[
https://issues.apache.org/jira/browse/YARN-5995?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15813739#comment-15813739
]
Sunil G commented on YARN-5995:
-------------------------------
Thank [~jianhe] for chipping in.
I also agree about the read op. We can skip them for now.
As mentioned, i think we can redefine the list of metrics as below.
- #total write operations
- Time series metric to store the cost of each write op (time duration taken
for each operation). Histogram will suit here I think.
- Write Latency (Data written for each op)
But I am also thinking in similar line. At time t1 if these datas are stored
as (#total write op, time cost of write ops, data written in last time
interval), then we can store same info at time t2 also. If we store all these
datas in time series way, then it may make external modules job easy. But its a
complex metric module to design internally. Do you whether we need to jump to
this immediately, or can do something like 3 metrics which I mentioned earlier.
Thoughts [~jianhe].
> Add RMStateStore metrics to monitor all RMStateStoreEventTypeTransition
> performance
> -----------------------------------------------------------------------------------
>
> Key: YARN-5995
> URL: https://issues.apache.org/jira/browse/YARN-5995
> Project: Hadoop YARN
> Issue Type: Improvement
> Components: metrics, resourcemanager
> Affects Versions: 2.7.1
> Environment: CentOS7.2 Hadoop-2.7.1
> Reporter: zhangyubiao
> Assignee: zhangyubiao
> Labels: patch
> Attachments: YARN-5995.0001.patch, YARN-5995.0002.patch,
> YARN-5995.patch
>
>
> Add RMStateStore metrics to monitor all RMStateStoreEventTypeTransition
> performance
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
