cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Stefania (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CASSANDRA-7392) Abort in-progress queries that time out
Date Mon, 05 Oct 2015 04:53:27 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-7392?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14942925#comment-14942925
] 

Stefania commented on CASSANDRA-7392:
-------------------------------------

bq. It kind of seems like they should both log at the same frequency

I've changed it so that they log at the same frequency and it is indeed better this way. However,
we introduced the no spam logger to avoid polluting the main log file whereas now we potentially
log one line every 5 seconds although not the details so it should be OK.

Than you for picking up the remaining points, they should be OK now. I converted {{numDroppedOperations}}
to an atomic log and in theory we could have a race in {{aggregateFailedOperations()}}, after
emptying the queue and before resetting the dropped operations to zero but the only way to
synchronize both would be via a lock and I don't think it strictly matters if we incorrectly
say that there were more dropped operations in one interval rather than the next. WDYT?

As for logging during unit tests, I've changed {{logback-test.xml}} since we cannot change
the log level at runtime, I hope that's OK. I think we have a problem with the {{TeeingAppender}},
I could not see any log messages in the log files, only on {{stdout}}.

> 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)

Mime
View raw message