[
https://issues.apache.org/jira/browse/HBASE-29084?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sanjeet Malhotra updated HBASE-29084:
-------------------------------------
Description:
When we call HTable#batch() API in HBase client, the runnable is created for
execution per region server. The runnable is executed in HTable thread pool.
Currently, no metrics are collected on the client side around time taken by
runnables in execution, waiting in queue, no. of such runnables submitted for
execution and among submitted runnables how many failed. These metrics can very
insightful when debugging latency issues.
The aim is to track metrics at Table level and metrics are recorded for each
runnable executed for a given table instance.
So far this is needed primarily for HBase 2.
was:
When we call HTable#batch() API in HBase client, the runnable is created for
execution per region server. The runnable is executed in HTable thread pool.
Currently, no metrics are collected on the client side around time taken by
runnables in execution, waiting in queue, no. of such runnables submitted for
execution and among submitted runnables how many failed. These metrics can very
insightful when debugging latency issues.
The aim is to track metrics at Table level and metrics are recorded for each
runnable executed for a given table instance.
> Track table level metrics on HBase client side
> ----------------------------------------------
>
> Key: HBASE-29084
> URL: https://issues.apache.org/jira/browse/HBASE-29084
> Project: HBase
> Issue Type: New Feature
> Reporter: Sanjeet Malhotra
> Assignee: Sanjeet Malhotra
> Priority: Major
>
> When we call HTable#batch() API in HBase client, the runnable is created for
> execution per region server. The runnable is executed in HTable thread pool.
> Currently, no metrics are collected on the client side around time taken by
> runnables in execution, waiting in queue, no. of such runnables submitted for
> execution and among submitted runnables how many failed. These metrics can
> very insightful when debugging latency issues.
> The aim is to track metrics at Table level and metrics are recorded for each
> runnable executed for a given table instance.
> So far this is needed primarily for HBase 2.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
