chore(chart): appVersion 0.2.5, chart 0.3.5 (elasticsql 0.20.4 train) - #20
Merged
Conversation
Last of the six-repo train: core 0.20.4 -> extensions 0.2.4 ->
arrow 0.2.5 -> jdbc 0.2.5 -> repl 0.20.4 -> THIS.
The chart deploys the federation server and its per-ES-version sidecars,
so it inherits the whole 0.20.4 train through the images:
elasticsql 0.20.4 the completeness family — GROUP BY (#205), window
PARTITION BY (#207), un-LIMITed SELECT (#209) and
an explicit LIMIT above index.max_result_window
(#224) no longer truncate silently; PIT paging
sorts by _doc and gates at 7.12 (#197); the WHOLE
statement must parse (#213)
extensions 0.2.4 materialized-view error integrity — a cluster
outage is no longer reported as 404 "does not
exist" (#45), and a metadata table that cannot be
created is no longer a 402 quota breach (#46)
arrow 0.2.5 the JOIN ceilings that killed the sidecar rather
than degrading: legs past exactly 4,032 rows
(#143), lazily streamed bounded legs (#147), a
writable and BOUNDED DuckDB spill directory
(#141), a real health probe that sees allocator
degradation (#146), and a default logback.xml so
the images stop booting at ROOT DEBUG (#138 — a
data-exposure fix as much as a performance one)
ALSO RECONCILES THE THIRD SURFACE: .github/workflows/federation-helm.yml
IMAGE_TAG was still '0.2.3' — the 0.3.4 hop bumped the chart and left the
workflow behind, exactly the drift the 0.2.3 hop had just fixed. The live
jobs pull and kind-load that tag independently of the chart, so CI was
loading 0.2.3 images while the goldens rendered 0.2.4. Now '0.2.5'.
All seven goldens regenerated. Checked for pre-existing drift FIRST (none
— all seven rendered byte-identical before the bump), so the diff is only
what appVersion drives: image tags, `helm.sh/chart`,
`app.kubernetes.io/version`, and the ConfigMap checksums that follow.
🔴 DO NOT MERGE UNTIL THE IMAGES EXIST. Probed at authoring time, all five
tags answer 404 on DockerHub:
softnetwork/softclient4es-federation:0.2.5
softnetwork/softclient4es{6,7,8,9}-arrow-flight-sql:0.2.5
appVersion is the DEFAULT image tag, so merging early ships a chart whose
pods cannot pull. Re-probe all five after the arrow 0.2.5 images are built.
Verified locally, the full static gate: helm lint clean on the chart and
all three examples; kubeconform -strict clean on base + three examples
(3/7/11/11 resources, 0 invalid); all seven goldens re-render
byte-identical; zero kind:Secret; the A2b discriminator holds (3 commented
duckdb-attach previews in heterogeneous-ready, 0 in three-region) and the
two example goldens stay byte-identical to each other.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.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 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.
Last of the six-repo train:
core 0.20.4→extensions 0.2.4→arrow 0.2.5→jdbc 0.2.5→repl 0.20.4→ this.Chart.yamlversion0.3.40.3.5Chart.yamlappVersion0.2.40.2.5federation-helm.ymlIMAGE_TAG0.2.30.2.5What the images inherit
GROUP BY(#205), windowPARTITION BY(#207), un-LIMITedSELECT(#209) and an explicitLIMITaboveindex.max_result_window(#224) no longer truncate silently; PIT paging sorts by_docand gates at 7.12 (#197); the whole statement must parse (#213)404 "does not exist"(#45), and a metadata table that cannot be created is no longer a402quota breach (#46)logback.xmlso the images stop booting at ROOT DEBUG (#138 — a data-exposure fix as much as a performance one).github/workflows/federation-helm.ymlIMAGE_TAGwas still'0.2.3'— the 0.3.4 hop bumped the chart and left the workflow behind, exactly the drift the 0.2.3 hop had just fixed. The live-install jobsdocker pullandkind loadthat tag independently of the chart, so CI was loading 0.2.3 images while the goldens rendered 0.2.4. Its own comment saysMUST match Chart.yaml appVersion.Goldens
All seven regenerated. Checked for pre-existing drift first (none — all seven rendered byte-identical before the bump), so the diff is only what appVersion drives: image tags,
helm.sh/chart,app.kubernetes.io/version, and the ConfigMap checksums that follow from them.✅ Image gate — cleared
appVersionis the default image tag, so this chart is only mergeable once the images exist. Re-probed at push time, all five answer HTTP 200 on DockerHub:(They were 404 when the commit was authored — the commit message records that state and is now superseded.)
Verified locally — the full static gate
helm lintclean on the chart and all three exampleskubeconform -strictclean on base + three examples (3 / 7 / 11 / 11 resources, 0 invalid)kind: Secretduckdb-attachpreviews inheterogeneous-ready, 0 inthree-region, and the two example goldens stay byte-identical to each other🤖 Generated with Claude Code