[
https://issues.apache.org/jira/browse/OAK-8673?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16973150#comment-16973150
]
Thomas Mueller commented on OAK-8673:
-------------------------------------
So with cach size 0 (no cache), the system is very slow (basically unusable).
So a cache is need. I see two problems:
* A: Having one cache per session is problematic if there is no limit in the
number of sessions: there is no way to guarantee the system will not run out of
memory. Is there no way to use just one cache (for all sessions)?
* B: Having a cache size in number of entries is problematic, if memory usage
of entries is very different: there is no way to guarantee the system will not
run out of memory. To solve this, in various places in Oak we use "weighted"
caches, and estimate memory usage of entries (e.g. for strings, 24 + number of
characters). I can help with this.
I think both A and B need to be addressed.
> Determine and possibly adjust size of eagerCacheSize
> ----------------------------------------------------
>
> Key: OAK-8673
> URL: https://issues.apache.org/jira/browse/OAK-8673
> Project: Jackrabbit Oak
> Issue Type: Technical task
> Components: core, security
> Reporter: Angela Schreiber
> Assignee: Angela Schreiber
> Priority: Major
>
> The initial results of the {{EagerCacheSizeTest}} seem to indicate that we
> almost never benefit from the lazy permission evaluation (compared to reading
> all permission entries right away). From my understanding of the results the
> only exception are those cases where only very few items are being accessed
> (e.g. reading 100 items).
> However, I am not totally sure if this is not a artifact of the random-read.
> I therefore started extending the benchmark with an option to re-read a
> randomly picked item more that once, which according to some analysis done
> quite some time ago is a common scenario specially when using Oak in
> combination with Apache Sling.
> Benchmarks with 10-times re-reading the same random item:
> As I would have expected it seems that the negative impact of lazy-loading is
> somewhat reduced, as the re-reading will hit the cache populated while
> reading.
> Result are attached to OAK-8662 (possibly more to come).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
