Ad Widget

Collapse

post-installation quickstart guide

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • smokey
    Junior Member
    • May 2006
    • 3

    #1

    post-installation quickstart guide

    Hello,

    First day with zabbix, so please forgive the stupidity in this post....

    I just got zabbix installed here at work using the installation section of the manual. Very good instructions, and the installation went smoothly.

    But I got stuck after logging in as an admin. I wanted to setup a basic test for one host, but couldn't find simple instructions for doing this. I managed to create a bogus host (Host.Unix), got some output in the logs about how zabbix couldn't find host "Host.Unix", then deleted the template "Host.Unix". Oops!

    Is there such a "quickstart guide" somewhere that I just missed?

    In addition to not knowing how to setup hosts (and host groups) properly, I guess I now also need to learn how to re-create the Host.Unix template that I somehow managed to delete.

    I was very surprised at the lack of post-installation instructions (ie. a configuration quickstart guide), especially given the awesomeness of the installation instructions. The great install instructions have lowered the entry barrier to zabbix, but without good tips on setting up a few basic hosts, I quickly got lost.

    A tutorial to walk the new user through setting up monitoring for one web server would give said user a nice overview of zabbix's features, layout, terminology, etc. It would be a nice intro before the new user started looking into more complex configurations and/or learning about the detailed config options for ITEMS, TRIGGERS, etc.

    Thanks in advance for any "getting started" pages you can point me to.

    Looking forward to (hopefully) replacing nagios with zabbix!

    Greg
  • Alexei
    Founder, CEO
    Zabbix Certified Trainer
    Zabbix Certified SpecialistZabbix Certified Professional
    • Sep 2004
    • 5654

    #2
    We have two highest priorities at the moment: outstanding issues (readyness for 1.1) and documentation. Hopefully the documentation will be significantly improved very soon.
    Alexei Vladishev
    Creator of Zabbix, Product manager
    New York | Tokyo | Riga
    My Twitter

    Comment

    • bdub
      Junior Member
      • Apr 2006
      • 24

      #3
      Quick start in 10 seconds or less:

      I've been using zabbix extensivley for the last two weeks, and have made great strides in figuring a lot of things out, so I'll try and share what I've learned with you.

      In this setup, i'm going to assume that you have the zabbix agent installed and running on your target host, and that you have it set up for passive checks. (uncomment DisallowActive). You'll also need to uncomment the UserDirectives for mysql at the bottom and change them to use your database user/password.

      The setup of hosts on zabbix is as follows:

      You need to first create the host. I click the checkbox 'use the IP address' and specify both the Hostname and IP of the server I want to monitor. This allows me to have simple hostnames like "testbed" and to specify and IP address that it will work with in the background.

      Once that is done, you'll need to create an item to monitor. The templates that are available provide many useful hints as to some of the valid item keys that you can use. At any rate, lets say you want to do a check to verify MySQL is up and running properly. Go to the items screen, and make sure you have the host you've created selected. Now click 'add item'. Once you've done that, you'll get a screen where you can name this item, as well as provide a key for it. Those default values should be self-explanatory, and they have all worked for me so far. The two fields you need to be concerned with are the name, where you can provide a simple name for the item, and key, where you specify which item type you want to use. Be careful to use an easy to understand naming convention for your items, as once you have a lot of servers running you'll want to know which one belongs to which servers. I use <hostname> - <service/item> check as mine. So in our example, you'll want your name to read 'testbed - MySQL availability' or something to the effect of that. Click save, and then triggers off the top bar.

      Create a new trigger, and assign it a similar type name from before. Make sure the type is set to "Zabbix Agent". Once you've got that done, the trigger expression for this check is as follows:

      {testbed:mysql.ping.last(0)}=0

      The setup for that expression can be found in the documentation under triggers.

