Bugzilla – Bug 3191
Prepared statements for queries in RFT
Last modified: 2005-10-03 13:17:24
You need to log in before you can comment on or make changes to this bug.
Investigate and implement poolable Prepared statements for all the queries in RFT.
A few notes from me, if you don't mind. When running a container, I see a number or persistent database connections. Initially it's just 1 or 2, but once I've hit the container with jobs, 6 or more connections to a Pg. These connections are still around when the container is idle again. If the Pg gets restarted even while all connections are idle, I don't think that RFT recovers (does it?). If the connections were to go away after some idle time, there are usually time windows where it is "safe" to restart the Pg. I believe RFT needs to deal with both cases: Suddenly gone database connections (and try to restart before giving up), and disconnect after idle.
*** Bug 3288 has been marked as a duplicate of this bug. ***
All frequently used RFT queries are converted into Prepared Statements now.
I may have asked this before: Are you re-connect safe? I mean, if I have an RFT up and idle, and I restart my rDBMS, will RFT recover from it? After all, preparted statements are usually associated with a database handle, and would need to be renewed, if the connection broke away.