Forum OpenACS Development: Re: Duplicate keys on assessment - possible CR bug
In my understanding about this thread, the Deadlock problem is solved when you use PostgreSQL 8.2.X, right? So, in OpenACS 5.4 is the lock removed? Because I'm watching the HEAD version of acs-content-repository and the lock is still there. I'm running an assessment with a lot of concurrent users and this lock seems a performance problem to me.
With PostgreSQL MVCC improvements for 8.2.X, can I just coment out the lock line in assessment? Or is somebody going to fix this for the next version?
So I'd encorage anyone who with a newer version of Postgres to make sure those LOCK TABLE statements are not being used. Look in acs-content-repository/tcl/content-item-procs.tcl and content-revision-procs.tcl.