each of those six system-test jobs and pushing them to ~20 minutes, well above the rest of the matrix. The test is executor/naming/version-independent, so run it once instead: a dedicated tests-kubernetes-lang-sdk job on a single default Python-Kubernetes combo that provisions the native Go/Java toolchains (cached) and runs only the lang-SDK test (-k test_lang_sdk_combined_dag_succeeds), not the full suite or the upgrade path. The regular system-test matrix no longer runs it at all, so those six jobs drop back to ~15 minutes. Also scope the Go/Java build caches by runner.arch: the previous key used only runner.os (Linux), which would collide between amd64 and arm64 runners now that the dedicated job runs on both. (cherry picked from commit 586e658) Co-authored-by: Jarek Potiuk <jarek@potiuk.com>
jason810496
requested review from
amoghrajesh, ashb, bugraoz93, gopidesupavan, jscheffl and potiuk
as code ownersJuly 8, 2026 05:15
potiuk
approved these changes
Jul 8, 2026
Uh oh!
There was an error while loading. Please reload this page.
vatsrahul1001 pushed a commit
that referenced
this pull request
Jul 9, 2026
… 6x per run (#69500) (#69587) The Multi-Lang KubernetesExecutor coordinator system test was gated on KubernetesExecutor + use-standard-naming==false, but that condition is true once per kubernetes-combo, so it ran redundantly across all six Python/K8s combos on ci-amd (and again on ci-arm) - adding ~4 minutes of Go/Java build and coordinator run to each of those six system-test jobs and pushing them to ~20 minutes, well above the rest of the matrix. The test is executor/naming/version-independent, so run it once instead: a dedicated tests-kubernetes-lang-sdk job on a single default Python-Kubernetes combo that provisions the native Go/Java toolchains (cached) and runs only the lang-SDK test (-k test_lang_sdk_combined_dag_succeeds), not the full suite or the upgrade path. The regular system-test matrix no longer runs it at all, so those six jobs drop back to ~15 minutes. Also scope the Go/Java build caches by runner.arch: the previous key used only runner.os (Linux), which would collide between amd64 and arm64 runners now that the dedicated job runs on both. (cherry picked from commit 586e658) Co-authored-by: Jarek Potiuk <jarek@potiuk.com>
vatsrahul1001 pushed a commit
that referenced
this pull request
Aug 5, 2026
… 6x per run (#69500) (#69587) The Multi-Lang KubernetesExecutor coordinator system test was gated on KubernetesExecutor + use-standard-naming==false, but that condition is true once per kubernetes-combo, so it ran redundantly across all six Python/K8s combos on ci-amd (and again on ci-arm) - adding ~4 minutes of Go/Java build and coordinator run to each of those six system-test jobs and pushing them to ~20 minutes, well above the rest of the matrix. The test is executor/naming/version-independent, so run it once instead: a dedicated tests-kubernetes-lang-sdk job on a single default Python-Kubernetes combo that provisions the native Go/Java toolchains (cached) and runs only the lang-SDK test (-k test_lang_sdk_combined_dag_succeeds), not the full suite or the upgrade path. The regular system-test matrix no longer runs it at all, so those six jobs drop back to ~15 minutes. Also scope the Go/Java build caches by runner.arch: the previous key used only runner.os (Linux), which would collide between amd64 and arm64 runners now that the dedicated job runs on both. (cherry picked from commit 586e658) Co-authored-by: Jarek Potiuk <jarek@potiuk.com>
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 freeto 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.
The Multi-Lang KubernetesExecutor coordinator system test was gated on KubernetesExecutor + use-standard-naming==false, but that condition is true once per kubernetes-combo, so it ran redundantly across all six Python/K8s combos on ci-amd (and again on ci-arm) - adding ~4 minutes of Go/Java build and coordinator run to each of those six system-test jobs and pushing them to ~20 minutes, well above the rest of the matrix.
The test is executor/naming/version-independent, so run it once instead: a dedicated tests-kubernetes-lang-sdk job on a single default Python-Kubernetes combo that provisions the native Go/Java toolchains (cached) and runs only the lang-SDK test (-k test_lang_sdk_combined_dag_succeeds), not the full suite or the upgrade path. The regular system-test matrix no longer runs it at all, so those six jobs drop back to ~15 minutes.
Also scope the Go/Java build caches by runner.arch: the previous key used only runner.os (Linux), which would collide between amd64 and arm64 runners now that the dedicated job runs on both.
(cherry picked from commit 586e658)