Forum Discussion

Wade's avatar
9 years ago

Add concept of "stopped" devices

We have an internal app that builds servers in AWS. It also controls the scheduling of downtime and is used by users to bring hosts down and up when needed in lieu of them having to do this in the AWS console. It also uses the LM api to add / remove devices in LM and to add/remove/configure properties so the devices get appropriate datasources in LM (such as tomcat, JVM, etc.)

We need the "concept" of a stopped device with appropriate icon in LM, so we can have our app tell LM the device is stopped - and then have LM display the "stopped" icon vs. "dead" icon, regardless of the fact that either state results in a No Data condition. Put another way, if the (new) "Stopped" property flag was set to "true", then LM would display stopped vs. dead - essentially overriding the normal display of "dead" in a No Data received situation. SDT isn't suitable because these devices are not on a schedule. They are brought down when not needed to control costs in AWS, and brought back up when they are needed.

As it it is now, there is no way to distinguish stopped devices from dead devices in the UI, or even in reports.

This feature would allow us to easily see the difference between intentionally stopped devices vs. dead devices (i.e. a problem) in the UI without the need to run reports and potentially compare the report to servers in the AWS consoles. The visual display would be very valuable.

  • Sarah_Terry's avatar
    Sarah_Terry
    Icon for Product Manager rankProduct Manager

    Are your EC2 instances monitored in LM as cloud devices (added via the Add AWS Account) or as regular devices (added via the normal add a device wizard)?  For EC2 instances monitored as cloud devices via our AWS Monitoring, we display the instance state as a property on the info tab - system.aws.stateName.  When an instance is stopped in AWS, you should see system.aws.stateName = stopped on the Info tab in LM.