We are running Postgres 7.3 on a Debian Linux system.
Our python web application uses psycopg to talk to the Postgres server.
Our problem is that our queries can be on millions of records of data
and take a long time. At times, the user (of the web app) loses
interest in waiting for the results. I can have the web app move on,
but the postmaster process which is taking care of the request stays
around until it completes. When we have several impatient users,
then the number of server processes left running makes the whole
system's performance degrade considerably.
I'd like to have the ability to "interrupt" that server process and
have it complete/abort.
Is there any plan to support this functionality in a future release?
If not, where should I submit the request?
Thank you!
Laura Arbilla
Deepfile Corporation
larbilla@deepfile.com