[Catalyst] Session Expiry

[Bill Moseley]

On Sun, Mar 02, 2008 at 01:31:42PM +0000, Ash Berlin wrote:
> 
> On 2 Mar 2008, at 03:50, Jonathan Rockway wrote:
> 
> >In the second case, it's just a bug in the Session::Store::Cache store
> >that can be fixed with the addition of a backend_expires configuration
> >key, probably set to "never" by default.
> 
> Not quite true. It would look like it fixes it for most cases. A cache  
> is just a cache. Don't rely on them (particularly memcahced) to keep  
> data around until you told it you would like it to expire at - there  
> is no guarantee of this happening, especially under heavy load.

Yes, very true.  A cache is a cache.  And the database server could
blow up, too.

What we are talking about is a situation where the session data is
(seemingly) suppose to persist for X seconds between requests, but
it's instead persisting X seconds between updates to the session
(which may not be every request).  It appears from the design of the
Session plugin that it's suppose to persist X seconds between requests
(the "expires" entry).  So, it looks like a design problem.

The fact that you might be using a store that deletes entries for
other reasons than expire time is another issue.

-- 
Bill Moseley
moseley at hank.org