Re: [PR] fix: correct XA connection hold fallback [incubator-seata-go]

["tsln1998 (via GitHub)" <gi...@apache.org>]

tsln1998 commented on PR #1150:
URL: https://github.com/apache/incubator-seata-go/pull/1150#issuecomment-5487735442

   I encountered the same issue while running Seata-Go XA integration tests with PostgreSQL.
   
     With seata-go v2.1.1-0.20260806094221-9992f1b7ae8f, race-enabled tests consistently reported a data race between the XA phase-two timeout checker and XAConn lifecycle operations. Regular test runs also occasionally panicked inside pgx with:
   
         BUG: slow write timer already active
   
     I also observed an occasional nil-pointer panic while XAConn.ExecContext was using a connection concurrently closed by the timeout checker. The business assertions often completed successfully before the process failed, indicating a physical
     connection ownership race rather than an application transaction error.
   
     After applying PR #1148, I could no longer reproduce these failures. Both race-enabled and repeated regular PostgreSQL integration tests passed, without concurrent-close panics or leaked prepared transactions.
   
     The synchronized connection ownership, state-aware timeout cleanup, and duplicate-close protection introduced by this PR address the root cause in my environment.
   
     I can confirm that PR #1148 is effective for the `PostgreSQL XA` path as well.


-- 
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: notifications-unsubscribe@seata.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: notifications-unsubscribe@seata.apache.org
For additional commands, e-mail: notifications-help@seata.apache.org