Gregory,
> On 18 Jul 2004, at 21:53, Werner Guttmann wrote:
> > that feels like you wanted to start digging right now .. ;-). Do you
> > need any other signs of encouragement than this very email ?
>
> Not right now this moment - I'm trying to ship a rewrite of our user
> registration system in preparation for an implementation of Liberty
> authentication, and am up to my ears in "can you make this pixel blue".
>
> However, probably immediately afterwards, yes - possibly this coming
> weekend.
If you happen to have any questions, please bear in mind that I will not be
able to help you as I am away next week, and won#t be returning before
August 3rd, 2004.
> > Having said that, please be aware that once I am back from my next
> > holiday (around August 3rd, that is .. ;-)), I'll be making changes to
> > the caching
> > mechanism to prepare myself (and the caching mechanism) for the
> > introduction of a distributed cache.
>
> In that case, I'm really better off waiting. ;)
Well, not really ... as all I'll be doing is make Cache.java implement
java.util.Map. As it stands right now, Cache.java does implememt some
(actually most of) of the Map methods, but not all.
Once I have done this, I'll need to update all current cache implementations
as well, adding default/dummy method implementations. Then, and only then,
I'll start to implememnt a distributed cache implementation.
The reward is that this will be absolutely easy once I have made above
changes.
> My issue is that I'm completely uninterested in an in-memory cache
> solution; I want to be able to shut down JVMs and take them out of
> service without affecting our performance in the mid-term. So whatever
> API you end up using, I'll probably end up writing a memcached plugin
> for it. I've already rewritten the client-side handling of memcached,
> and added a client-side "short term memory" cache for minimising
> short-term thrashing of objects in and out of serialization. I may
> have to move to a (preexisting?) distributed cache for that element of
> the design.
>
> It is a problem I have right now; so I am interested in getting it
> solved. But it's not on fire... (at least, I don't think it's on fire
> - I guess I won't really know till I deploy this beast in the next
> couple of days)
-----------------------------------------------------------
If you wish to unsubscribe from this mailing, send mail to
[EMAIL PROTECTED] with a subject of:
unsubscribe castor-dev
