Hello,

> As far as I understand JR has it own in-memory cache 
> implementation and surely it does not use any of the existing 
> open source cache solutions.

IMHO an open source cache solution would be an overkill (simple weak
reference maps or some lru memory cache (hashmap on combination with
linkedmap) are IMO more transparant then some dependency on a caching
framework from which 95% is not used. I am used to only use caching
frameworks like ehcache or jcscache when I need filesystem caches / jvm
restart persistent caches or, though i never used before, cluster caches
with communication)

> 
> Caching will definitely not affect cluster as it was designed 
> to improve only local performance. If data source for JR data 
> is local to each cluster node, then you have to use 
> journaling in order to make them synchronized.
> 
> But I also have questions about cache mechanics and as far as 
> I know it's options are not realy configurable in JR 
> configuration file.

see [1] about configuring cache

[1] https://issues.apache.org/jira/browse/JCR-1011

-Ard


>  
> Best regards,
> 
> Andrey Adamovich
> 

Reply via email to