      A simple useful expression might look like:

      {<server>:<key>.<function>(<parameter>)}<operator> <const>
      Basically, the mysql.ping item returns a 1 if MySQL is up and running, and a 0 if it is not. Therefore, your trigger is set to go off if during the last check your expression came back as 0, indicating that mysql was not running. To verify that this is working, click the Monitoring tab, followed by Latest Data.

      Some of the more useful links I have found over the last two weeks are:

      Trigger Config
      Item Types
      Converting 1.0 items to 1.1bx syntax

      As for your unix host template, maybe you can reinstall the database before you get into configuring everything?

      Also, sorry for any minor errors in field names or anything like that in my above tutorial, as I would need to actually VPN into work to look at it, and since I'm lazy on Sunday mornings, it's all done from memory.

      Anyways, I hope this helps a little.

      Comment

      • smokey
        Junior Member
        • May 2006
        • 3

        #4
        thanks much!

        bdub, your message helped tremendously! thank you for taking the time on a sunday -- I understand the "lazy sunday" part very well. Also, I did end up re-populating the database and starting over since I didn't have any configs that I couldn't lose.

        alexei -- thanks for your reply. I noticed there was a wiki, but that it doesn't have as much information as the "official docs". Is this by design? I wonder if putting "official documentation" on the wiki would ease the burden on core developers. It seems like there would be willing documentation contributors from forum participants.

        Comment

        • avmnd.com
          Junior Member
          • Jul 2006
          • 10

          #5
          Originally posted by bdub
          I've been using zabbix extensivley for the last two weeks, and have made great strides in figuring a lot of things out, so I'll try and share what I've learned with you.

          In this setup, i'm going to assume that you have the zabbix agent installed and running on your target host, and that you have it set up for passive checks. (uncomment DisallowActive). You'll also need to uncomment the UserDirectives for mysql at the bottom and change them to use your database user/password.

          The setup of hosts on zabbix is as follows:

          You need to first create the host. I click the checkbox 'use the IP address' and specify both the Hostname and IP of the server I want to monitor. This allows me to have simple hostnames like "testbed" and to specify and IP address that it will work with in the background.

          Once that is done, you'll need to create an item to monitor. The templates that are available provide many useful hints as to some of the valid item keys that you can use. At any rate, lets say you want to do a check to verify MySQL is up and running properly. Go to the items screen, and make sure you have the host you've created selected. Now click 'add item'. Once you've done that, you'll get a screen where you can name this item, as well as provide a key for it. Those default values should be self-explanatory, and they have all worked for me so far. The two fields you need to be concerned with are the name, where you can provide a simple name for the item, and key, where you specify which item type you want to use. Be careful to use an easy to understand naming convention for your items, as once you have a lot of servers running you'll want to know which one belongs to which servers. I use <hostname> - <service/item> check as mine. So in our example, you'll want your name to read 'testbed - MySQL availability' or something to the effect of that. Click save, and then triggers off the top bar.

          Create a new trigger, and assign it a similar type name from before. Make sure the type is set to "Zabbix Agent". Once you've got that done, the trigger expression for this check is as follows:

          {testbed:mysql.ping.last(0)}=0

          The setup for that expression can be found in the documentation under triggers.



          Basically, the mysql.ping item returns a 1 if MySQL is up and running, and a 0 if it is not. Therefore, your trigger is set to go off if during the last check your expression came back as 0, indicating that mysql was not running. To verify that this is working, click the Monitoring tab, followed by Latest Data.

          Some of the more useful links I have found over the last two weeks are:

          Trigger Config
          Item Types
          Converting 1.0 items to 1.1bx syntax

          As for your unix host template, maybe you can reinstall the database before you get into configuring everything?

          Also, sorry for any minor errors in field names or anything like that in my above tutorial, as I would need to actually VPN into work to look at it, and since I'm lazy on Sunday mornings, it's all done from memory.

          Anyways, I hope this helps a little.
          Is there a way to move agents from the server to hosts without downloading the code on the host servers and doing the make/install etc. Would link static handle that... what process did you go through?

          Comment

          Working...