cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Marcus Eriksson (JIRA)" <j...@apache.org>
Subject [jira] [Comment Edited] (CASSANDRA-8562) Fix checking available disk space before compaction starts
Date Mon, 12 Jan 2015 10:18:34 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-8562?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14273287#comment-14273287
] 

Marcus Eriksson edited comment on CASSANDRA-8562 at 1/12/15 10:18 AM:
----------------------------------------------------------------------

bq. Delete default no-op implementation of reduceScopeForLimitedSpace from DiskAwareRunnable
as it's no longer used
fixed

bq. is throwing an RTE the best thing for us to do when we don't have sufficient disk space
for a node to compact files?
Perhaps not, using something like a compaction_failure_policy might be good, problem is though
that in the compaction case there is a chance that we might actually recover - we could have
10 compactions going eating up disk space but once those are done we will have plenty of room
on the drive. Perhaps if we do something like "if we fail finding space to compact to for
X minutes, die"?

I wanted a quick fix in for 2.0.12 though, as we broke this in CASSANDRA-8329 (which is not
yet released)

edit: oops misread, thought you suggested adding compaction_failure_policy in this ticket,
will get it committed


was (Author: krummas):
bq. Delete default no-op implementation of reduceScopeForLimitedSpace from DiskAwareRunnable
as it's no longer used
fixed

bq. is throwing an RTE the best thing for us to do when we don't have sufficient disk space
for a node to compact files?
Perhaps not, using something like a compaction_failure_policy might be good, problem is though
that in the compaction case there is a chance that we might actually recover - we could have
10 compactions going eating up disk space but once those are done we will have plenty of room
on the drive. Perhaps if we do something like "if we fail finding space to compact to for
X minutes, die"?

I wanted a quick fix in for 2.0.12 though, as we broke this in CASSANDRA-8329 (which is not
yet released)

> Fix checking available disk space before compaction starts
> ----------------------------------------------------------
>
>                 Key: CASSANDRA-8562
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-8562
>             Project: Cassandra
>          Issue Type: Bug
>            Reporter: Marcus Eriksson
>            Assignee: Marcus Eriksson
>             Fix For: 2.0.12, 2.1.3
>
>         Attachments: 0001-Check-for-available-disk-space-before-starting-compa.patch,
8562-v2.patch
>
>
> When starting a compaction we check if there is enough disk space available to start
it, otherwise we might (for STCS) reduce the compaction so that the result could fit. Now
(since CASSANDRA-8329) we check for the directory to write to a lot later and this can reduce
the compaction after we have created the scanners.



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

Mime
View raw message