imbajin commented on PR #710: URL: https://github.com/apache/hugegraph-toolchain/pull/710#issuecomment-6432825778
Closing in favor of #858, which takes over this contribution from current master and is ready for review. The related issue #706 remains tracked by that replacement until it is merged. The reproduced HDFS NPE comes from scheduling each direct file twice: one parser task completes and removes the file checkpoint while the duplicate still updates it. #858 fixes directory traversal and the separate shared-regex-matcher race, rather than treating an unexpected null reader as end-of-file. It also restores four parser threads in the previously restricted regression cases and carries forward the two HDFS concurrency cases proposed here. The final replacement passes the full Loader/HDFS CI and the Spark, Flink, and Hubble integrations: https://github.com/apache/hugegraph-toolchain/actions/runs/40438801720. The affected cases also pass locally against the locked candidate Server and real Hadoop 3.3.6, with the original vertex-count assertions. Thank you for the original contribution and investigation. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
