We have one core and 4 remote probes with a few thousand sensors each. I need to copy the groups/devices/sensors from each remote probe to remote probes in an alternate locate for fault tolerance/high availability. I tried to clone a group which worked but when I cloned the group again it created a new identical group. Our environment is somewhat dynamic and the backup site will need to be updated regularly so this method will not work. Here is what I fed it to create the group (and attempted to first delete the group then re-create with error)

/api/deleteobject.htm?id=60270&approve=1 /api/duplicateobject.htm?id=8513&name=CloneTest1&targetid=60270

How should I architect this so that I have two sites with identical data as an active/passive setup? Do I need to move the devices off of remote probes and onto cores and have multiple cores with HA pairs instead?

Article Comments

Hi Tvc15,

I currently see two ways to accomplish this:

  • You can either use the cluster feature of PRTG to install one master server and a failover node. Every remote probe will also be able to send the results to both cluster nodes. The downside of this is, if I understood you correctly, that this scenario will not cover the outage of a remote probe.
  • You could organize the structure of the remote probe so that all objects are located below one group. It will then be possible to move this group to another device using a single API call. You can move this group to a remote probe, which is idling without sensors until the API command is executed. The command could look like this, replace the : 
http://<your-prtg>/moveobjectnow.htm?id=<object-id>&targetid=<id-of-target-folder>

You can then create a HTTP action to trigger the call.

Best regards, Felix


Jan, 2017 - Permalink

But I dont want to move the group I want a copy that is paused, and I want to re-copy everyday so that the copy reflects the current state of the originating remote probe. Thoughts on that?


Jan, 2017 - Permalink

Hi Tvc15,

Thanks for the explanation. What exactly do you want to achieve with the new group after cloning the object? If you are looking for a way to backup the configuration of the probe, cloning the objects will not be required as the configuration is always stored on the core server. If one remote probe fails, you can start cloning the devices to the new probe, they will not be lost.

Please consider that some sensor types cannot be cloned, as they're probe specific (like flow sensor).

Which error message do you receive when you execute the API command to create the clones?

Best regards, Felix


Jan, 2017 - Permalink

Well I wanted to create a duplicate probe so if one went down I could bring this second clone up with little to no manual labor. It sounds like I could have a remote probe online with no sensors and change the GID to the failed remote probe and accomplish the same thing.


Jan, 2017 - Permalink

Hi,

This will work, good idea! The idling probe will then automatically get the config of the other probe and continue the monitoring.

Best regards, Felix


Jan, 2017 - Permalink