© 2004
Canonical Ltd.
•
Terms of use
•
Data privacy
•
Contact Launchpad Support
•
Blog
•
Careers
•
System status
•
35a32c5
(Get the code!)
@Alex,
So initially the customer was running Percona XtraDB Cluster 5.6.15-25.4 when he hit this bug, this is before my comment #6. After Percona XtraDB Cluster 5.6.15-25.5 was released on 3/20, I upgraded the nodes to 5.6.15-25.5 and still had the crash.
Comment #6 and following comments are from crashes on 5.6.15-25.5.
This bug is being hit perhaps in a different way?
In comment #1 Seppo indicated that "if mysql would use full table scan, we would not hit this issue".
However, based on the crash that I had and the stack traces in comment #8 and #9, we can see that MySQL is doing a full table scan, but still hits this issue.
Unfortunately, I do not have the specific test case anymore. The customer added primary keys to the tables that were hit with the issue, so I won't be able to reproduce it anymore.
I suppose you might get some required information from the attached gdb back traces?