[
https://issues.apache.org/jira/browse/SLING-12879?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18112568#comment-18112568
]
Julian Reschke commented on SLING-12879:
----------------------------------------
bq. With this setup, we end up with more than 480,000 aliases. As a result:
That is a lot.
bq. The background alias cache population also takes about 2 hours before the
instance reaches acceptable performance.
Yes, there's a toggle in OSGi.
bq. Specifically, would it be possible to fall back to the non-cached alias
resolution mechanism until the cache has been fully built, since this appears
to be significantly faster in our environment?
That's what it's supposed to do during init.
Q: do you have a resourceresolver index in the system? (will have to check the
name in CS)
Q: could you set DEBUG level logging on AliasHandler and PagedQueryIterator and
attach the logs?
> ResourceResolver: add option to run alias initializer in background
> -------------------------------------------------------------------
>
> Key: SLING-12879
> URL: https://issues.apache.org/jira/browse/SLING-12879
> Project: Sling
> Issue Type: Sub-task
> Components: ResourceResolver
> Reporter: Julian Reschke
> Assignee: Julian Reschke
> Priority: Major
> Fix For: Resource Resolver 1.12.12, Resource Resolver 2.0.2
>
>
> This will add that option, but keep it disabled.
> For now it will only be tested in unit tests.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
