This discussion has been locked. The information referenced herein may be inaccurate due to age, software updates, or external references.
You can no longer post new replies to this discussion. If you have a similar question you can start a new discussion in this forum.

Discussion about alerting.

I forgot to tune into the lab 54 and am now watching it (I really should find a hobby for the evenings.) But I find monitoring so interesting. Generally I find IT still interesting.  From my first memory of using a sinclair 48k, to using large personal servers that run virtual environments.  I love IT.  Even though I hate mornings, I love IT.

Anyways, alerting.

I have a grand design of how I want our new environment to look like.

My plan was to build an environment that represents a real world environment, involving dependencies, all new alerting and a new flashy SAM environment.

Anyways, looking/listening to labs 54, they were talking about measuring CPU and the intricate details associated with an alert, such as jobs. (in this instance, high CPU alert.)

My question is how do other people do alerts and even though I know that every environment is different, some ideal tips on alerting.

Thanks!

  • We try to monitor and alert on critical applications, interfaces and devices. Then use reports for others that are not as critical or that can how some down time. We are also starting to work more and more with maintenance windows to cut down on false alerts, we find if the false alerting gets out of hand no one pays attention to the alerts as they need to

  • 1. I would prefer alerts being more generic than the specific, specific alerting at times can be a mess and difficult to maintain in a huge environment

    2. We generally have 2 sets of Generic Alerts, the first one for business critical apps and devices and second set for rest of them in our environment.

    3. Its equally important to think through the custom attributes that you have in your environment, well designed custom attributes can play a major role in alerting and you can easily tag apps and devices to the generic alerts that are created in your environment.

    My take on this would be "Keep the alert as Generic as possible, reduce number of specific alerts in your environment, think through the custom attributes with respect to your environment and lay them down before you create the alerts"