On the PostgreSQL lists there has been some discussion recently as to
the mechanism behind, benefits and drawbacks, of persistent connections.
In particular a scenario similar to the following was brought up:
Browser A connects to Apache child N, and calls a web page that calls a
script which issues SQL commands that involve opening and committing a
transaction:
- begin work;
- do some selects, inserts, deletes, updates . . .
Now, the question is: Is it conceivable that in the meantime, whilst
Browser A is waiting for the output from the script, that Browser B
talks to same Apache child, uses the same persistent connection, and
messes up the transaction that Browser A initiated in some unpredictable
way?
- do more selects, inserts, deletes, updates . . .
- commit work;
My guess would be that the Apache child wouldn't allow Browser B to talk
to it whilst Browser A is waiting for output from the script and that
that means that everything is fine and hunky-dory . . . I'm not sure though,
and there is nagging trace of murkiness.
Cheers, Frank
--
PHP General Mailing List (http://www.php.net/)
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
To contact the list administrators, e-mail: [EMAIL PROTECTED]
