[ 
https://issues.apache.org/jira/browse/CASSANDRA-17002?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17506577#comment-17506577
 ] 

Stefan Miklosovic edited comment on CASSANDRA-17002 at 3/14/22, 9:49 PM:
-------------------------------------------------------------------------

[~frankgh] I did few modifications here (as the last commit on top of your 
stuff,  it is rebased on top of current trunk, you might cherry-pick it)

I have basically filtered out all states which are unsupported (these X's 
states) + TOKENS. You will see it from the code. So by doing this, if there is 
some other state added, or removed,  tests would fail because the number of 
columns in your test which you are expecting (40) would not match anymore. And 
this is also your chance to take a look what state was added / removed and act 
accordingly in the tests or in the vtable class, like filtering that state out 
if not appropriate and similar.

In other words, iterating over all states and explicitly filtering out ones we 
are not interested in seems to be better approach in the long run than 
explicitly enumerating them.

[~yifanc] what do you think?

https://github.com/instaclustr/cassandra/tree/CASSANDRA-17002


was (Author: smiklosovic):
[~frankgh] I did few modifications here (as the last commit on top of your 
stuff,  it is rebased on top of current trunk, you might cherry-pick it)

I have basically filter out all states which are unsupported (these X's states) 
+ TOKENS. You will see it from the code. So by doing this, if there is some 
other state added, or removed,  tests would fail because the number of columns 
in your test which you are expecting (40) would not match anymore. And this is 
also your chance to take a look what state was added / removed and act 
accordingly in the tests or in the vtable class, like filtering that state out 
if not appropriate and similar.

In other words, iterating over all states and explicitly filtering out ones we 
are not interested in seems to be better approach in the long run than 
explicitly enumerating them.

[~yifanc] what do you think?

https://github.com/instaclustr/cassandra/tree/CASSANDRA-17002

> Gossipinfo should be available through a Virtual Table
> ------------------------------------------------------
>
>                 Key: CASSANDRA-17002
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-17002
>             Project: Cassandra
>          Issue Type: Improvement
>          Components: Feature/Virtual Tables
>            Reporter: Dinesh Joshi
>            Assignee: Francisco Guerrero
>            Priority: Normal
>          Time Spent: 4h 40m
>  Remaining Estimate: 0h
>
> The {{nodetool}} utility provides a utility to print out the {{gossipinfo}} 
> for the cluster. The output is then printed out as text to stdout and 
> consumed by operators and tools built around {{nodetool gossipinfo}}.
> With the introduction of virtual tables in Cassandra, the {{gossipinfo}} 
> virtual table can help operators and the tools built around {{gossipinfo}} to 
> consume the gossip information more easily by providing a tabular output of 
> the data, which can be easily queried by different clients.



--
This message was sent by Atlassian Jira
(v8.20.1#820001)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to