BIND 10 #2459: call to getCachedZoneWriter must be protected by mutex
BIND 10 Development
do-not-reply at isc.org
Wed Nov 7 06:12:10 UTC 2012
#2459: call to getCachedZoneWriter must be protected by mutex
-------------------------------------+-------------------------------------
Reporter: | Owner: jinmei
jinmei | Status: accepted
Type: | Milestone:
defect | Sprint-20121120
Priority: | Resolution:
medium | Sensitive: 0
Component: | Sub-Project: DNS
b10-auth | Estimated Difficulty: 3
Keywords: | Total Hours: 0
Defect Severity: High |
Feature Depending on Ticket: |
Add Hours to Ticket: 0 |
Internal?: 0 |
-------------------------------------+-------------------------------------
Comment (by jinmei):
trac2459 is ready for review.
The branch is exactly what's proposed in the ticket's description
(and it's small).
I also confirmed my theory was (seemingly) correct by the following
setup:
- run b10-auth with an sqlite3-only data source
- keep sending queries to b10-auth at a high volume using
queryperf(-like tool)
- at the same time, repeat "Auth loadzone" from bindctl in a shell
loop
Eventually b10-auth crashed due to segmentation fault (normally within
a minute or so). It died somewhere around sqlite3 API, so (while I
didn't looked into it any further) it's quite likely to be a race I
suspected. After applying this change, it didn't happen.
I don't think we need a changelog for this fix as it's an intermediate
regression.
--
Ticket URL: <https://bind10.isc.org/ticket/2459#comment:5>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development
More information about the bind10-tickets
mailing list