Assigning an alert to a group of nodes

Perhaps(maybe), this is what you are looking for?

Create a Group of Nodes alert, using a Custom property to Identify Group name and Group Member Name With Variables - Sol…

Thank you for the link, this tells me how to configure the 'Scope of alert' whereby I can specify the Custom Property (server group) by alerting against 'Group Member'.

Then when configuring the 'actual trigger condition' there is no option to set the condition to trigger against AppInsight for SQL whereby I want to report when a SQL DB File runs Low on Space because this option isn't available when alerting against 'Group Member'.

I think you have a misconception about the way objects are handled in Solarwinds, you don't create alerts and then assign groups to those alerts. 

Lets start by defining what exactly you want and knocking this out one piece at a time. 

In the first alert you show it seems that the main thrust of what you want is just to alert when Database File Free Space is less than or equal to 5% right?

As far as I can tell none of the other lines you have on there are needed.  Do you have servers that have appinsight monitors on them that you wouldn't want to alert if db free space gets low?

Hi Mesverrum

I'm really struggling with some alerts on this newer version of SAM that were easy to create a year ago via NPM, I know the components are different but concepts should be the same. Unfortunately for me as an engineer I'm looking for Logic; create groups and assign a trigger condition, not create Custom Properties of all kinds for the alerts to hook into. I can't seem to simply do this, however, very simple to do when assigning to a single node and I have 4 alerts that are taking a long time to figure out. I shouldn't have to go on a course to do this but feel it necessary to do so just to understand what used to be basic, is no longer.

There are 120 nodes assigned to AppInsight for SQL - I have 2 alert requests from the DBAs asking for the following, but don't want all 70 SQL components monitored that are part of the AppInsight for SQL package;

1. Alert One - 1 Group of Critical SQL servers (Custom Group created) to trigger an alert if: 'Average Lock Wait Time' is equal to or more than 120 seconds
2. Alert Two - A different Group of Critical SQL servers (Custom Group created) to trigger an alert if: 'Database File-Used Space Percentage' is greater than or equal to 90
3. Alert Three - ALL 120 SQL Servers to trigger an alert if Application Pool/IIS service stops; Trigger Condition to 'Recycle/Restart App Pools/IIS

Simple 1, 2, 3 Steps for the above would be perfect :-)

Ok so the first thing is you need to disregard groups for the purposes of alerting.  I've seen lots of people fall into this trap but for whatever reason Solarwinds does not allow you to tie alert logic directly to group memberships within the UI.  You can do some round about stuff but ultimately that's not how the designers programmed it so don't fight an uphill battle trying to make it work that way. To be honest you probably will be better off just ignoring the groups feature for most purposes.

I would create a custom node property or find some other way to designate which nodes are subject to these thresholds.  For my demo I'll use the Comment property.  You can pretty easily set the property for every node in that group by going to the custom property editor at yourserver/Orion/Admin/CPE/InlineEditor.aspx

Select Group from the group by column > select your lock wait group > click the top box to select all > edit multiple values > set the property for them all in one go.

Do the same to tag your second group with a different custom property or different value.

Create a new alert, set the "I want to alert on" to Component and set up your parameters like I have here

Set up your actions and such however you want but this is all there is to it.  For the used space percentage you could do something similar.

For SAM components I prefer to just build a single alert for all components that has a trigger condition of "if any component is critical/down then email applicationowner" (applicationowner is a custom property i create on applications).  Then I just make sure that all my applications have their monitors set up correctly with appropriate thresholds (if you want to monitor something but never alert on it just erase the warning and critical thresholds).  For your use case I would then go into all the servers where I needed to trigger at 120 seconds of lock wait and edit the application, over ride the template and set the critical threshold there. Since you have a lot of servers you might be better off going the first route unless you are familiar enough to just do bulk edits in the database to change all the thresholds you want in one go.