Forum Discussion

Nader's avatar
4 years ago

Request to improve the alert rules resources/instances selection options

As of now, if I want to create an alert rule for multiple resources I need to filter/select based on something common between the devices, e.g. device name includes specific text/regex, or they are under specific resources group, or in specific location, .. etc. and I can monitor - under the same alert rule - either everything on these devices (*) or the datasources/configsources which are common between these devices e.g. Memory, CPU, HostStatus, Ping, ... etc.

But if I want to alert for a specific interface on a specific device but not to monitor the same interface (same interface name) on another device, (e.g. Trigger alert for the status of Interface Gig 1/1 on Switch_A1 but DO NOT trigger alert for the status of Interface Gig 1/1 on the Switch_B2), I can't do it in 1 alert rule, and I need to have separate alert rule for each device.

In my case, I created alert rule for the common things but I will need more than 100 alert rules to cover the specific interfaces on each device that I want to include in the alert rule since I want to monitor everything and alert only on specific things which are not common between all the devices.

So, I'm requesting to have the capability of selecting specific datasources/configsources of multiple devices (might be similar or different) and be able to put all of them in one alert rule. For example, to have a matrix (y-axix: devices, x-axis: datasources, config sources) and I tick the cells that matches the devices with the datasources I want to monitor.

 

2 Replies

  • Oh this would definitely be helpful for our junior admins and admins in general.

    I made a slight workaround to this, which is very useful if the number of devices covered by the alert rule is very high.

    1) Download the relevant devices (with Device name and Device ID, etc) in a csv
    2) Added column for where I indicate to follow the alert rule or not (leave empty if not) and save
    3) Make a script with BASH/powershell/python/go/etc and import the CSV and filter it only have the items where the last column indicator is not empty
    4) Make a Foreach, and have it run on Device ID
    5) In the Foreach, utilize the SDK or API calls to PATCH the devices, and add a property like alert.rule with the value somerule
    6) And then proceeded to create the Alert rules I needed based by specifying the property

     

    All this scripting and coding wouldn't be as necessary though if we had the option as you stated, especially if scripting is not a strong point for the admins involved.

     

     

     

  • Can we get an update regarding this feature request?

    Thanks.