Monitoring IIS application pool

[Bhaskar_Khanal]

Hi,

I am trying to add monitoring for all the IIS application pool. I want to get alerted when the pool is stopped/down.

What would be the ideal way of doing it?

I tried using the app agent availability check which is not always giving the right status. If the application pool is not getting any traffic for a certain time, the agent availability becomes 0% and it gives the false alert. Another way of achieving this is through the count of working process under the IIS metric but I am unable to apply the same rules to all the nodes in a particular application. I am trying to avoid creating a separate rule for individual tiers/nodes as it is going to be very cumbersome. I would really appreciate it if you could share the way to simplify this monitoring.

[Blaine_Johnson]

Hi Bhaskar,

Did you ever get a reply or come up with a workable solution?  We are at the same point as you, the working processes metric will let us know when the app pool is down, but we'll have to write a health rule for each and every app pool.  For the application in question, this would be a huge undertaking.

Thanks.

-Blaine

[iamryan]

Hi @Blaine.Johnson,

I found this AppD Documentation I wanted to share: https://docs.appdynamics.com/display/PRO45/Monitor+IIS

If you find anything that helps in it, please share those learnings as a reply to this post.

[Blaine_Johnson]

Hi Ryan,

Thanks for the document.  It is one the first documents I came to when researching monitoring app pools with AppDynamics.  While it is helpful to describe what can be monitored, what is really needed is a "best practices" document for monitoring app pools with health rules.  Both Bhaskar and I came to the same conclusion that a single health rule must be written for each and every app pool - an exhausting task.  A combined health rule can be written for a Tier, with a condition written for each app pool, but even that's a lot of work and will not fire an event for an app pool that is down if there is an open event for a down app pool within the same health rule.

Thanks.

-Blaine

[Ben_Andress]

Hey Blaine,

If you have a minute and wouldn't mind,

Could you share the details around the rules you put in place?

Within the health rule, i am having trouble finding where to specify the app pool state (even for the individual nodes).

If not i totally understand, just trying to gather as much info as possible before implementing our rules.

Thanks!

[Blaine_Johnson · Accepted solution]

Hi Ben,

Sure.  It's quite a drill down.  Choose custom for type in the Affected Entities and select application performance.  In the condition criteria, select "single metric", "value", and drill down to Application Infrastructure Performance|"your tier"|individual nodes|"your node name"|IIS|Application Pools|"your application pool name"|number of working processes .  Set the specific value to < 1.

Without a true "status" metric, we went with 0 working processes is "down".

I hope this helps.

-Blaine 

[Ben_Andress]

This helps a ton! thank you so much!

I really appreciate it!

[Morelz]

Hi Ben

To simplify the health rule to one rule, you can use the wildcard in the setup

Within the health rule you select Tier/Node Health - Hardware, select nodes instead of tiers to include in series

Then use a relative path as below

IIS|Application Pools|*|Number of working processes

This will then look at all application pools within an Application, and alert when it breaches the threshold

Ciao

[iamryan]

Hi @Blaine.Johnson,

Thanks for that info. I've passed this feedback to the docs team. In the meantime, let's see if the community is able to jump in and provide some best practices help. 

[Ben_Andress]

Has there been anything that has come of this since this point?

I am also looking for a best practice for the health rules for App pool state.

I am about to start the long and tedious process of a health rule per node, but wanted to check in on this before doing so.

thanks,