Hello,
I was thinking about deploying a few EC2 instances as Clustered probes, however i can't get it to work. Is this due to the way PRTG forces you to use a specific IP set by the nic?
I've tried adding elastic IP's to my instances, and adding those to the list of allowed IP's as well.
All i get is "20 connection attempts failed with error Socket Error # 10060" ie a timeout
What am i doing wrong here? Any guides? Help!
Ps. You REALLY need to get going with Linux probes, been waiting years for this now.
Article Comments
This doesn't seem to want to work for me. I'll go through the steps I went through and maybe you can point out where I went wrong.
I firstly installed PRTG on my ec2 probe. I did this using the install from my main install as the download on the internet was a different version.
I then changed the main install to a master cluster and then changed the ec2 install to connect to a probe using the settings provided from the master. I then put the elastic ip on and changed the ip addresses on the master cluster to the external ip address. I've also looked at firewalls on both sides ensuring that traffic can pass both ways (to the extent of disabling both for a short period).
I *think* this is all I've done re prtg however its not working. As it stands EC2 is connecting to our local however our local will not connect to the EC2 probe.
Any thoughts as to potential solutions?
Dec, 2012 - Permalink
Dear MacAodh,
please note that you can either install a PRTG cluster, consisting of (at least) two core servers in clustering mode,
or you can install one PRTG core server and connect one or more remote probes to it.
In order to extend an existing monitoring configuration by a monitoring performed from an EC2 instance, you would usually install a Remote Probe inside of your EC2 instance and connect it to your (local) PRTG core server.
For details about remote probes, please see PRTG Manual: Add Remote Probe (and following sections).
Dec, 2012 - Permalink
I finally managed to sort this out.
Basicallyy, you'd follow the normal guide for setting up PRTG on EC2, but instead setup the standard-install and join a cluster (see How to set up a failover cluster in PRTG in 6 steps).
After you've joined the cluster you should assign elastic-ip's to your instances and set allow-rules for incoming traffic on port 23570, please double-check your sources.
Once this is done you need to change the active clusters, and change their internal IP's as reported by PRTG to the assigned Elastic IP's given by EC2.
This is actually fairly easy once you figure out all the kinks with the windows firewalls and amazon's 1:1 routing.
If anyone needs any more in-depth explanation, just reply to this thread and i will help you.
Jul, 2012 - Permalink