[ 
https://issues.apache.org/jira/browse/CASSANDRA-17948?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17613424#comment-17613424
 ] 

Stefan Miklosovic commented on CASSANDRA-17948:
-----------------------------------------------

I have put my vision into this PR (1). Please try it and tell me if there are 
any issues with that.

You can do, for example, these queries:

{code}
select timestamp,message from system_views.instance_logs where timestamp > 
'2022-10-06 10:12:42+0000' AND timestamp < '2022-10-06 10:13:30+0000' AND level 
= 'INFO' ALLOW FILTERING;
{code}

The reason there is "order_in_millisecond" column introduced is that logging 
message contains resolution in milliseconds but there are some cases when we 
have multiple messages withing that millisecond. If we didnt have 
order_in_millisecond column, it might happen that one message would replace 
another on insert as it would be of same milliseconds as the previous one. 
Hence, we have millisecond-wide partition.

I have also added the logging level among clustering columns so you can filter 
on that. I would say that using ALLOW FITLERING on tables like these is OK. 

_ If the query would fail in this case, would it be possible to tell it is 
because of the partition size?_

Is not this already case? I am not familiar with denylist mechanism but I would 
expect that if you try to insert something into a partition which is too big, 
it would tell you what partition it is? Maybe try to re-formulate your 
question, please.

(1) https://github.com/apache/cassandra/pull/1900

> Get warning and errors through virtual tables
> ---------------------------------------------
>
>                 Key: CASSANDRA-17948
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-17948
>             Project: Cassandra
>          Issue Type: New Feature
>            Reporter: Michiel Saelen
>            Priority: Normal
>
> The warnings and errors are currently only accessible through Cassandra logs. 
> Automating the monitoring of the nodes would be much easier/secure if we can 
> make use of virtual tables to get the logs.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

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

Reply via email to