zhangyue19921010 opened a new issue, #11341: URL: https://github.com/apache/paimon/issues/11341
### Search before asking - [x] I searched the issues and pull requests and found no matching report. ### Paimon version master, `ad4f6e09ef51c6b9932cbf22f0be563c6dde683f` (2.2-SNAPSHOT). ### Compute Engine Java catalog API on JDK 11, using a SQLite JDBC catalog. Flink and Spark rename DDL calls delegate to the same catalog method; no full engine job was used for this reproduction. ### Minimal reproduce step 1. Create two ordinary append tables in a JDBC catalog with `lock.enabled=false`. 2. Choose four distinct names A, B, C and D so A < B and C < D. A and D share one local lock stripe; B and C share another. 3. Run `renameTable(A, B, false)` and `renameTable(C, D, false)` concurrently. One operation can hold the first stripe while the other holds the second, and both wait for the remaining stripe. The [regression](https://github.com/zhangyue19921010/flink-table-store/blob/73d03eb203417a2f3aa400506057eba442d2baec/paimon-core/src/test/java/org/apache/paimon/jdbc/JdbcRenameLockOrderTest.java) selects colliding identifiers and coordinates real public catalog calls. On unchanged production code it fails with `Independent JDBC renames did not finish`; the same regression passes after the fix. ### What doesn't meet your expectations? Renaming different tables should finish. `runWithLocks` sorts identifier keys, but different key pairs can map to the same physical stripes in opposite order, causing a deadlock. ### Anything else? The proposed fix acquires local stripes in `Striped.bulkGet` order while retaining the existing identifier order for distributed locks. It also covers `renameView`, which uses the same helper. The new regression and neighboring catalog/view, lease and executor serialization tests pass (111 tests, normal Maven checks and RAT). ### Are you willing to submit a PR? - [x] Yes, I am willing to submit a PR. -- 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]
