This past week, we rolled out some new functionality for Sonar monitoring checks. If you have signed in recently, you may have noticed your dashboard looks a little different.
If you click on the little camera icon, you will see the latest results of that check in both table and chart form.
Hover over the different bars in the chart to see the query response and DNS lookup times from each monitoring node. This is a great way to see the recent performance of your domain from multiple locations.
We also released three new features for Sonar checks to help fine tune monitoring and alert policies.
When you create a check, you have the option to select which monitoring nodes you want to run the check from. Sonar currently offers 80+ nodes in over two dozen locations.
Most clients will choose to monitor from a few different nodes. In the past, each node was checked simultaneously when the check was run. Now, you can choose to run checks once per site. This can reduce your monitoring usage and thin out your data for quicker analysis.
If you have used a monitoring solution before, at one point in time you’ve likely been inundated with alerts of what appeared to be downtime. In some cases, the perceived downtime is only for a few seconds (end-users usually won’t even notice). If this happens frequently, you can try our new verification policies.
When an endpoint is detected as down from any location, we check that endpoint from the next closest location to verify that it is down.
Please Note: The “next closest location” may not be in the same region where the original check was specified.
Let’s say you have a check that’s being monitored from eight different nodes. If you choose the majority policy, when the endpoint is detected as down we will check from all other specified monitoring nodes. If the majority of nodes report it as down (in this case four) you will be alerted.
If you have 14 or more nodes per check, it will be considered down when a maximum of seven nodes reports it as down.
Please Note: This setting is only available if you choose the simultaneous setting for check intervals.
This is not a new feature, but it can be very helpful if you are trying to cut down on false positive (down) reports. You can set a delay for when Sonar detects an end-point as down before it sends you an alert. You can set windows anywhere from one minute to ten minutes. Sonar will continue to check the status of your endpoint during this time. Only if it is still detected as down at the end of the window will Sonar issue an alert.
Every Sonar alert comes with a link to a snapshot of the status of your check with response times from multiple nodes. This link is shareable, but only for a limited period of time. You can use the Notification Report Timeout setting to set the expiration of this link from 30 minutes to four hours.
Please Note: You can also turn this feature off in the same menu.
If you have any new features you would like to suggest for future updates, please contact us at firstname.lastname@example.org