Any monitoring/systems management product need to have built-in alerting capabilities. Too bad most of them aren't very useful. The tricky thing is knowing what thresholds to set. The truth of the matter is, you are monitoring so many different things in any application environment, it's onerous to set up and configure alerts, and it's almsot impossible to know what are the right levels to attach to your yellow, red, or 6 different shades of orange alerts. As a result, alerts become the least useful part of any management product, too often it's a sea of red and lots of false alarms, that IT simply becomes desensitized until something really big happens and brings down the farm. Well ExtraHop is trying to change all that. Call it Alert 2.0, truly useful alerts that you don't even have to set up! There are 2 pieces to this idea: 1. We know that most Data Center share commonalitites in their application infrastructure, i.e. everyone uses HTTP, everyone has databases. So for these common systems, ExtraHop ships with alerts targeted at them, no need to set up anything. It's just there, viola, out-of-the-box. 2. Instead of hard coded/configured threshold alerts, ExtraHop learns about your system and generates a baseline over time, when key metrics start to trend far above or below the norm, automatic trend-based alerts will fire off to let you know something abnormal is happening. No more configuration of thresholds. Finally! Alerting made simple, and hopefully truly useful. 2.0 isn't officially out yet, but some of our customers are already betaing the platform. Here's an actual story of how Trend-based Alerts helped save the day: One of our customer had an application upgrade that was causing their servers to send out lots and lots of 500 errors, worst part is it was completely intermittent! This customer was already using Keynote for end user experience monitoring. But due to the intermittent nature of the problem, Keynote did not catch it. (Monitoring via synthetic transactions have an inherent under-sampling problem.) Luckily they were running the latest firmware (2.0 beta) for the ExtraHop Application Delivery Assurance system. With our trend-based alerting, the IT group was made aware of this problem and was able to address it proactively. The customer was amazed at how easy it was for ExtraHop to surface this really important information, and all without any work on their part. Alerts with absolutely zero configuration, that's how you deliver value!
Blog
When Alerts Actually Become Useful
Any monitoring/systems management product need to have built-in alerting capabilities. Too bad most of them aren't very useful. The tricky thing is knowing what thresholds to set. The truth of the matter is, you are monitoring so many different things in any application environment, it's onerous to set up and configure alerts, and it's almsot impossible to know what are the right levels to attach to your yellow, red, or 6 different shades of orange alerts. As a result, alerts become the least useful part of any management product, too often it's a sea of red and lots of false alarms, that IT simply becomes desensitized until something really big happens and brings down the farm. Well ExtraHop is trying to change all that. Call it Alert 2.0, truly useful alerts that you don't even have to set up! There are 2 pieces to this idea: 1. We know that most Data Center share commonalitites in their application infrastructure, i.e. everyone uses HTTP, everyone has databases. So for these common systems, ExtraHop ships with alerts targeted at them, no need to set up anything. It's just there, viola, out-of-the-box. 2. Instead of hard coded/configured threshold alerts, ExtraHop learns about your system and generates a baseline over time, when key metrics start to trend far above or below the norm, automatic trend-based alerts will fire off to let you know something abnormal is happening. No more configuration of thresholds. Finally! Alerting made simple, and hopefully truly useful. 2.0 isn't officially out yet, but some of our customers are already betaing the platform. Here's an actual story of how Trend-based Alerts helped save the day: One of our customer had an application upgrade that was causing their servers to send out lots and lots of 500 errors, worst part is it was completely intermittent! This customer was already using Keynote for end user experience monitoring. But due to the intermittent nature of the problem, Keynote did not catch it. (Monitoring via synthetic transactions have an inherent under-sampling problem.) Luckily they were running the latest firmware (2.0 beta) for the ExtraHop Application Delivery Assurance system. With our trend-based alerting, the IT group was made aware of this problem and was able to address it proactively. The customer was amazed at how easy it was for ExtraHop to surface this really important information, and all without any work on their part. Alerts with absolutely zero configuration, that's how you deliver value!
