Bug #39809 falcon.falcon_bug_29332 fails sporadically on Windows

• Submitted: 2 Oct 2008 13:26
• Modified: 14 Apr 2009 7:00
• Reporter: Alexander Nozdrin
• Status: Can't repeat
• Category: MySQL Server: Falcon storage engine
• Severity: S3 (Non-critical)
• Version: 6.0-TRUNK
• OS: Windows (vm-win2003-32-a)
• Assigned to: Hakan Küçükyılmaz
• CPU Architecture: Any
• Tags: F_TEST, pushbuild, sporadic

[2 Oct 2008 13:26] Alexander Nozdrin

Description:
Symptoms:
------------------------------------------------------
@@ -43,13 +43,15 @@
 9
 # Switch to connection conn1
 UPDATE t1 SET a = 88 WHERE a = 8;
+a
+8
 UPDATE t1 SET a = 99 WHERE a = 9;
 SELECT * FROM t1 ORDER BY a;
+COMMIT;
 a
 2
 88
 99
-COMMIT;
 # Switch to connection default
 COMMIT;
 ------- Test 2 - with partitions ------
------------------------------------------------------

The bug reported after a failure on Wed Oct 1 13:08:25 2008 in 6.0-rpl-merge:
http://tinyurl.com/5yccgb

The test started to fail on 2008-07-09.

How to repeat:
XRef: http://tinyurl.com/48ap55

[4 Oct 2008 4:55] Kevin Lewis

Hakan, This looks like a test case issue with bad synchronization between threads in MTR.  Can a strategic --real_sleep 1 help with this?

[13 Mar 2009 16:18] Kevin Lewis

Hakan, what is the status of this?

[13 Mar 2009 18:57] Hakan Küçükyılmaz

I checked the test case already, but I could not find anything which could lead to a failure, nor I could reproduce the failure.

Our last failure is from February 26th. I would like to wait another month and revisit Xref to see whether this test case still fails.

[14 Apr 2009 7:00] Hakan Küçükyılmaz

Test did not fail since February 26th 2009 anymore. I checked with xref.