[
https://issues.apache.org/jira/browse/CASSANDRA-15519?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jordan West updated CASSANDRA-15519:
------------------------------------
Test and Documentation Plan: Added unit test and manually tested JMX
modifications via jconsole
Status: Patch Available (was: Open)
A couple notes for reviewers:
* I used the BiConsumer (remover) pattern in EndpointPayloadTracker to make
the code more testable but if there is concern about the memory overhead I can
change it
* Two concurrent calls to setEndpointLimit could leave the node in a state
where some endpoints have one limit and some have the other. This would require
human’s to make JMX calls concurrently so I didn’t see much value in adding
more concurrency control (especially since the fix is to just set the limit one
more time w/ human coordination). Open to revisiting though.
> Make native_transport_max_concurrent_requests_in_bytes(_per_ip) changeable at
> runtime
> -------------------------------------------------------------------------------------
>
> Key: CASSANDRA-15519
> URL: https://issues.apache.org/jira/browse/CASSANDRA-15519
> Project: Cassandra
> Issue Type: Improvement
> Components: Local/Config, Messaging/Client
> Reporter: Jordan West
> Assignee: Jordan West
> Priority: Normal
> Labels: pull-request-available
> Fix For: 4.0-beta
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> CASSANDRA-15013 added configurable global and per endpoint limits on the
> number of in flight requests, measured in bytes. During times when the
> cluster is overloaded, it can be useful to change this setting without having
> to restart the Cassandra process. Changing the limits should affect all
> existing and new connections.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
