How we monitor everything

A It's never more clear how important a campus' web presence is to day-to-day operations than when the website goes down. Around eight months ago, we had edge load balancer issues which resulted in the website going down randomly for five to fifteen minutes up to twenty times a day. Since the vendor fixed the issue, we have had 100% uptime for the past four months.

To keep on top of everything, we use several tools. We took a Chrome Box from Google I/O and use it to display real-time visitors from Google Analytics, but use the Heartbeat extension to turn the real-time page into one really big number. This picture was taken at the end of the day in the summer, so our traffic is low, but we have an idea of what that number should be at any time, and if it drops significantly it could point to a problem. We also use Statusboard and an old iPad to display green/red data from Pingdom's API, our current tickets, how many emails we have, and a search of "csumb" on twitter (sometimes we know about issues from Twitter before anything else). Aside from monitoring, Pingdom also emails/text messages us if one of our websites is either down, or responding too slowly. We also depend on tools (Big Brother/Cacti) from our friends in Network Operations to know about problems going on in our servers.

With all that monitoring, we hope to know about problems before they occur, or at least don't hear about any of our sites going down by the CIO calling us. When students start to return for the fall semester, we'll be glued to our graphs and blinking lights to make sure everything runs smoothly.
The large number indicates how many real-time users are on CSUMB.EDU.

Comments

  1. Very interesting, Kevin. We're glad to hear that you're using our services to keep track of the uptime of your sites and services. Could you get in touch with us please, we'd like to find out more? Just email press@pingdom.com :)

    ReplyDelete
  2. And a pig monitors the whole process...

    ReplyDelete
  3. So you are using so many tools to monitor everything, but why? Don't you think that it is much better to install one tool which is going to be responsible for everything? I for instance use the tool Anturis which is all-in-one pack solution.

    ReplyDelete

Post a Comment