Difference between revisions of "Monitoring dCache with ELK"

From Public PIC Wiki
Jump to navigation Jump to search
Line 6: Line 6:
 
== Monitoring the number of transfers, stores, restores, removes & hits in time per VO ==
 
== Monitoring the number of transfers, stores, restores, removes & hits in time per VO ==
 
* Link: '''[https://kibana4.pic.es:8443/#/dashboard/dCache-Statistics-in-Time-per-VO?_g=%28%29&_a=%28filters:!%28%29,panels:!%28%28col:1,id:Count-%23hits-in-time-per-VO,row:6,size_x:6,size_y:5,type:visualization%29,%28col:7,id:Count-%23removes-in-time-per-VO,row:6,size_x:6,size_y:5,type:visualization%29,%28col:1,id:Count-%23restores-in-time-per-VO,row:11,size_x:6,size_y:5,type:visualization%29,%28col:1,id:Count-%23transfers-in-time-per-VO,row:1,size_x:12,size_y:5,type:visualization%29,%28col:7,id:Count-%23stores-in-time-per-VO,row:11,size_x:6,size_y:5,type:visualization%29%29,query:%28query_string:%28analyze_wildcard:!t,query:%27*%27%29%29,title:%27dCache%20Statistics%20in%20Time%20per%20VO%27%29 dCache Statistics in Time per VO]'''
 
* Link: '''[https://kibana4.pic.es:8443/#/dashboard/dCache-Statistics-in-Time-per-VO?_g=%28%29&_a=%28filters:!%28%29,panels:!%28%28col:1,id:Count-%23hits-in-time-per-VO,row:6,size_x:6,size_y:5,type:visualization%29,%28col:7,id:Count-%23removes-in-time-per-VO,row:6,size_x:6,size_y:5,type:visualization%29,%28col:1,id:Count-%23restores-in-time-per-VO,row:11,size_x:6,size_y:5,type:visualization%29,%28col:1,id:Count-%23transfers-in-time-per-VO,row:1,size_x:12,size_y:5,type:visualization%29,%28col:7,id:Count-%23stores-in-time-per-VO,row:11,size_x:6,size_y:5,type:visualization%29%29,query:%28query_string:%28analyze_wildcard:!t,query:%27*%27%29%29,title:%27dCache%20Statistics%20in%20Time%20per%20VO%27%29 dCache Statistics in Time per VO]'''
* In this Dashboard we can find how many stores and restores have run in the last X hours (default value is 1 week), as well as the amount of transfers and hits. Number of removed files is also shown.
+
* In this Dashboard we can find how many stores and restores have run in the last X hours (default value is 15 minutes), as well as the amount of transfers and hits. Number of removed files is also shown.
  
 
== Monitoring transfers ==
 
== Monitoring transfers ==

Revision as of 09:02, 7 August 2015

PIC & ELK

  • PIC has an ELK (ElasticSearch + LogStash + Kibana) service in order to monitor the different batch & storage services.
  • A Kibana4 Dashboard can be accessed in the following link: Kibana4

Monitoring Dashboards

Monitoring the number of transfers, stores, restores, removes & hits in time per VO

  • Link: dCache Statistics in Time per VO
  • In this Dashboard we can find how many stores and restores have run in the last X hours (default value is 15 minutes), as well as the amount of transfers and hits. Number of removed files is also shown.

Monitoring transfers

  • A plot showing information about moved data in the last X hours for the top 20 dCache pools. Here also is specified how many data has been moved by protocol and which protocols have been used per pool.
  • A plot showing information about moved data in the last X hours for all dCache pools. Here also is specified how many data has been moved by protocol and which protocols have been used per pool.
  • A stack plot showing the amount of data moved by DN.
  • A window with percentil statistics.

Transferred data by protocol

  • Link: dCache Transferred Bytes (Read & Writes) Dashboard
  • In this Dashboard we have several plots, basically showing the amount of data moved by each protocol.
  • A big plot is showing the amount of data moved by all protocols
  • We have also 2 plots showing the amount of data moved due to stores and restores.