cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jörg Heinicke (JIRA) <j...@apache.org>
Subject [jira] Commented: (COCOON-2151) Sub-optimal implementation of AbstractDoubleMapEventRegistry
Date Tue, 27 Nov 2007 02:09:43 GMT

    [ https://issues.apache.org/jira/browse/COCOON-2151?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12545682
] 

Jörg Heinicke commented on COCOON-2151:
---------------------------------------

Ellis Pritchard wrote at COCOON-2146:

Thanks for your comment, Ard.

Your analysis may well explain some performance problems we were seeing on our system; unfortunately
I was not around at the time to analyze them, so I can't really be sure what caused them.
The action they took at the time was simply to cache less, i.e. only the expensive stuff,
rather than, well, basically everything!

Presumably though, if site-usage is fairly uniform, stuff that expires and is still referenced
from AbstractDoubleMapEventRegistry will get regenerated again at some fairly near point,
and replace the old entry with the new one, allowing it to be freed by the JVM? Only if you're
generating lots of unique pages or short-lived pages and caching them for a long time may
it become a 'real' problem?

I think that ehcache 1.4 (now in beta) may provide at least one solution to this problem,
with its new listeners functionality, which can get notification of the keys of entries removed
(for any reason) from the cache.

However, maybe there's a 'redesign the whole thing' kind of solution too?

As for whether we need it or not: according to "the powers that be" we do need a persistent
cache, so that when we bring the site up under load, on n instances (each with their own cache),
it doesn't hammer our database. The fact that the site was running without one for 8 months
due to this bug is not enough to convince them otherwise! It's either that, or having to pre-warm
the cache during site-down time (and making site down time longer), or even worse, running
the whole site behind mod_cache with an expiry time of about an hour or so, to cover any lag
in getting it up again. A persistent EventCache just looks so attractive compared to those
options... 

> Sub-optimal implementation of AbstractDoubleMapEventRegistry
> ------------------------------------------------------------
>
>                 Key: COCOON-2151
>                 URL: https://issues.apache.org/jira/browse/COCOON-2151
>             Project: Cocoon
>          Issue Type: Improvement
>          Components: Blocks: Event Cache
>    Affects Versions: 2.1.10, 2.2-dev (Current SVN)
>            Reporter: Jörg Heinicke
>
> This is just a follow-up from COCOON-2146 where Ard pointed out some issues with AbstractDoubleMapEventRegistry.
I just didn't want to lose the information when I actually fixed the issue. So I will add
it here.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message