So, that means Lyla could not be used reliably in a clustered
environment, where load balancing between servers might jump a user from
one server to another. It might not happen often (our load balancer
uses a 10-minute sticky session based on IP address, for example), but I
sure wouldn't want to have to debug problems caused by this in the real
world.
It's one thing to use memory as a cache (LRU makes sense), but to use
memory exclusively is not safe.
(The only reason I'm saying anything is because I've been following this
thread about Lyla with interest... it looks like a really nice
implementation that I may use. I guess making something that's good
attracts hopefully-constructive criticism.)
Thanks
Mark
-----Original Message-----
From: Peter J. Farrell [mailto:[EMAIL PROTECTED]
Sent: Sunday, August 06, 2006 9:27 PM
To: CF-Talk
Subject: Re: LylaCaptcha Hash Issue
Lyla uses an in-memory LRU cache for holding hash references (think
about your CPU). The reason why you are losing the hashes and the
validation isn't working is because you need to init() Lyla into a
persistent scope (recommended application scope). The salt is getting
reset on each time the captcha is getting init()ed. Be sure to stick it
in the application scope.
....Peter
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~|
Introducing the Fusion Authority Quarterly Update. 80 pages of hard-hitting,
up-to-date ColdFusion information by your peers, delivered to your door four
times a year.
http://www.fusionauthority.com/quarterly
Archive:
http://www.houseoffusion.com/groups/CF-Talk/message.cfm/messageid:248986
Subscription: http://www.houseoffusion.com/groups/CF-Talk/subscribe.cfm
Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4
