Skip to content

Commit 4ab062c

Browse files
zifeozifeo
committed
feat: improve migration
1 parent 2a81964 commit 4ab062c

File tree

2 files changed

+85
-60
lines changed

2 files changed

+85
-60
lines changed

‎README.md‎

Lines changed: 78 additions & 59 deletions
Original file line numberDiff line numberDiff line change
@@ -52,11 +52,11 @@ kubectl --kubeconfig single-server.rke2.yaml get nodes
5252
# k8s-pool-a-1 Ready <none> 119s v1.21.5+rke2r2
5353
# k8s-server-1 Ready control-plane,etcd,master 2m22s v1.21.5+rke2r2
5454

55+
# get SSH and restore helpers
56+
terraform output -json
57+
5558
# on upgrade, process node pool by node pool
5659
terraform apply -target='module.rke2.module.servers["server-a"]'
57-
# for servers, apply on the majority of nodes, then for the remaining ones
58-
# this ensures the load balancer routes are updated as well
59-
terraform apply -target='module.rke2.openstack_lb_members_v2.k8s'
6060
```
6161

6262
See [examples](./examples) for more options or this
@@ -68,78 +68,40 @@ Note: it requires [rsync](https://rsync.samba.org) and
6868
can disable this behavior by setting `ff_write_kubeconfig=false` and fetch
6969
yourself `/etc/rancher/rke2/rke2.yaml` on server nodes.
7070

71-
## Infomaniak OpenStack
72-
73-
A stable, performant and fully equipped Kubernetes cluster in Switzerland for as
74-
little as CHF 26.90/month (at the time of writing):
75-
76-
- load-balancer with floating IP (perfect under Cloudflare proxy)
77-
- 1 server 2cpu/4Go (= master)
78-
- 1 agent 2cpu/4Go (= worker)
79-
80-
| Flavour | CHF/month |
81-
| ---------------------------------------------------------------------------------- | --------- |
82-
| 2×5.88 (instances) + 0.09×2×(4+6) (block storage) + 3.34 (IP) + 10 (load-balancer) | 26.90 |
83-
| single 2cpu/4go server with 1x4cpu/16Go worker | ~37.— |
84-
| 3x2cpu/4go HA servers with 1x4cpu/16Go worker | ~50.— |
85-
| 3x2cpu/4go HA servers with 3x4cpu/16Go workers | ~85.— |
86-
87-
See their technical [documentation](https://docs.infomaniak.cloud) and
88-
[pricing](https://www.infomaniak.com/fr/hebergement/public-cloud/tarifs).
89-
90-
## More on RKE2 & OpenStack
91-
92-
[RKE2 cheat sheet](https://gist.github.com/superseb/3b78f47989e0dbc1295486c186e944bf)
71+
## Restoring a backup
9372

9473
```
95-
# alias already set on the nodes
96-
crictl
97-
kubectl (server only)
98-
99-
# logs
100-
sudo systemctl status rke2-server.service
101-
journalctl -f -u rke2-server
102-
103-
sudo systemctl status rke2-agent.service
104-
journalctl -f -u rke2-agent
105-
106-
less /var/lib/rancher/rke2/agent/logs/kubelet.log
107-
less /var/lib/rancher/rke2/agent/containerd/containerd.log
108-
less /var/log/cloud-init-output.log
109-
110-
# restore s3 snapshot (see restore_cmd output of the terraform module)
74+
# ssh into one of the server nodes (see terraform output -json)
75+
# restore s3 snapshot (see restore_cmd output of the terraform module):
11176
sudo systemctl stop rke2-server
11277
sudo rke2 server --cluster-reset --etcd-s3 --etcd-s3-bucket=BUCKET_NAME --etcd-s3-access-key=ACCESS_KEY --etcd-s3-secret-key=SECRET_KEY --cluster-reset-restore-path=SNAPSHOT_PATH
11378
sudo systemctl start rke2-server
114-
# remove db on other server nodes
79+
# exit and ssh on the other server nodes to remove the etcd db
80+
