Ad Widget

Collapse

Host monitoring from different location

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • stewie83
    Junior Member
    • Oct 2013
    • 9

    #1

    Host monitoring from different location

    Hello,

    I'm implementing a Demo installation of Zabbix for an important customer. One of the requirements is to be able to monitor a single host from different location. For example, to monitor the reachability of a Server from a Zabbix Node in DMZ, another in Server Farm and another on the Customer's client network. In addiction they want that the trigger notification has the detail of the single Zabbix from where the host is unreachable (and if more than one Zabbix trigger the alarm they want to have this information). I've read the documentation about Proxy and Node distributed monitoring, but I'm not really sure that this type of notification is possibile.
    I hope I was clear enough, if not so please ask for detail.

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

    #2
    You need multiple monitoring. So you can setup many Zabbix servers and then only setup agents to these Zabbix server. Is it possible in zabbix_agentd.config:
    Code:
    ### Option: Server
    #       List of comma delimited IP addresses (or hostnames) of Zabbix servers.
    #       Incoming connections will be accepted only from the hosts listed here.
    #       No spaces allowed.
    #       If IPv6 support is enabled then '127.0.0.1', '::127.0.0.1', '::ffff:127.0.0.1' are treated equally.
    #
    # Mandatory: no
    # Default:
    # Server=
    
    Server=SERVER1,SERVER2,SERVERN,...
    IMHO: this solution is overkill
    What exactly do you want to monitor? Only ping availability?
    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

    • stewie83
      Junior Member
      • Oct 2013
      • 9

      #3
      Thanks for your answer.

      I need only ping availability or maybe tcp port availability. I've thinked about creating N copies of the same host and connect each of them to a specified Zabbix Proxy to achieve the requirements, but I don't like this solution too much because it's expensive to configure and to mantain.

      Comment

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

        #4
        You can use only standard Zabbix agent, when you need only ping, tcp check.
        Create new item for Zabbix agents with correct item type (Zabbix agent active or Simple check) and use these keys for checks:
        Code:
        icmpping[<target>,<packets>,<interval>,<size>,<timeout>]
        net.tcp.port[<ip>,port]
        This is better and lighter solution as full Zabbix servers
        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

        • stewie83
          Junior Member
          • Oct 2013
          • 9

          #5
          Maybe you've not well undestood my requirements. I know how to create a Simple Check, what I need to achieve is to monitor a single host (no matter what kind of monitoring) from different location.

          EXAMPLE:

          4 Location -> DMZ, Server Farm, Remote Office Branch, VPN

          I want to check if HOST A is reachable from those 4 location and if any of them mark the HOST as unreachable the customer wants to know from where.

          Comment

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

            #6
            That's right.

            Proof of concept:
            - Zabbix server/agent in DMZ
            - Zabbix agents only in Server Farm, Remote Office Branch, VPN
            - monitored host: ping_host

            Every agent has central configured on Zabbix server:
            - item icmpping[ping_host]
            - trigger "when icmpping[ping_host] is unsuccessful" raise a problem

            If is problem, you will receive message: "[host_in_office]: Problem with icmpping[ping_host]" and you will see, where is server unreachable. If is it only one location or more.

            You don't check with icmpping agent yourself, but you check ping_host (another host).
            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

            Working...