DBP-285-research-loadbalancer #10
Conversation
| prefix = format("%s/24", data.ionoscloud_servers.k8s_nodes.servers[0].nics[local.nicIndex].ips[0]) | ||
| nicIndex = index(data.ionoscloud_servers.k8s_node.servers[0].nics.*.lan, tonumber(var.lan_id)) | ||
| prefix = format("%s/24", data.ionoscloud_servers.k8s_node.servers[0].nics[local.nicIndex].ips[0]) | ||
| node_lan_ips = flatten(data.ionoscloud_servers.k8s_nodes[*].servers[*].nics[local.nicIndex].ips[0]) |
There was a problem hiding this comment.
In this we get the ips of the alb target lan. These ips were initially for the two nodes 10.7.224.11 and 10.7.224.13, but after maintenance, they changed to 11 and 12.
I do not know why ip 12 was not at first assigned but 13. All other clusters seem not to skip any numbers. Maybe this is only something that happens when setting up a cluster/lan and will get standardized after maintenance?
There was a problem hiding this comment.
for the time being we will be careful when adding application load balancer target lans to nodes, especially if it is is a prod instance, and see which values they will receive. If they should not match after a maintenance a terraform apply will update the ips. In the meantime we have enabled a health check on the targets in the target group. So worst case is that the traffic will only be forwarded to half the nodes.
Description
extend module cidr-workaround so i know the ips of Node Lans
Add IONOS Application Loadbalancer Module
Links to Tickets or other PRs
Notes
Approval for review