[
https://issues.apache.org/jira/browse/SLING-3332?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13879867#comment-13879867
]
Antonio Sanso commented on SLING-3332:
--------------------------------------
here the details about 100 and 1000 children with non existing resource and
alias without and with the optimization in SLING-2521
{code}
# ResolveNonExistingWith100AliasTest min 10% 50% 90% max
jcr.resource-2.2.10 1 1 1 2 16
jcr.resource-2.2.10 0 0 0 0 2
{code}
and
{code}
# ResolveNonExistingWith1000AliasTes min 10% 50% 90% max
jcr.resource-2.2.10 14 15 18 19 27
jcr.resource-2.2.10 0 0 0 0 7
{code}
so obviously the optimization in SLING-2521will bring some benefit in terms of
resource resolution time (but the improvement is not so dramatic in this case)
The real slow down would be with something like 5000 direct-children
{code}
# ResolveNonExistingWith10000AliasTe min 10% 50% 90% max
jcr.resource-2.2.10 195 197 224 248 274
jcr.resource-2.2.10 0 0 0 0 1
{code}
but hopefully this is not a common case
> Long startup time with many sling:alias
> ---------------------------------------
>
> Key: SLING-3332
> URL: https://issues.apache.org/jira/browse/SLING-3332
> Project: Sling
> Issue Type: Improvement
> Components: ResourceResolver
> Reporter: Antonio Sanso
> Priority: Minor
>
> Similarly to SLING-3290 (Long startup time with many vanityPath) .
> When many sling:alias are present the system take long time to startup ,
> Same when alias is removed or updated (and many sling:alias are present) .
> The reason behind is the usage of a query that updates the aliasMap used as
> cache for sling:alias.
> Performance test report to follow
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
