Go to All Forums

chained thresholds and IT automations

Hi folks, has anyone set up a "chained" or "escalating" responses to, let's say, an outage?  You can set up multiple automations, but I want them to occur in a waterfall fashion.  For example:

-- If apache fails (threshold), then start apache service (IT automation)
-- --  If that doesn't succeed, then restart the server (IT automation)
-- -- --  If that doesn't succeed, then restart AWS EC2 instance (IT automation)

Anyone set something like this up? Can you share your setup? Or, where do I start?


Thanks! Kris

Like (1) Reply
Replies (4)

Re: chained thresholds and IT automations

Hi,

Currently, our IT Automation feature does not support checking of one automation's result before triggering the second automation. We will take this up as a feature request and update this post once it's live. 

However, the first two steps can be combined together using a single server script automation and can be associated with the apache process, for the process down attribute.

Please look at the above image for reference, in which whenever the NGINX process goes down it'll trigger Action Script1. This will execute the command to start the NGINX process and if the command fails, it will restart the corresponding server.

Please feel free to post your comments on this.

 

Thanks,

Vinoth M

 

 

Like (1) Reply

Re: Re: chained thresholds and IT automations

Thanks Vinoth! I've setup that standard automation and it's worked great. I'll keep an eye open for a possible "chained" feature too. Thanks! Kris

Like (0) Reply

Re: Re: Re: chained thresholds and IT automations

I spoke too soon. I moved my IT automation to attached to the "process" threshold (versus apache "plugin"). but it simply doesn't fire when I stop apache.  Any ideas?  See attached.

 

 

Attachments
hpt2.png
Size: 72.75 KB
shpt1.png
Size: 60.57 KB
Like (0) Reply

Re: Re: Re: Re: chained thresholds and IT automations

Hi Kris,

Can you change the notification type from DOWN to TROUBLE and check again? That should solve the problem.

We are facing a temporary technical issue after our latest build update. We are working on the fix and will update this thread once that's done. Sorry for the inconvenience.

 

Yours sincerely,

Vinoth

 

Like (0) Reply

Was this post helpful?