Ad Widget

Collapse

Is there any option like 'Caching Graph'?

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • EvilEugene
    Junior Member
    • Mar 2012
    • 20

    #1

    Is there any option like 'Caching Graph'?

    Hello all.
    I have a screen which have 'many' of graphs and metric.
    It works well in 1H ~ 12H range, however, CPU spikes when I choose 1D range; It seems like DBMS uses CPUs for selecting all data for all metrics.
    and the real problem is, If my colleague open this screen with me, the CPU usages getting higher - I'm not sure but zabbix server selects metrics data everytime requested...

    so, If zabbix server caching graph result, then It may not need to select bunch of data again, thus CPU usages will be lower.

    but unfortunately, I did googled but couldn't find anything...
    Is there any solution for issue like this?

    thanks for your help
  • Colttt
    Senior Member
    Zabbix Certified Specialist
    • Mar 2009
    • 878

    #2
    yes, increase the cache on database side..
    Debian-User

    Sorry for my bad english

    Comment

    • jan.garaj
      Senior Member
      Zabbix Certified Specialist
      • Jan 2010
      • 506

      #3
      Graphs are generated on the fly by PHP (from data which can be in DB cache), you can't cache them out of the box. But you can cache them as elements, which are served by your web server. Article for nginx http://blog.sepa.spb.ru/2016/01/spee...ith-nginx.html
      Devops Monitoring Expert advice: Dockerize/automate/monitor all the things.
      My DevOps stack: Docker / Kubernetes / Mesos / ECS / Terraform / Elasticsearch / Zabbix / Grafana / Puppet / Ansible / Vagrant

      Comment

      • kloczek
        Senior Member
        • Jun 2006
        • 1771

        #4
        Originally posted by EvilEugene
        It works well in 1H ~ 12H range, however, CPU spikes when I choose 1D range; It seems like DBMS uses CPUs for selecting all data for all metrics.
        and the real problem is, If my colleague open this screen with me, the CPU usages getting higher - I'm not sure but zabbix server selects metrics data everytime requested...
        If it is iowait CPU time it means that yur DB engine spends time in waiting on results of IOs and as consequence of this it means tht you have no enough memory used by DB caches.
        http://uk.linkedin.com/pub/tomasz-k%...zko/6/940/430/
        https://kloczek.wordpress.com/
        zapish - Zabbix API SHell binding https://github.com/kloczek/zapish
        My zabbix templates https://github.com/kloczek/zabbix-templates

        Comment

        • EvilEugene
          Junior Member
          • Mar 2012
          • 20

          #5
          thanks for all your response It was really helpful.
          and at this moment, nginx caching method can be solution, I guess.
          I'll try it. thanks

          Comment

          Working...