[
https://issues.apache.org/jira/browse/CASSANDRA-19393?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17817325#comment-17817325
]
Stefan Miklosovic edited comment on CASSANDRA-19393 at 2/14/24 10:36 AM:
-------------------------------------------------------------------------
Well, this is something I would like to address as well. The current
implementation of dumping in MBean is creating a file, as you said, and it will
return just a file name. However, nodetool is meant to be used anywhere,
remotely, so what happens when I execute nodetool on a machine different from a
machine the node runs on? You see? E.g. admins can just have some jumpbox to a
cluster they manage, they do not necessarily have access to nodes themselves.
So they would not be able to read it.
Also, now it creates tmp file which is not deleted so /tmp will be populated
with these dumps until node is turned off which might take a lot of time and
can consume a lot of data if dumps are done frequently and they are big.
I think that we could return the whole string and it would be then parsed on
the nodetool. It might be also possible to send the format as an agrument to
MBean method and server would returned that formatted already.
If the string is "too big", we could compress it on server and decompress it on
the client but this is not the objective of this ticket and it is a subject of
further discussion.
was (Author: smiklosovic):
Well, this is something I would like to address as well. The current
implementation of dumping in MBean is creating a file, as you said, and it will
return just a file name. However, nodetool is meant to be used anywhere,
remotely, so what happens when I execute nodetool on a machine different from a
machine the node runs on? You see? E.g. admins can just have some jumpbox to a
cluster they manage, they do not necessarily have access to nodes themselves.
So they would not be able to read it.
I think that we could return the whole string and it would be then parsed on
the nodetool. It might be also possible to send the format as an agrument to
MBean method and server would returned that formatted already.
If the string is "too big", we could compress it on server and decompress it on
the client but this is not the objective of this ticket and it is a subject of
further discussion.
> Introduce nodetool command to dump cluster metadata
> ---------------------------------------------------
>
> Key: CASSANDRA-19393
> URL: https://issues.apache.org/jira/browse/CASSANDRA-19393
> Project: Cassandra
> Issue Type: Improvement
> Components: Transactional Cluster Metadata
> Reporter: n.v.harikrishna
> Assignee: n.v.harikrishna
> Priority: Normal
> Time Spent: 10m
> Remaining Estimate: 0h
>
> Adding a nodetool command for dumping cluster metadata would be handy when
> compared to making a JMX call. Nodetool has two more commands related to
> cluster metadata service (describecms and reconfigurecms). These commands can
> be grouped for simplicity.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
