[
https://issues.apache.org/jira/browse/SOLR-12210?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Kevin Risden resolved SOLR-12210.
---------------------------------
Resolution: Duplicate
This duplicates SOLR-7393. Not much detail added here over SOLR-7393 though.
The HDFS block cache shouldn't be involved with indexing though.
> Perfomance degradation at High Load
> -----------------------------------
>
> Key: SOLR-12210
> URL: https://issues.apache.org/jira/browse/SOLR-12210
> Project: Solr
> Issue Type: Bug
> Security Level: Public(Default Security Level. Issues are Public)
> Components: Hadoop Integration
> Affects Versions: 7.2
> Reporter: ROMAN NAZARENKO
> Priority: Minor
>
> When hdfsDataFactory is used and the incoming data stream is large, the rate
> of indexation is degraded. This is not a problem when using a local file
> system. I think the problem is in cleaning hdfsBlockCache. Because the
> degradation of the indexing rate occurs precisely when this cache reaches its
> maximum value. The basis for hdfsBlockCache is caffeine cache, but it is
> configured to clean only when the maximum size is reached. At the same time,
> nothing prevents using timeEviction, I suggest testing the indexing speed
> using this type of cache cleaning.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
