> Given the description, it really looks to me like a race condition. > At this point, I think we should try to sync everything transaction related > and see what happens. > Could you try and add a synchronized keyword to all the methods in > AbstractSimpleService and try again ? > It probably would be worth trying to sync the transaction manager methods > too. > > The transaction itself doesn't need to be synced, since there's a 1-1 > mapping between the transaction object and the thread. >
Thanks Remy. I've rebuilt slide with these synchronized attributes added, and I'm about to start stress-testing things on our server. If I have anything to report soon, I'll do so, otherwise it could be a few days until I get another chance to look at any of this (been a bit sick, not doing much work at the moment). I've found it easy to reproduce this (and/or similar problems) by simply doing _anything_ involving 2 or more users writing to the repository at the same time (more makes the problem show up more often, of course). It doesn't seem to make much difference what is being written to the repository, either. Additional load from users doing other things (not writing anything) might also make it more likely to cause problems, but doesn't directly trigger anything. Michael -- To unsubscribe, e-mail: <mailto:[EMAIL PROTECTED]> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
