You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Andres de la Peña (Jira)" <ji...@apache.org> on 2021/07/14 13:05:00 UTC
[jira] [Commented] (CASSANDRA-16104) Wrong warning about data
volumes capacity
[ https://issues.apache.org/jira/browse/CASSANDRA-16104?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17380576#comment-17380576 ]
Andres de la Peña commented on CASSANDRA-16104:
-----------------------------------------------
[~saprykin] Thanks for the patch, and sorry for the delayed review.
Getting concurrent compactors through JMX when validating the {{-j/--jobs}} option makes sense to me. Indeed, there is [a TODO comment|https://github.com/apache/cassandra/blob/cassandra-3.11/src/java/org/apache/cassandra/tools/NodeProbe.java#L302] about this in 3.11, dating back to CASSANDRA-9054. I have rebased the patch and created versions for 3.11, 4.0 and trunk, and CI is running:
||patch||CI||
|[3.0|https://github.com/apache/cassandra/compare/cassandra-3.0...adelapena:16104-3.0]|[j8|https://app.circleci.com/pipelines/github/adelapena/cassandra/661/workflows/3e91ee7c-eb67-416b-af65-3e0314e5d359]|
|[3.11|https://github.com/apache/cassandra/compare/cassandra-3.11...adelapena:16104-3.11]|[j8|https://app.circleci.com/pipelines/github/adelapena/cassandra/664/workflows/2fe0f649-b846-47e4-8a42-74c04aef75c4]|
|[4.0|https://github.com/apache/cassandra/compare/cassandra-4.0...adelapena:16104-4.0]|[j8|https://app.circleci.com/pipelines/github/adelapena/cassandra/663/workflows/e58e0821-b6a7-4b18-923d-f0eddf743734] [j11|https://app.circleci.com/pipelines/github/adelapena/cassandra/663/workflows/be12b427-30ed-429b-9a4e-ca0ec1a5e53a]|
|[trunk|https://github.com/apache/cassandra/compare/cassandra-trunk...adelapena:16104-trunk]|[j8|https://app.circleci.com/pipelines/github/adelapena/cassandra/662/workflows/bf7ad882-b6f7-4c13-a534-0e64b740c748] [j11|https://app.circleci.com/pipelines/github/adelapena/cassandra/662/workflows/b1ff8a2f-b268-407f-98ef-cb13b10fc575]|
From 3.11 and beyond we don't need to add the mbean methods since they already exist since CASSANDRA-12248, so we only need to use them from {{NodeProbe}} to get rid of all the dependencies from {{DatabaseDescriptor}}.
I wonder whether it's worth adding a dtest setting a couple of nodes with different values for {{concurrent_compactors}} and call them with nodetool to verify that the {{-j/--jobs}} option issues the warning depending on the connected node. I think it should be a Python dtest since, if I'm right, Java dtests use an internal probe that doesn't use JMX.
> Wrong warning about data volumes capacity
> -----------------------------------------
>
> Key: CASSANDRA-16104
> URL: https://issues.apache.org/jira/browse/CASSANDRA-16104
> Project: Cassandra
> Issue Type: Bug
> Components: Tool/nodetool
> Reporter: Dmitrii Saprykin
> Assignee: Dmitrii Saprykin
> Priority: Normal
>
> I see the following warning trying to run *nodetool upgradesstables*
> {noformat}
> WARN 16:09:24 Only 34988 MB free across all data volumes. Consider adding more capacity to your cluster or removing obsolete snapshots
> {noformat}
> This warning is wrong because the wrong storage device capacity gets tested.
> All my cassandra data paths are subdirectories of */data* mount point which has enough of space:
> {noformat}
> $ df -h /data
> Filesystem Size Used Avail Mounted on
> ................. 1.2T 200G 1T /data
> {noformat}
> However what Warning reports is a OS mount which has nothing to do with Cassandra configuration:
> {noformat}
> df -h /
> Filesystem Size Used Avail Use% Mounted on
> ............ 40G 5.7G 35G 15% /
> {noformat}
> I see this error running Cassandra 3.0.22
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@cassandra.apache.org
For additional commands, e-mail: commits-help@cassandra.apache.org