[
https://issues.apache.org/jira/browse/SOLR-15892?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17468925#comment-17468925
]
Ankur Gupta commented on SOLR-15892:
------------------------------------
In our use case, the data deletion is triggered/automated by our users. And
could be in large numbers depending on their use case. The data selected for
deletion does not always fall into groups/categories that can be efficiently
formulated into the queries for DBQs.
Basically, one delete operation that is generated by the user does not often
map to many ids to effectively use DBQs. And such operations can be very
frequent and in large numbers. So in order to delete the required number of
records using DBQ, we'll need to form a large number of DBQs, and processing
such a large number of DBQ is not optimal and is taking a huge hit in
performance and throughput. Thus we rely on DeleteByIds for most deletes.
> RoutedAliasUpdateProcessor not distributing delete and commit requests.
> -----------------------------------------------------------------------
>
> Key: SOLR-15892
> URL: https://issues.apache.org/jira/browse/SOLR-15892
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Affects Versions: 8.1, 8.2, 8.3, 8.4, 8.5, 8.6, 8.7, 8.8, main (9.0), 8.9,
> 8.10, 8.11
> Reporter: Ankur Gupta
> Priority: Minor
> Fix For: main (9.0), 8.11
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> Currently, the RoutedAliasUpdateProcessor is not distributing and forwarding
> all delete and commit requests for a RoutedAlias to a fixed shard leader.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
