lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mark Miller (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (SOLR-8855) The HDFS BlockDirectory should not clean up it's cache on shutdown.
Date Mon, 21 Mar 2016 17:47:25 GMT

    [ https://issues.apache.org/jira/browse/SOLR-8855?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15204727#comment-15204727
] 

Mark Miller commented on SOLR-8855:
-----------------------------------

bq. Is this safe? It looks like we'd be dropping data that is still in the cache but hasn't
been written yet.

Not sure I understand that - this stops dropping from the cache, it does not add it.

bq. Why does the core container need to be volatile?

Because it is set in init - and we don't really know when an external user may call that or
in what thread. Technically, if that is to be thread safe, it should be set in the constructor
and be finally or be volatile. Given volatile reads without updates are basically as fast
as regular reads, just better to be airtight IMO.

> The HDFS BlockDirectory should not clean up it's cache on shutdown.
> -------------------------------------------------------------------
>
>                 Key: SOLR-8855
>                 URL: https://issues.apache.org/jira/browse/SOLR-8855
>             Project: Solr
>          Issue Type: Improvement
>            Reporter: Mark Miller
>            Assignee: Mark Miller
>             Fix For: master, 6.1
>
>         Attachments: SOLR-8855.patch
>
>
> The cache cleanup is done for early close and the global cache. On shutdown it just burns
time.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org


Mime
View raw message