[
https://issues.apache.org/jira/browse/CASSANDRA-21048?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18043145#comment-18043145
]
Alan Wang commented on CASSANDRA-21048:
---------------------------------------
Just ran CI and attached it and also changed the default to 100 SSTables.
> Log queries scanning too many SSTables per read
> -----------------------------------------------
>
> Key: CASSANDRA-21048
> URL: https://issues.apache.org/jira/browse/CASSANDRA-21048
> Project: Apache Cassandra
> Issue Type: Improvement
> Components: Observability/Logging
> Reporter: Alan Wang
> Assignee: Alan Wang
> Priority: Normal
> Fix For: 6.x
>
> Attachments: ci_summary-2.html, result_details.tar.gz
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> Debugging expensive reads currently is difficult. Attempts using fqldump,
> probabilistic tracing, tablehistograms, and debug logging could not find the
> expensive read query because we can not leave them on for extensive periods.
> We want to log queries that scan more than 50 sstables per read by default,
> and make the threshold configurable.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
