[
https://issues.apache.org/jira/browse/CASSANDRA-18022?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jordan West updated CASSANDRA-18022:
------------------------------------
Test and Documentation Plan: Existing tests and manual demonstration /
reviewer manual test. Existing tests cover scheduled sampling only not what is
sampled.
Status: Patch Available (was: Open)
Patch: [https://github.com/jrwest/cassandra/tree/jwest/18022]
Running tests to see what I broke:
[j11|https://app.circleci.com/pipelines/github/jrwest/cassandra/127/workflows/2ecef45a-7883-48f5-98e8-1fead328ed29]
[j8|https://app.circleci.com/pipelines/github/jrwest/cassandra/127/workflows/92416dfd-13ee-4643-a21c-bb4ca30cd5e8]
A few notes:
* I didn't add latency since it exists already by query
* I opted to keep these metrics by partition key instead of by query like
latency but I am not strongly opposed to switching it. My initial
implementation pre-4.x was by key and usually I've wanted to debug by key but
for range queries I could see a case for by query. At least for tombstone and
row count.
* The currentKey check in ReadCommand instead of putting the code in
SinglePartitionReadCommand could likely be factored out if someone feels
strongly about it moving it to SinglePartitionReadCommand.
> Extend profileload to track by additional metrics
> -------------------------------------------------
>
> Key: CASSANDRA-18022
> URL: https://issues.apache.org/jira/browse/CASSANDRA-18022
> Project: Cassandra
> Issue Type: Improvement
> Components: Tool/nodetool
> Reporter: Jordan West
> Assignee: Jordan West
> Priority: Normal
>
> Add option to toppartitions to track by:
> * Row Count
> * Tombstone Count
> * SSTable Count
> * Latency
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
