[
https://issues.apache.org/jira/browse/CASSANDRA-7392?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14791741#comment-14791741
]
Stefania commented on CASSANDRA-7392:
-------------------------------------
bq. I'm not sure which part of the documentation you are reading.
The last two lines of the [first
paragraph|http://bugs.java.com/bugdatabase/view_bug.do?bug_id=6275329]:
{code}
(or equivalently, might not be visible to other threads) until some other
volatile write or synchronizing action occurs).
{code}
So, if I understood you correctly, you are positive that a store-store fence
(rather than a store-load fence) is sufficient to ensue other CPUs see a
changed state because of the Intel Cache coherence protocol (MESI/MESIF)? Well
I suppose the atomic reference get() is a volatile read so that would ensure a
store-load fence anyway.
I therefore reverted back to {{lazySet}}, decreased the number of entries from
50 to 30, and picked them at random.
> Abort in-progress queries that time out
> ---------------------------------------
>
> Key: CASSANDRA-7392
> URL: https://issues.apache.org/jira/browse/CASSANDRA-7392
> Project: Cassandra
> Issue Type: New Feature
> Components: Core
> Reporter: Jonathan Ellis
> Assignee: Stefania
> Priority: Critical
> Fix For: 3.x
>
>
> Currently we drop queries that time out before we get to them (because node
> is overloaded) but not queries that time out while being processed.
> (Particularly common for index queries on data that shouldn't be indexed.)
> Adding the latter and logging when we have to interrupt one gets us a poor
> man's "slow query log" for free.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
