[
https://issues.apache.org/jira/browse/SLING-13284?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18100755#comment-18100755
]
Julian Reschke commented on SLING-13284:
----------------------------------------
Yes, but this is a workaround to a deeper problem.
As the index may not reflect the current state of the repository, paging will
always be (slightly) unreliable, and miss stuff on cache init.
Maybe we need a better strategy - use the equivalent of the JCR rows iterator
(which will use the query result directly). Also, as observed, the cache might
be out of date for various reasons; maybe we too to re-init it regularly (in
the background).
> PagedQueryIterator aborts alias/vanity cache init when async index is stale
> ---------------------------------------------------------------------------
>
> Key: SLING-13284
> URL: https://issues.apache.org/jira/browse/SLING-13284
> Project: Sling
> Issue Type: Bug
> Components: ResourceResolver
> Affects Versions: Resource Resolver 1.12.12
> Reporter: Sagar Miglani
> Assignee: Sagar Miglani
> Priority: Major
>
> {{PagedQueryIterator.getNext()}} throws {{QueryImplementationException}} when
> a row's live {{values[0]}} (read from repo) is less than the previous row's -
> aborting the entire alias/vanity cache build. This check assumes the async
> index ordering matches live property values, which cannot hold for an
> eventually-consistent async index.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
