You are here

need best practice for monitoring a clients windows env behind vpn when my master/slave is redhat 6.

3 posts / 0 new
Last post
gdohmeier
gdohmeier's picture
need best practice for monitoring a clients windows env behind vpn when my master/slave is redhat 6.

We have opsview 4.4 enterprise running on redhad 6 with a master and failover master and currently 4 slaves... masters are in aws and slave are in aws and client environments - using both ssh and reverse ssh. 

Now I need to monitor a client that is all Windows (Win 2007) and its only accessible thru a Vpn (they are using Vpn that comes packaged with Win) - so none of the servers are available over ssh and I cannot put a RedHad 6 slave into this client environment.

I would like to know best practive for a solution in this case.. but in not finding much information in the opsview online docs dealing with setting up passive checks or maybe configuring a slave (but it would have to be running on Windows).  I dont like the idea of opening a vpn from master and sending checks that way.  I really want to stick with the distributed architecture and keep all the checks local in the client environment and pass them back to the master via a slave, but not sure what my options are. 

The solution that I can think of at this point is running opsview agent for windows (nsclient++) - on all the servers and passing data back to the host passive; but there is 12 hosts.   I would like to consolidate that traffic like is done with a slave, but the OS is different between my master and this client...so not sure what other options might be available.  I would prefer to know what is opsview best practice in this situation and propose that .

thanks

awijntje
awijntje's picture
Re: need best practice for monitoring a clients windows env ...

he there,

hum well in my opinion the "best-pratice" would be to deploy a slave into that environment (possibly a virtual machine) but it seems this is not possible from what you describe.

Now I personally have never been a fan of passive checks (for various reasons) so if you are looking for an active setup you might consider deploying an NRPE proxy (have a search on "nsclient++ proxy").

We use something similar to run checks against multi-homed server where certain interfaces are not directly reachable from the slave (we run the check over another "slave").

I also vageuly remember reading something about a Windows webserver package for something similar (can't remember the search terms anymore).

Hope this helps,

Alan

gdohmeier
gdohmeier's picture
Re: need best practice for monitoring a clients windows env ...

Thanks Alan, 

I had asked for the slave as 'best practice' - but I dont know that I will get it.  

I will follow up on the (nsclient++ proxy) thread to see what I can find.  This sounds like the best option at the moment.

I just had another query from one of my team about starting a vpn and doing a check over the vpn, but I have not had pleasant experiences trying to monitor over vpns in the past...  Thoughts on checks thru a vpn ?- I suppose I could make the vpn the parent to all the other checks and monitor it as well.