Ad Widget

Collapse

Graphs with several items on them - like currentthreads/active/maxtreads

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • sorin
    Junior Member
    • Feb 2014
    • 11

    #1

    Graphs with several items on them - like currentthreads/active/maxtreads

    I am quite new to zabbix and I am trying it especially because I wanted to monitor JMVs via JMX. THe current support in nagios is very poor and I need to update nagios configuration if I modify the memory for a JVM, and I find this unacceptable.

    I need to monitor memory/heap/current/max, some with threads and number of requests.

    Still, I find quite hard to configure it because the default templates do not work and I need to rewrite them, things like port number being called "http-bio-8080" instead of "http-8080", missing double quotes and unclear errors from zabbix, clearly not the most friendly experience. Still, I do get progress on these.

    Still, I wasn't able to find-out how to make meaningful graphs, obviously it doesn't make sense to have 3-4 different graphs for the same set of metrics.
  • jan.garaj
    Senior Member
    Zabbix Certified Specialist
    • Jan 2010
    • 506

    #2
    Clone existing template(s) and edit names, items, graphs, how do you need:

    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

    • sorin
      Junior Member
      • Feb 2014
      • 11

      #3
      Thanks, this seems the way to go but now is far too late. I've spent 4 working days installing and configuring zabbix in order to monitor JVMs and based on the progress I would have needed up to two weeks in order to obtain the information I was looking for.

      Luckily, I discovered Java Melody which seems to provide all the desired metrics and even more, and it took me less than 2 hours to configure it for the first instance.

      Due to this the experimental zabbix instance will probably be dropped because even if it was better than nagios regarding configuration, its default settings and templates do not work directly. Clearly I was looking for a solution that does not require this amount of customizations in order to make it work with, what I call, quite basic use-cases, tomcat instances.

      I am writing this not to complain, but just to give a feedback so maybe the next versions will address these problems.

      Things that are not quite friendly:
      * having to install an agent on each machine
      * having to configure the agent on each machine, so you tell who is the server. take a look at saltstack and see that the default installation is looking for "salt" server on the same network/domain and connects to it. Means that you can deploy it with a single command.
      * failure to connect to a JVM with cryptic message
      * spending a lot of time to reconfigure JVM templates just because the tomcat is running on port different than 8080
      * wrong attributes to monitor, the famous problem of the quotes... in JVM templates
      * inability to monitor two JVMs on the same host because of the duplicate templates... (need to use alternative host names to add 2nd JVM)

      Comment

      Working...