I am needing to setup monitoring at several remote sites (school districts) to monitor network equipment and want to have the monitoring report back to a central location at our main office. All of the remote sites have a firewall in place and so does our main office. What’s the best setup for me to do for this situation? And are there any guides that discuss this further other than the main docs
What’s the transport between the sites?
Yes all docs are at docs.librenms.org
All sites are just on their own independent ISP connection.
No vpn back to the main site?
Not currently. However, all remote sites have a fortinet firewall so VPNs could be setup
Look for the section on distributed pollers in the Docs. We are using them to monitor multiple locations.
How many locations are you monitoring?
I recommend you setup vpn connection first. Distributing polling was not meant for romote polling.
Hmm, ok. I know in the distributed polling section of the manual it mentions devices being in different physical locations though.
They are “distributed” polling to help even out the load they are not meant for “remote” polling dont get me wrong it can work…
It dose but it best with VPN thats just my two cents…
I gotcha. But doing it the VPN route, all of the remote sites would have to be setup on different internal IP schemes wouldn’t they? All the remote sites have pretty the same IP schemes that we created as a standard.
Thats outside the scope of LibreNMS. Thats something your district needs to take a hard look at your infrastructure.
I have the same use case - the ability to poll devices at client sites, all behind their own NAT firewalls.
The ‘Distributed Polling’ section of the LibreNMS documentation specifically states…
“This will provide the ability to have a single poller behind a NAT firewall monitor internal devices and report back to your central system. You will then be able to monitor those devices from the Web UI as normal.”
Are you saying this is not possible?
You can do it behind a NAT but be ready to open ports and etc…
Same use case. I think I have emailed you about my case last week. I have distributed poller at remote site, links to the Central Server via fortigate SSL vpn. i would like to monitor all switches at remote sites (but the remote poller is on different subnet as remote switches)
The Central server has full access (open all ports) to the remote site Poller only. But the Central server is not able to ping the remote switches. When I add remote devices, the central server is not able to add the remote devices because of not able to ping the adding device.
Is it possible to add a remote device to Central server if the central server is not able to ping the remote devices?
if it is not possible. is that mean I have to make Central server to ping all adding devices in a Distributed environment?
Can you please give me an advice?
You can force add device.
thanks for your reply. I tried force add devices already. but my central server is keep showing “unpolled”
my central server is in my office, and distributed poller is in one of our client office. The Central server has full access (SSL VPN) to the distributed poller, but do not has access to our client’s switches.(the distributed poller has full access to client’s switches)
Do you think it is possible to monitor remote site (our clients) machines which central server is not reachable directly?
if you have a kind of this experience, please share with me
Have you assigned that device to the correct poller?
Yes, I believe I have. I created a new poller group for the remote site (distributed) poller. So when I add a device on Central Server, i select the new group.
Do you think it is possible to monitor remote site (our clients) machines which central server is not reachable directly? Because some users are saying distributed poller does not support remote sites… i try to confirm this…
what is the purpose to have “force add device” feature on LibreNMS? is it use for adding “IP not reachable devices”? I could not find any documents to state “force add device”