<https://docs.k3s.io/datastore/cluster-loadbalance...
# general
b
c
It is one or more additional addresses to add to the subject alternative name (SAN) attribute on the server certificate.
As shown in that example, could be the name or address of a load balancer that you are putting in front of the servers
b
Then wouldn’t it again be single point failure ?
c
wouldn’t what be a single point of failure?
b
the loadbalancer address i use for tls san
c
ideally you would use an external load-balancer that was more reliable or has better availability than any individual server node
but if your load-balancer is less reliable than just picking one of your servers to use when joining new nodes to the cluster, then I’m not sure why you’d use it either
b
so instead of it, using round robin dns can have better availability 🤔 ?
c
I don’t see many scenarios where people have load-balancers that are more of a single point of failure than a single server
I don’t know, how much of a single point of failure is your DNS?
is that more or less reliable than your load balancer?
b
i am using cloudflare
c
I would probably decide what is most reliable, and use that. If you’re using a DNS alias, you’d still need a tls san entry for the DNS alias.
use it however makes sense in your environment. that page just shows an example.
👍 1