[imp] session does not timeout

Eddie Urenda eurenda@cts.ucla.edu
Thu, 3 Oct 2002 09:47:09 -0700 (PDT)


Forgot to mention that the settings I used to accomplish this were:

session.gc_probability = 100
session.gc_maxlifetime = 30

-Eddie

On Thu, 3 Oct 2002, Eddie Urenda wrote:

> Jonathan,
> 
> The settings you mention below worked for me, and the garbage collector
> runs 100% of the time, when a new session is initiated.
> 
> In other words, it doesn't time out on its own after 60 secs, but after 60
> secs if I initiate a second login session, then the gabage collection is
> started (since I set the probability to 100%), and the first session times
> out.
> 
> The key is that another session must be started for garbage collection to
> be run (at least in my tests), it doesn't run on it's own after 60 secs
> (this would probably be resource intensive for a large system).
> 
> -Eddie
> 
> On Thu, 3 Oct 2002, Jonathan Giles wrote:
> 
> > Mike:
> > 
> > Thanks for the clarification on the use of garbage collection.
> > 
> > Do you have any suggestions on how I would force timeouts on sessions
> > whether or not the system is in high load?  I don't expect the system to
> > have real high loads for some time, but still want to prevent people
> > from walking away from open sessions and having them stay open. 
> > 
> > I have also tried this...
> > session.gc_probability = 100
> > and the session still doesn't time out.
> > 
> > I have also tried this...
> > session.cookie_lifetime = 60
> > and the session still doesn't time out gosh darn.
> > 
> > Thanks so much for the help!
> > 
> > Jon Giles
> > 
> > 
> > 
> > Mike Cochrane wrote:
> > 
> > > >From what you're describing here, it doesn't sound like you're testing in a way
> > > that would allow the garbage collector to run....
> > >
> > > The gc doesn't run every request to php, that's over kill... instead it has a
> > > probability set in php.ini... if you're not getting many requests to your server
> > > then it's not going to get run very often and the session aren't going to get
> > > timed out...
> > >
> > > if your server is under higher load then the gc is going to run more often and
> > > timeout the sessions, or if you incread the probability in php.ini the same
> > > result will happen.
> > >
> > > if it's just you prodding you're server, it's not likely that the gc would have run.
> > >
> > > - Mike :-)
> > >
> > > > ----- Message from jong@clinedavis.com ---------
> > > >
> > > > I looked all over the faq and lists for this one and couldn't find it.
> > > >
> > > > I am trying to force a timeout for imp sessions, and the usual ways
> > > > don't seem to work.
> > > >
> > > > In the older imp, it was done in the local.inc if I remember correctly,
> > > > but there is none around the new version as best as I see.
> > > >
> > > > On this list, a few different things were mentioned to be found in the
> > > > php.ini file.
> > > >
> > > > I went to http://www.php.net/manual/en/ref.session.php for info on each
> > > > value.
> > > >
> > > > I found that session.gc_maxlifetime would do it for me, if it worked.  I
> > > > set it to something really small, 60 seconds, so I can see the results.
> > > > I restarted the apache server.  I would think that after the 60secs, the
> > > > session file would be removed.  No such thing happens, and the session
> > > > stays open even after the 60 secs.
> > > >
> > > > I also tried session.cookie_lifetime as well, but that didn't work
> > > > either.
> > > >
> > > > BTW I have set the session file to be written to /var/www/tmp, which
> > > > I created to keep session cookies from intruders as best as I can.
> > > >
> > > > Is there something in horde/imp that overides these values, or maybe I
> > > > need to turn something else on or off in php.ini?
> > > >
> > > > Any help would be really appreciated!
> > > >
> > > > Thanks,
> > > >
> > > > Jon Giles
> > > >
> > > >
> > > >
> > > > --
> > > > IMP mailing list
> > > > Frequently Asked Questions: http://horde.org/faq/
> > > > To unsubscribe, mail: imp-unsubscribe@lists.horde.org
> > > >
> > > > ----- End message from jong@clinedavis.com -----
> > >
> > > --
> > > IMP mailing list
> > > Frequently Asked Questions: http://horde.org/faq/
> > > To unsubscribe, mail: imp-unsubscribe@lists.horde.org
> > 
> > --
> > Jonathan Giles
> > Senior Unix Administrator
> > Cline Davis Mann, Inc.
> > 
> > --
> > Privileged/Confidential Information may be contained in this
> > message.  If you are not the addressee indicated in this message
> > (or responsible for delivery of the message to such person), you
> > may not copy or deliver this message to anyone.  In such case,
> > you should destroy this message and kindly notify the sender
> > by reply e-mail.  Please advise immediately if you or your
> > employer do not consent to Internet e-mail of this kind.
> > Opinions, conclusions, and other information in this message
> > that do not relate to the official business of CDM shall
> > be understood as neither given nor endorsed by it.
> > 
> > 
> > 
> > 
> > -- 
> > IMP mailing list
> > Frequently Asked Questions: http://horde.org/faq/
> > To unsubscribe, mail: imp-unsubscribe@lists.horde.org
> > 
> 
> 
> -- 
> IMP mailing list
> Frequently Asked Questions: http://horde.org/faq/
> To unsubscribe, mail: imp-unsubscribe@lists.horde.org
>