[SERVER-5918] invalid memory writes in sharding test suite Created: 24/May/12 Updated: 11/Jul/16 Resolved: 11/Jun/12 |
|
| Status: | Closed |
| Project: | Core Server |
| Component/s: | Sharding |
| Affects Version/s: | 2.1.1 |
| Fix Version/s: | 2.1.2 |
| Type: | Bug | Priority: | Major - P3 |
| Reporter: | Eric Milkie | Assignee: | Greg Studer |
| Resolution: | Done | Votes: | 0 |
| Labels: | buildbot | ||
| Remaining Estimate: | Not Specified | ||
| Time Spent: | Not Specified | ||
| Original Estimate: | Not Specified | ||
| Issue Links: |
|
||||||||
| Operating System: | ALL | ||||||||
| Participants: | |||||||||
| Description |
|
To reproduce, run "test sharding" with Valgrind
|
| Comments |
| Comment by auto [ 08/Jun/12 ] | ||||
|
Author: {u'login': u'gregstuder', u'name': u'Greg Studer', u'email': u'greg@10gen.com'}Message: | ||||
| Comment by Greg Studer [ 24/May/12 ] | ||||
|
Probably fastest actually to keep the current override and use ConnectionString::OTHER to indicate. Otherwise we'll need pluggable connection types, which is a pretty big change. | ||||
| Comment by Greg Studer [ 24/May/12 ] | ||||
|
Thanks for tracking this down - the issue is that, for testing, it's useful to replace the connections to shards with connections to the local DB. May need to create a separate kind of connection string. | ||||
| Comment by auto [ 24/May/12 ] | ||||
|
Author: {u'login': u'milkie', u'name': u'Eric Milkie', u'email': u'milkie@10gen.com'}Message: | ||||
| Comment by Eric Milkie [ 24/May/12 ] | ||||
|
Looks like the commit d1fd245ea8ec1a75974169fc0f07d21dc3217a96 for | ||||
| Comment by Eric Milkie [ 24/May/12 ] | ||||
|
Findings:
The C style cast is causing memory corruption due to a bug where somehow a DBDirectClient connection is unexpectedly getting added to the connection pool. It remains to be discovered why the DBDirectClient* was added to the pool. |