SEZ9 commented on PR #12299: URL: https://github.com/apache/seatunnel/pull/12299#issuecomment-5852160997
Checked the actual current diff at `38d0f805c111` on `dev@025d08f9bcb0`: - **F1/F3/F4:** [FileUtils.readFileTail](https://github.com/apache/seatunnel/blob/38d0f805c111c2a22df20507f9db2d62088d9e92/seatunnel-common/src/main/java/org/apache/seatunnel/common/utils/FileUtils.java#L124) returns `FileTail` containing the retained window, size snapshot and truncated flag. Positive-limit reads use one channel/size snapshot. [LogContentReader.read](https://github.com/apache/seatunnel/blob/38d0f805c111c2a22df20507f9db2d62088d9e92/seatunnel-engine/seatunnel-engine-server/src/main/java/org/apache/seatunnel/engine/server/rest/LogContentReader.java#L64) consumes that result without a second `Files.size(path)`. Both full-window and truncated decoding explicitly use UTF-8. The [v2 servlet](https://github.com/apache/seatunnel/blob/38d0f805c111c2a22df20507f9db2d62088d9e92/seatunnel-engine/seatunnel-engine-server/src/main/java/org/apache/seatunnel/engine/server/rest/servlet/LogBaseServlet.java#L74) and [v1 handler](https://github.com/apache/seatunnel/blob/38d0f805c111c2a2 2df20507f9db2d62088d9e92/seatunnel-engine/seatunnel-engine-server/src/main/java/org/apache/seatunnel/engine/server/rest/RestHttpGetCommandProcessor.java#L418) call this same reader, including its prefix. I added a complete notice example to the PR description. - **F2/F6:** Both compatibility entries are in this head: [en](https://github.com/apache/seatunnel/blob/38d0f805c111c2a22df20507f9db2d62088d9e92/docs/en/introduction/concepts/incompatible-changes.md#L349) / [zh](https://github.com/apache/seatunnel/blob/38d0f805c111c2a22df20507f9db2d62088d9e92/docs/zh/introduction/concepts/incompatible-changes.md#L310). So are the v1 Response Size Limit sections: [en](https://github.com/apache/seatunnel/blob/38d0f805c111c2a22df20507f9db2d62088d9e92/docs/en/engines/zeta/rest-api-v1.md#L1010) / [zh](https://github.com/apache/seatunnel/blob/38d0f805c111c2a22df20507f9db2d62088d9e92/docs/zh/engines/zeta/rest-api-v1.md#L1012). No separate unmerged documentation commit is required. - **F7:** Both v2 pages describe the configured limit with 64 MB as its default, and the stray Chinese double backtick is removed: [en](https://github.com/apache/seatunnel/blob/38d0f805c111c2a22df20507f9db2d62088d9e92/docs/en/engines/zeta/rest-api-v2.md#L1456) / [zh](https://github.com/apache/seatunnel/blob/38d0f805c111c2a22df20507f9db2d62088d9e92/docs/zh/engines/zeta/rest-api-v2.md#L1433). - **F8/F5:** [HttpConfig.getLogResponseMaxSizeBytes](https://github.com/apache/seatunnel/blob/38d0f805c111c2a22df20507f9db2d62088d9e92/seatunnel-engine/seatunnel-engine-common/src/main/java/org/apache/seatunnel/engine/common/config/server/HttpConfig.java#L100) is the sole production conversion for this option; both handlers delegate to it. `FileTail.getContentWithPrefix` decodes the retained window directly into the prefixed response builder, without an intermediate full content String. I have no measured peak-heap number and do not claim one; the description explicitly distinguishes file-content bytes from total request allocations. Current-head local Spotless/full-reactor verify, 25 unit tests and the real v1/v2 HTTP limit integration test passed. Remote CI is still running on attempt 5, retaining 80 successful jobs. Direct failure evidence: [OceanBase/Java 8](https://github.com/SEZ9/seatunnel/actions/runs/36242600513/job/108529136135) / [Java 11](https://github.com/SEZ9/seatunnel/actions/runs/36242600513/job/108529136166), both `NotSerializableException: io.debezium.relational.TableId`; independently fixed and runtime-validated in #12489. [Java 11 engine tests](https://github.com/SEZ9/seatunnel/actions/runs/36242600513/job/108529136173) are the active retry; the [previous engine attempt](https://github.com/SEZ9/seatunnel/actions/runs/36242600513/job/108516960990) failed `testStreamJobFailsAfterCheckpointTriggerDispatchFailure`, covered by the checkpoint guard under review in #12454. Neither shared patch is included in this feature diff, and none of these partial results is being presented as a green Build. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
