Page 1 of 2

Distributed Monitoring Advice/Questions

Posted: Wed Aug 27, 2014 4:41 pm
by BanditBBS
Ok, I've asked similar in the past, but new job=new requirements :)

First off, what ever came about with Nagios Core 4 remote workers working like gearman? Are we able to distribute monitoring like gearman yet using the native Core 4 workers? If not, how close is that to coming?

Ok, now for the fun question........

We host a bunch of customers in house and for those it is easy, they are all on my NagiosXI server. We also have managed customers where their servers are located in their own data centers and we just have a VPN connection between the two of us. That VPN connection is very secured and I must put a device of some sort in their DC to limit the amount of ports required. So, the question is basically, how would you handle this? I can think of many ways to do it. I would love to have a method where the configuration is centralized on the current NagiosXI server.

So, go with advice...... :)

EDIT: I also prefer active checks, just wanted to throw that out there. I like being able to recheck a few times before notifications are sent

Re: Distributed Monitoring Advice/Questions

Posted: Thu Aug 28, 2014 1:47 pm
by tmcdonald
If it weren't for having the configs central to the XI server I would say use NRPE. You could still do this with dont_blame_nrpe set to 1, but I'm not sure how this would fit into your current security policy. Shouldn't be an issue though if you lock it down to a specific IP and use encryption.

As for the remote workers, I'd have to pull a dev in for that one.

Re: Distributed Monitoring Advice/Questions

Posted: Thu Aug 28, 2014 1:51 pm
by BanditBBS
Currently right now we monitor these customers with Nimbus and a nimbus robot(think gearman worker) is located on their network.

With the lack of input so far, I'm leaning towards using gearman again...

Re: Distributed Monitoring Advice/Questions

Posted: Thu Aug 28, 2014 2:54 pm
by BanditBBS
Can anyone argue with me and tell me I am stupid for thinking of this.....

1.) Send raspberry pi to managed users
2.) Use check_by_ssh to the run check_by_ssh and/or check_nrpe and/or other plugins from the pi

This is all of course to say screw gearman and not have to add another application to the mix here.

Re: Distributed Monitoring Advice/Questions

Posted: Fri Aug 29, 2014 10:00 am
by slansing
No, I think that would be perfectly logical, I've played around with this a bit, and there have been talks the last two years at the Conference on this I believe. You could set up a VPN connection from it for the traffic.

Mike Weber's DM talk: https://www.youtube.com/watch?v=Nm9mQ6zitZs

Re: Distributed Monitoring Advice/Questions

Posted: Fri Aug 29, 2014 10:08 am
by BanditBBS
slansing wrote:No, I think that would be perfectly logical, I've played around with this a bit, and there have been talks the last two years at the Conference on this I believe. You could set up a VPN connection from it for the traffic.

Mike Weber's DM talk: https://www.youtube.com/watch?v=Nm9mQ6zitZs
Yeah, I sat through this one last year https://www.youtube.com/watch?v=ONCFbmQitOE, which gave me the idea. I had a Pi kit ordered yesterday so I can get one in house and play with it so see if it will work for us.

Re: Distributed Monitoring Advice/Questions

Posted: Fri Aug 29, 2014 11:18 am
by mikew
Once you get your Raspberry Pi and you have questions let me know as I can help with some of that.

I am working on a new project to put a device like Raspbery Pi on 500 oil rigs world wide. I am going to be using NRDS with some special config settings including the ability to store-and-forward data if the link goes down. I am hoping for a beta to be working by the conference.

Re: Distributed Monitoring Advice/Questions

Posted: Sat Aug 30, 2014 3:22 pm
by BanditBBS
Yeah, I'll let you know Mike. I think it came in Friday but the receiver wasn't in the office to open the package. Not like I would have had time anyway. Our largest customer is coming in Tuesday and I spent all day trying to get ready. (that's why you never received that email from me yet)

Edit: Yeah, it came in. I'll probably start playing with it tomorrow.

Re: Distributed Monitoring Advice/Questions

Posted: Tue Sep 02, 2014 9:33 am
by tmcdonald
I know we've strayed *slightly* off-topic since we got on the Pi topic, but if that works for your datacenter situation I say do it!

If you have Pi questions related to Nagios, post 'em here. I'm not 100% sure it would be smiled upon to talk about the Pi in general, but I do use one at home for some home automation so if you have general questions we can PM.

Re: Distributed Monitoring Advice/Questions

Posted: Wed Sep 03, 2014 1:29 pm
by BanditBBS
Dangit, the Pi isn't going to work for me. I need to run the check_oracle_health plugin which requires the oracle instant client to be installed. There is no ARM support for it, time to rethink my options....

Edit: highly considering an Intel NUC with Celeron CPU. Couple hundred more bucks, but whatever, we'll be charging the customer, lol.