# How to monitor a device using JIT client?

**URL:** <https://discourse.sensu.io/t/how-to-monitor-a-device-using-jit-client/441>\
**Category:** Sensu Classic (EOL)\
**Created:** [June 8, 2015, 10:38pm UTC](https://discourse.sensu.io/t/how-to-monitor-a-device-using-jit-client/441 "2015-06-08T22:38:45Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Farzad\_Panahi](https://avatars.discourse-cdn.com/v4/letter/f/a183cd/32.png) [@Farzad\_Panahi](https://discourse.sensu.io/u/Farzad_Panahi)\
**Post date:** [June 8, 2015, 10:38pm UTC](https://discourse.sensu.io/t/how-to-monitor-a-device-using-jit-client/441/1 "2015-06-08T22:38:45Z")

</div>

Hi,

Now that sensu has JIT client implemented, what are the good practices for monitoring bunch of networking devices that you cannot install sensu-client on?

For instance if I want to ping monitor 10 devices, should I have 10 different check scripts (configured with “source” attribute) each targeting a device or there is a way to have a single common ping script and somehow register all those 10 devices with that common ping script?

Any help regarding how to go about monitoring agent-less devices is appreciated.

Cheers

Farzad

---

<div class="post-metadata">

**Author:** ![Kyle\_Anderson](https://avatars.discourse-cdn.com/v4/letter/k/e5b9ba/32.png) [@Kyle\_Anderson](https://discourse.sensu.io/u/Kyle_Anderson)\
**Post date:** [June 9, 2015, 5:27am UTC](https://discourse.sensu.io/t/how-to-monitor-a-device-using-jit-client/441/2 "2015-06-09T05:27:11Z")

</div>

We are still experimenting with this, but we are using this  
module/script to configure sensu server-side checks:  
[https://github.com/Yelp/puppet-monitoring\_check/blob/master/manifests/server\_side.pp](https://github.com/Yelp/puppet-monitoring_check/blob/master/manifests/server_side.pp)  
[https://github.com/Yelp/puppet-monitoring\_check/blob/master/files/check\_server\_side.rb](https://github.com/Yelp/puppet-monitoring_check/blob/master/files/check_server_side.rb)

We use common scripts (nagios check\_ping for example) but the puppet  
type allows us to deploy checks en-mass using yaml files and puppet's  
create\_resources function.

The key thing here is that we have yaml files for config data, but the  
flexibility to do even more fancy things.  
Also the script runs server-side on the cluster of sensu-servers. That  
means the check is not tied to any one server, so we get the HA aspect  
of that and no SPOF, which is cool.

The important part I think is having a good config language. In puppet  
this is hiera, which makes is really good for describing how we  
monitor differently in prod/dev, or different switches in different  
datacenters, etc  
Then you have to run the check \*somewhere\*, the sensu-servers seem  
like as good as a place as any.

> **···**
>
> On Mon, Jun 8, 2015 at 3:38 PM, Farzad Panahi \<farzad.panahi@gmail.com\> wrote:
> 
> > Hi,
> > 
> > Now that sensu has JIT client implemented, what are the good practices for  
> > monitoring bunch of networking devices that you cannot install sensu-client  
> > on?
> > 
> > For instance if I want to ping monitor 10 devices, should I have 10  
> > different check scripts (configured with "source" attribute) each targeting  
> > a device or there is a way to have a single common ping script and somehow  
> > register all those 10 devices with that common ping script?
> > 
> > Any help regarding how to go about monitoring agent-less devices is  
> > appreciated.
> > 
> > Cheers
> > 
> > Farzad