Good Morning,
I'm having issues installing PRTG as it doesn't seem to be able to start and stop its installed services properly.
My system setup is that I have two servers (VMs) which are both domain members. None of the domain is connected to the internet. Both will have a full installation of PRTG on it, and will be configured as a cluster with one Master and one Failover. I have disabled the Firewall on both machines. Some GPOs apply, but they are either concerning Firewall access or to allow WMI access.
The issues I will describe occur on both machines.
Firstly, during installation, when PRTG is trying to start the services at the very end, it hangs for a minute or so and then I get a pop-up saying "PRTG's Local Probe did not start within 60 seconds." I then check the services and both PRTG Core and Probe Services seem to be running, so I just close this window and the rest of the "frozen" installation and I am able to move on.
Next, I try and configure the cluster. I convert one instance of PRTG to the Master using the Administration Tools and everytime there is a variety of issues; PRTG cannot stop the core server service, and then it can't start it. Then the local probe wont stop or start. I managed to fudge my way through by manually disabling the services, then restarting and then when the cluster configuration fails, go in and start them again.
Finally, when trying to convert my second PRTG to a failover, it just doesn't work altogether. No amount of manual tweaking has helped so far.
It may be possible to bumble my way through, but I am trying to present PRTG to a client as a solution and it doesn't look very professional when the basic function of the software doesn't function properly on my machines.
Any help would be appreciated.
Thanks!
Article Comments
Hello thanks for the reply.
We are running Windows 2012 R2 using Esxi 6 Hypervisor. We have 4 VCPUs, 6G Memory and 40G Hard disk. There are < 100 sensors as its just the installation baseline I am dealing with at the moment.
I tried to edit my original post, but it seems that I do not have the issue when I install PRTG on a fresh machine that is not connected to the domain. This seemed to work for a Master node configuration, and I am just about to test the same conditions for Failover node. I will report back shortly.
May, 2017 - Permalink
Update: So playing around with the fresh VMs and my existing ones, it seems to boil down to PRTG just being very *slow* at managing its services. In all cases it seems that PRTG declares a timeout before the service has had a chance to stop or start. This is problematic when Administration Tools tries to restart a service, as it gives up on stopping it and then tries to start another service over the top of the existing one.
The process I have worked out so far for configuring either Master or Failover Nodes: 1. Restart VM. 2. Go to 127.0.0.1 and wait until activation is complete, around 5 minutes or so. 3. Open Administration Tools and configure the Node. 4. Wait for the service requests to timeout and then restart the VM. Here I assume that I could probably just force a restart during the services requests as that will properly restart the services without worrying about timeout.
So the question is why the PRTG services are so slow. Is there any guidance from the community that could point me towards a solution? Could it be the way the CPUs are being rationed out to the other configured VMs?
May, 2017 - Permalink
Dear trevor-l
Those application service issue should not happen. What kind of VM are you using (number of CPU cores, model of the CPU hardware used, amount of RAM) and roughly now many sensors are you operating in your cluster?
May, 2017 - Permalink