| Fri, 06 Aug 2010 17:33:07 +0200 |
Sylvain Thénault |
[session] should be _cw, not req
stable
|
file |
diff |
annotate
|
| Wed, 28 Jul 2010 16:31:32 +0200 |
Sylvain Thénault |
backport stable
|
file |
diff |
annotate
|
| Wed, 28 Jul 2010 11:37:02 +0200 |
Sylvain Thénault |
[session] this log should be at the debug level, not info
stable
|
file |
diff |
annotate
|
| Mon, 26 Jul 2010 12:15:11 +0200 |
Sylvain Thénault |
[dbapi] cleanup shared data api: let access to transaction from dbapi, we can write it after all... Also, querydata is better named txdata
|
file |
diff |
annotate
|
| Wed, 07 Jul 2010 13:01:44 +0200 |
Sylvain Thénault |
backport stable
|
file |
diff |
annotate
|
| Tue, 06 Jul 2010 17:09:14 +0200 |
Sylvain Thénault |
[repository] fix so that when repository is shutting down, internal session in transaction are interrupted
stable
|
file |
diff |
annotate
|
| Mon, 05 Jul 2010 19:05:56 +0200 |
Sylvain Thénault |
[session] don't log reraised error occured during precommit
|
file |
diff |
annotate
|
| Mon, 05 Jul 2010 18:25:19 +0200 |
Sylvain Thénault |
[schema migration] import refactoring to fix #1109558 and enhances things on the way
|
file |
diff |
annotate
|
| Thu, 01 Jul 2010 17:06:37 +0200 |
Sylvain Thénault |
backport stable
|
file |
diff |
annotate
|
| Wed, 23 Jun 2010 09:27:52 +0200 |
Alexandre Fayolle |
[session] refactor session handling so that when calling session.close(), pool of long running transaction of the same session is properly freed
stable
|
file |
diff |
annotate
|
| Mon, 21 Jun 2010 15:34:46 +0200 |
Sylvain Thénault |
backport stable
|
file |
diff |
annotate
|
| Mon, 21 Jun 2010 15:32:26 +0200 |
Sylvain Thénault |
[transaction w/ separated web/repo processes] the dbapi should explicitly specify a transaction id to avoid confusion when web server / repository run in separated processes
stable
|
file |
diff |
annotate
|
| Mon, 21 Jun 2010 13:23:11 +0200 |
Sylvain Thénault |
backport stable
|
file |
diff |
annotate
|
| Mon, 21 Jun 2010 08:52:13 +0200 |
Sylvain Thénault |
[session] when rollback_on_failure is false, we should not try to reconnect if the sql raised an error
stable
|
file |
diff |
annotate
|
| Fri, 18 Jun 2010 15:41:39 +0200 |
Sylvain Thénault |
[session] update session's timestamp in session.execute, so long running transactions are not erroneously closed by the repository
stable
|
file |
diff |
annotate
|