[
https://issues.apache.org/jira/browse/HDDS-17168?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Chung-En Lee reassigned HDDS-17168:
-----------------------------------
Assignee: 蕭昌沅 (Vincent Hsiao)
> Expose datanode gRPC server thread pool utilization and request queue length
> ----------------------------------------------------------------------------
>
> Key: HDDS-17168
> URL: https://issues.apache.org/jira/browse/HDDS-17168
> Project: Apache Ozone
> Issue Type: Improvement
> Components: metrics, Ozone Datanode
> Reporter: Siyao Meng
> Assignee: 蕭昌沅 (Vincent Hsiao)
> Priority: Major
> Labels: pull-request-available
>
> XceiverServerGrpc runs on a Netty event loop group plus the ChunkReader
> ThreadPoolExecutor, but neither is visible in metrics. When reads slow down
> it is not possible to tell whether the datanode is saturated on request
> handling threads or on disk.
> Expose for the datanode gRPC server:
> * Active and maximum threads of the read executor.
> * Read executor queue length.
> * Number of active client connections.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]