Hey, I work in IT support for a post production facility which has grown a lot over the past few years, and these days we are considering purchasing PRTG. We run a lot of services (transcode servers, FTP servers, several databases, 3D render farm, etc), and my manager has asked for the implemented solution to offer a dashboard that displays a list of every service, with a green/yellow/red light that indicates when a service is down. This is meant to be a map for him and other less-tech-oriented staff at our company -they don't need to know where in the infrastructure the problem has occurred, but just what's actually affected by it.

e.g. if the service in question is an FTP server we are hosting for a client, it can fail because the actual machine running it is down, because the switch is broken, because of firewall issues or due to internet connection problems. This means I would need to group the four related hosts (server, switch, firewall, router) together, so that when any of them is down, the service on his dashboard called "FTP" would appear red.

I have tried creating said dashboard as a map, but I don't need for him to see the actual hardware components or dependencies involved, just the bottom line. Something like these examples from apple's or AWS sites: https://www.apple.com/support/systemstatus/ http://status.aws.amazon.com/

What is the best way to achieve this using PRTG? My team really wants to get going with purchase, but we need to be sure this is possible to get his approval first.

Thanks, Michael

Article Comments

If PRTG monitors the FTP server, would this sensor not automatically go down if one of the components in its path (switch, router) fails?


Mar, 2015 - Permalink

Yes, but my manager doesn't need to see a list of every sensor we have (since we have around 300 of them).. he just want to have a representation of the actual logical services we provide.

Let my try another example, since FTP might not be the best way to explain this.. say I want to group the 3D render farm - We have 2 controllers + 12 render nodes. If one of them goes down the rest might still work, but I want the service ("Render farm") as a whole to show an alert - not only the sensor for the specific node.

I'm sorry I'm not very good in describing this in english, if you look at the links I gave as examples in my original massage - that is basically what I want to achieve. Hope that's more clear :)

Thanks!


Mar, 2015 - Permalink

Did you have a look at the Sensor Factory Sensor?

If you scroll down on this page to

Define Sensor Status—status() Function

You will find this example

status(2031) OR status(2044)

This changes the factory sensor to a Down status if at least one
of the sensors with ID 2031 or ID 2044 is Down.
Otherwise the factory sensor shows an Up status.

Mar, 2015 - Permalink

Great, this seems like what I've been looking for. Will try it out.

Thanks again! :)


Mar, 2015 - Permalink