[jira] Updated: (CHUKWA-571) Add digested system metrics to SystemMetrics demux parser

["Eric Yang (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/CHUKWA-571?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Eric Yang updated CHUKWA-571:
-----------------------------

    Attachment: CHUKWA-571.patch

Digest metrics is parse and stored without number suffix.  For example, the total bytes received by all network card is in network:RxBytes column.  Network Loopback device is omitted from the summary number.

Full list of digest metrics:

* Averaged CPU utilization:
* cpu:combined (user+sys)
* cpu:idle
* cpu:sys
* cpu:user

Sum of network devices throughput stats at interval:
* network:RxBytes
* network:RxDropped
* network:RxErrors
* network:RxPackets
* network:TxBytes
* network:TxColisions
* network:TxErrors
* network:TxPackets

Sum of disk devices throughput stats at interval:
* disk:ReadBytes
* disk:Reads
* disk:WriteBytes
* disk:Writes


> Add digested system metrics to SystemMetrics demux parser
> ---------------------------------------------------------
>
>                 Key: CHUKWA-571
>                 URL: https://issues.apache.org/jira/browse/CHUKWA-571
>             Project: Chukwa
>          Issue Type: New Feature
>          Components: MR Data Processors
>         Environment: Java 6, Mac OS X 10.6
>            Reporter: Eric Yang
>            Assignee: Eric Yang
>         Attachments: CHUKWA-571.patch
>
>
> For now, the system metrics demux parser are storing raw metrics.  Each system components are measured independently.  It would be nice to make a summary grouping for cpu, disk, network.  Instead of storing cpu:combined.0 ... cpu:combined.8, for 8 cpu cores in the system, it would be nice to have a average number for all cpus combined.  The digested metrics are easier to aggregate at cluster level.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.