Comment on I’m about to throw my entire Pihole out the window

friend_of_satan@lemmy.world ⁨11⁩ ⁨months⁩ ago

One of your biggest problems here is what we call high availability. You have a single point of failure: your one pi-hole server. For any service to be resilient, it needs to be highly available. This means full redundancy for all things in the whole stack that provide the service.

For pihole, this means running a pihole instance on two separate devices, with two separate IP addresses. Your dhcp server will send out two DNS servers with every lease. Most DNS clients will use at least two DNS servers.

If one of the servers goes down, your other, redundant pihole server will continue serving DNS.

This is why, contrary to other recommendations in this thread, I run pihole in docker on regular machines. If one of those machines dies, the other machine will continue serving requests, and it’s easy to launch the docker pihole on another machine and reconfigure my DNS server to hand out the new DNS server address.

source
Sort:hotnewtop