[jira] [Updated] (TC-392) Traffic Server Log aggregation in 1 min intervals for delivery service charting

["Ryan Durfey (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/TC-392?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Ryan Durfey updated TC-392:
---------------------------
         Labels: log_aggregation logs  (was: chart influx log_aggregation logs)
    Description: 
Breaking down bytes, requests,  time to serve (ttms), and users into various buckets allows for rapid charting for analytics, reporting, and monitoring
Sum of bytes, sum of requests, sum ttms(time to serve milliseconds), count chi (user IPs)
by CDN
by Service (CDN service)
by Host (server)
by Cache Group (site or geo)
by IPtype (IPv4 or IPv6)
by pssc (http return code)
by crc (cache response code)
by minute

  was:
Today, log aggregation is performed in a separate commercial system.  This should be done in an open source system and aggregated into a an open source time series database.

Once traffic logs is complete, we should begin aggregating logs on 1 min intervals for charting and graphing in a time series database like influx.db.
Key Fields to aggregate:
bandwidith, tps, http codes, crc, tier, etc...  

Start with what exists in the portal today.  We would also like to expand on this based on our monitoring experience.

    Component/s:     (was: Traffic Ops API)
                     (was: Traffic Portal)
        Summary: Traffic Server Log aggregation in 1 min intervals for delivery service charting  (was: Log aggregation in 1 min intervals for delivery service charting)

> Traffic Server Log aggregation in 1 min intervals for delivery service charting
> -------------------------------------------------------------------------------
>
>                 Key: TC-392
>                 URL: https://issues.apache.org/jira/browse/TC-392
>             Project: Traffic Control
>          Issue Type: New Feature
>          Components: Traffic Analytics
>            Reporter: Ryan Durfey
>              Labels: log_aggregation, logs
>
> Breaking down bytes, requests,  time to serve (ttms), and users into various buckets allows for rapid charting for analytics, reporting, and monitoring
> Sum of bytes, sum of requests, sum ttms(time to serve milliseconds), count chi (user IPs)
> by CDN
> by Service (CDN service)
> by Host (server)
> by Cache Group (site or geo)
> by IPtype (IPv4 or IPv6)
> by pssc (http return code)
> by crc (cache response code)
> by minute



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)