Uh oh!
There was an error while loading. Please reload this page.
Fix review findings: integration/s3-sink (hy-dq5u) - #15
Merged
Conversation
Address the major review finding on PR #7: the s3 sink's `exportBatch` terminal-failure return (`s3_credentials_missing`, `s3_access_denied`, `s3_bucket_missing`, `s3_region_mismatch`, `s3_config_invalid`) omitted `retryPartitions`. The sink driver at `src/core/sinks/driver.js:146` falls back to the entire batch in that case, so a terminal error landing on partition N caused the driver to outbox every partition in the batch — including the N-1 partitions that had already been PUT to S3. Subsequent retries would then double-export those objects. Return `retryPartitions: failures` so the driver outboxes only the partitions that actually failed. `failures` always has at least one entry when `lastConfigFailure` is set because both are pushed in the same catch block. Adds `test/plugins/s3-export-batch.test.js` covering the terminal, partial, and all-success exit paths, exercised through the real `activate` registration path with an injected `__clientFactory`. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Uh oh!
There was an error while loading. Please reload this page.
philcunliffe added a commit
that referenced
this pull request
May 22, 2026
* chore: seed integration/s3-sink for feature flow * feat(s3): add @hypaware/s3 blob destination plugin (hy-myln) (#10) Implements hy-s3-1: a bundled first-party plugin providing hypaware.blob-store@1.0.0 and contributing the `s3` sink. Plugin tree under hypaware-core/plugins-workspace/s3/: - manifest declares network + read_env permissions and the s3 sink - src/config.js: validates bucket/prefix/region/profile/storage_class/ server_side_encryption + endpoint_url/force_path_style; emits stable s3_config_invalid errors - src/keys.js: renders <prefix>/<dataset>/<segment>/<filename> matching local-fs layout; normalizes prefix slashes; rejects keys outside the configured prefix - src/client.js: AWS SDK v3 wiring with injectable client factory and a credential_source_kind detector that never leaks secret material - src/errors.js: AWS SDK error -> stable error_kind mapping (s3_credentials_missing, s3_access_denied, s3_bucket_missing, s3_region_mismatch, s3_throttled, s3_put_failed, encoder_failed) - src/index.js: sink composes ctx.query schema + ctx.storage rows + the paired encoder; emits s3.client.init / s3.put_object / s3.put_object.failed telemetry; closes terminal errors as status=failed and surfaces transient failures as status=partial Bundled wiring + dependencies: - @hypaware/s3 added to V1_BUNDLED_PLUGIN_ALLOWLIST - @aws-sdk/client-s3 + @aws-sdk/credential-provider-ini added to root dependencies; npm pack --dry-run ships the new plugin tree Tests (test/plugins/s3-*.test.js): - s3-keys: key composition, prefix normalization, path-separator stripping, within-prefix guard - s3-config: required/optional fields, storage_class allowlist, endpoint_url URL parsing, S3-compatible (MinIO) shape - s3-client: credential_source_kind precedence + redaction contract - s3-errors: full AWS SDK error -> stable kind mapping table Hermetic smoke (s3_sink_export_fixture): - Activates real format-parquet + s3 plugin trees with a fake injected S3 client; fires one driver tick; asserts PutObject bucket/key/body, decodes the body back into Parquet rows, and verifies the s3.client.init / s3.put_object / sink.encode_partition telemetry carries the expected attributes with no credential-shaped substring Smoke fix (cli_bundled_plugins_activated): - The allowlist now has 8 entries; the smoke selects 6, so plugins_skipped is 2 and both format-jsonl and s3 emit plugin.skipped logs Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com> * fix(s3): return retryPartitions on terminal failure (hy-dq5u) (#15) Address the major review finding on PR #7: the s3 sink's `exportBatch` terminal-failure return (`s3_credentials_missing`, `s3_access_denied`, `s3_bucket_missing`, `s3_region_mismatch`, `s3_config_invalid`) omitted `retryPartitions`. The sink driver at `src/core/sinks/driver.js:146` falls back to the entire batch in that case, so a terminal error landing on partition N caused the driver to outbox every partition in the batch — including the N-1 partitions that had already been PUT to S3. Subsequent retries would then double-export those objects. Return `retryPartitions: failures` so the driver outboxes only the partitions that actually failed. `failures` always has at least one entry when `lastConfigFailure` is set because both are pushed in the same catch block. Adds `test/plugins/s3-export-batch.test.js` covering the terminal, partial, and all-success exit paths, exercised through the real `activate` registration path with an injected `__clientFactory`. Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com> --------- Co-authored-by: feature-launch <feature-launch@gas.city> Co-authored-by: Claude Opus 4.7 (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 freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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.
Summary
Fix review findings: integration/s3-sink
Implemented: s3 exportBatch terminal-failure path now returns retryPartitions=failures so driver outbox excludes already-uploaded partitions. Added test/plugins/s3-export-batch.test.js covering terminal, partial, and all-success paths.
Delivery
This PR is auto-merged into the integration branch once CI is green. Human
review for the full feature happens on the downstream PR
(
integration/s3-sink -> master, PR #7), which is opened separately when thefeature is ready to ship.