[
https://issues.apache.org/jira/browse/SOLR-8858?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15364857#comment-15364857
]
Caleb Rackliffe commented on SOLR-8858:
---------------------------------------
bq. I am not very happy with this solution because from a Solr user's
perspective, this feature adds no benefit but causes stored fields to be read
twice for an uncached read? I must also admit that I do not have a good
suggestion on how to avoid that.
[~shalinmangar] Unless there's a less invasive solution I'm overlooking, I
think it might be best to abandon this issue as something to handle in Solr
proper.
[~dsmiley] Our fork actually reads most stored fields from an embedded database
and relies on the visitor's fields information to make decisions about when
(and when not) to read stored fields from Solr itself. We don't actually use
{{documentCache}} at all, so the fixes I made around the initial patch to get
the unit tests passing won't even be necessary.
Let me know if there are any objections to closing this...
> SolrIndexSearcher#doc() Completely Ignores Field Filters Unless Lazy Field
> Loading is Enabled
> ---------------------------------------------------------------------------------------------
>
> Key: SOLR-8858
> URL: https://issues.apache.org/jira/browse/SOLR-8858
> Project: Solr
> Issue Type: Bug
> Affects Versions: 4.6, 4.10, 5.5
> Reporter: Caleb Rackliffe
> Labels: easyfix
>
> If {{enableLazyFieldLoading=false}}, a perfectly valid fields filter will be
> ignored, and we'll create a {{DocumentStoredFieldVisitor}} without it.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
