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-14201) Add a few options to nodetool verify
Date Wed, 31 Jan 2018 09:12:00 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-14201?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16346475#comment-16346475
] 

Marcus Eriksson edited comment on CASSANDRA-14201 at 1/31/18 9:11 AM:
----------------------------------------------------------------------

pushed another commit with option to not mutate repair status (it also fixes a bug where the
corrupt sstable was kept in the wrong compaction strategy)

Running {{nodetool verify}} by default runs the same way as before (with repair status change,
disk failure policy invoke and no version check), but with 9947 in mind, maybe we should change
the defaults to be 'safe' in trunk at least? Ie, don't invoke the disk failure policy and
don't mutate repair status - then the operator can decide what to do (the best way to handle
this is probably to just replace the node if verify fails)


was (Author: krummas):
pushed another commit with option to not mutate repair status (it also fixes a bug where the
corrupt sstable was kept in the wrong compaction strategy)

Running {{nodetool verify}} by default runs the same way as before (with repair status change,
disk failure policy invoke and version check), but with 9947 in mind, maybe we should change
the defaults to be 'safe' in trunk at least? Ie, don't invoke the disk failure policy and
don't mutate repair status - then the operator can decide what to do (the best way to handle
this is probably to just replace the node if verify fails)

> Add a few options to nodetool verify
> ------------------------------------
>
>                 Key: CASSANDRA-14201
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-14201
>             Project: Cassandra
>          Issue Type: Improvement
>            Reporter: Marcus Eriksson
>            Assignee: Marcus Eriksson
>            Priority: Major
>             Fix For: 3.0.x, 3.11.x, 4.x
>
>
> {{nodetool verify}} currently invokes the disk failure policy when it finds a corrupt
sstable - we should add an option to avoid that. It should also have an option to check if
all sstables are the latest version to be able to run {{nodetool verify}} as a pre-upgrade
check



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@cassandra.apache.org
For additional commands, e-mail: commits-help@cassandra.apache.org


Mime
View raw message