4 Replies Latest reply on Dec 2, 2010 5:06 PM by byrona

    Assistance with customizing the Web experience

    JaredC

      Hi Everyone, I am new to both Orion and thwack.  I've been on the other side of Orion for several years, but I've never had to implement or support the application before; it was just one of my tools as a NOC engineer.

      I've already imported all of the nodes into the database so data is being collected. 

      After showing management the "summary" page on the Web, I could tell it wasn't what they were hoping for.

      I will need to completely tear down and redesign the summary page to be more high level, but include very detailed SNMP trap and syslog events.

      Where should I begin?  I've been to the Orion Website Administration page, but it doesn't look to be as customizable as I'm going to need.

      Any help or ideas is greatly appreciated!

      thanks,

      -Jared Castiglione

        • Re: Assistance with customizing the Web experience
          byrona

          Could you please provide a summary of what exactly you are expecting your summary page to communicate?  Who will be the audience, how will they be using it, what do they want to be able to know when they look at it?

          Also, I should point out that you say "more high level" as compared to the default which is pretty high level; however, you also mention SNMP Trap and Syslog which is the exact opposite of high level.

            • Re: Assistance with customizing the Web experience
              JaredC

              Thanks for the quick reply.  I get the impression that the front page needs to be able to give a lot of information through red/amber/green status lights.  I say SNMP and Syslog because one of the requirements is to have Orion show when a user locked his/her account by logging in incorrectly 5 times in a row, and when an admin has incorrectly logged in to any server is a specific domain.

              The overall experience on the page needs to be simple and attention-grabbing, with abilities to either drill down from there, or search for any more detailed information about an interface, node, or application status.

              The audience will be a mix of upper IT management and entry-level Help Desk personnel.  SMEs will use the more advanced views on other tabs.

              Does this help clarify what I need to accomplish?

              Thanks again,

              -Jared

                • Re: Assistance with customizing the Web experience
                  lhorstma

                  I use custom properties for this and an All-Nodes tree sorted by those properties. So management can open the page, see that datacenter 1 is green, and datacenter 2 is yellow, etc. It can then be drilled down to the device level to see details. You could do this with maps too, management types like maps. "Oh no, New York City is red! I need to call the NOC to see what's wrong!". Do the same thing with APM, organize by application so managers can look and see, "oh no, Exchange is yellow! I need to call the NOC!". You can put all kinds of stuff in custom properties to give your users a view that's meaningful to them, you just need to play with it.

                  • Re: Assistance with customizing the Web experience
                    byrona

                    A few different ideas come to mind on how this could be done at a very high level with Green/Yellow/Red lights.

                    You can use the groups feature that was implemented in version 10.1 grouping your systems and applications in such a way as to represent the different systems and servers as they pertain to your business.  You can do this manually or dynamically using things such as Custom Properties.

                    Alternatively, you could use Network Atlas and create embedded maps for drill down and have the main page be nothing more than a giant high level map.

                    For the Failed Logins, to represent this with a color change you will need to have Orion APM and use an Event Log (assuming these logins are logged on an Active Directory server) monitor that looks for the specific logs.

                    The real problem with doing something like this is that the health of a complex network is hard to boil down to a simple Green/Yellow/Red, I have personally had difficulty explaining that to management on more than one occasion.

                    These are a few ideas to get you going, hope this helps.