How do I configure failover in Linode using FRR?
I would like to configure Failover on Linode based on FRR protocol. I am following this doc - https://www.linode.com/docs/products/compute/compute-instances/guides/failover/#configure-failover
One of the point says to add an IP to one of the instances and it also says to configure IP sharing. The problem is when I configure IP sharing, I am unable to SSH into that node.
Moreover, I don't understand the following instruction:
On the Compute Instance that is not assigned the IP address you selected in the previous step, add that IPv4 address or IPv6 range as a Shared IP using Linode’s IP Sharing feature.
1 Reply
CassandraD
Linode Staff
While I haven't used FFR for configuring failover, it's possible that you're running into some of the same issues I talked about in this post about Lelastic.
I'll highlight a few things I'd recommend looking at:
First, I'd recommend deploying your instances with Network Helper on globally, and then turn it off for each instance when prompted by the guide (preferably after you do add a second IP, which I'll talk about next).
I'd recommend adding a second public IPv4 address to your Primary Node, which may involve opening a Support Ticket.
If you've already added the second IP, you want to make sure that the shared IP is not configured under the ETH0 interface. Both servers should have an independent IP configured under ETH0 and the shared IP configured under the lo interface. If you added the second IP to the first server AFTER turning off Network Helper for the instance, the second IP shouldn't be automatically added to ETH0.
While it's possible to do this without a third IP, you'll find this warning in our failover guide:
Warning
When IP Sharing is enabled for an IP address, all connectivity to that IP address is immediately lost until it is configured on Lelastic, FRR, or another BGP routing tool. This is not an issue when adding a new IP address, but should be considered if you are enabling IP Sharing on an existing IP address that is actively being used.
That means if something goes wrong while configuring the BGP routing tool, you'll lose access to your server. Adding a second IP to the primary server after Network Helper is turned off should prevent this.
I also wanted to address this here:
On the Compute Instance that is not assigned the IP address you selected in the previous step, add that IPv4 address or IPv6 range as a Shared IP using Linode’s IP Sharing feature.
In this case, assume the primary server has 2 IPs (one dedicated and one that it will share with the secondary server) and the secondary server has 1 dedicated IP at that point in the guide. The previous step asked you to determine which IP you want to share. Since that is on the primary server, "the Compute Instance that is not assigned the IP address you selected in the previous step" is talking about the secondary server. So that step is saying to add the shared IP to the secondary node in Cloud Manager.