# (recall that you may need to ssh into one node as a bastion then to the others):
11581
sudo systemctl stop rke2-server
116-
sudo rm -rf /var/lib/rancher/rke2/server/db
82+
sudo rm -rf /var/lib/rancher/rke2/server
11783
sudo systemctl start rke2-server
11884
# reboot all nodes one by one to make sure all is stable
11985
sudo reboot
120-
121-
# check san
122-
openssl s_client -connect 192.168.42.3:10250 </dev/null 2>/dev/null | openssl x509 -inform pem -text
123-
124-
# defrag etcd
125-
kubectl -n kube-system exec $(kubectl -n kube-system get pod -l component=etcd --no-headers -o custom-columns=NAME:.metadata.name | head -1) -- sh -c "ETCDCTL_ENDPOINTS='https://127.0.0.1:2379' ETCDCTL_CACERT='/var/lib/rancher/rke2/server/tls/etcd/server-ca.crt' ETCDCTL_CERT='/var/lib/rancher/rke2/server/tls/etcd/server-client.crt' ETCDCTL_KEY='/var/lib/rancher/rke2/server/tls/etcd/server-client.key' ETCDCTL_API=3 etcdctl defrag --cluster"
126-
127-
# increase volume size
128-
# shutdown instance
129-
# detach volumne
130-
# expand volume
131-
# recreate node
132-
terraform apply -target='module.rke2.module.servers["server"]' -replace='module.rke2.module.servers["server"].openstack_compute_instance_v2.instance[0]'
13386
```
13487

13588
## Migration guide
13689

13790
### From v2 to v3
13891

13992
```
140-
# use the previous patch version to setup an additional san: 192.168.42.4
141-
# this will become the new VIP inside the cluster and replace the load-balancer
142-
# run an full upgrade with it, then switch to the new major, then you can upgrade
93+
# use the previous patch version (2.0.7) to setup an additional san for 192.168.42.4
94+
# this will become the new VIP inside the cluster and replace the load-balancer:
95+
source = "zifeo/rke2/openstack"
96+
version = "2.0.7"
97+
# ...
98+
additional_san = ["192.168.42.4"]
99+
# run an full upgrade with it, node by node:
100+
terraform apply -target='module.rke2.module.servers["your-server-pool"]'
101+
# and so on for each node pool
102+
# you can now switch to the new major:
103+
source = "zifeo/rke2/openstack"
104+
version = "3.0.0"
143105
# 1. create the new external IP for servers access with:
144106
terraform apply -target='module.rke2.openstack_networking_floatingip_associate_v2.fip'
145107
# 2. pick a server different from the initial one (used to bootstrap):
@@ -257,3 +219,60 @@ terraform state rm module.rke2.openstack_networking_floatingip_v2.external
257219
# 7. continues with other nodes step-by-step and ensure all is up-to-date with a final:
258220
terraform apply
259221
```
222+
223+
## Infomaniak OpenStack
224+
225+
A stable, performant and fully equipped Kubernetes cluster in Switzerland for as
226+
little as CHF 16.90/month (at the time of writing):
227+
228+
- 1 floating IP for admin access (ssh and kubernetes api)
229+
- 1 server 2cpu/4Go (= master)
230+
- 1 agent 2cpu/4Go (= worker)
231+
232+
| Flavour | CHF/month |
233+
| ------------------------------------------------------------- | --------- |
234+
| 2×5.88 (instances) + 0.09×2×(4+6) (block storage) + 3.34 (IP) | 16.90 |
235+
| single 2cpu/4go server with 1x4cpu/16Go worker | ~27.— |
236+
| 3x2cpu/4go HA servers with 1x4cpu/16Go worker | ~40.— |
237+
| 3x2cpu/4go HA servers with 3x4cpu/16Go workers | ~75.— |
238+
239+
You may also want to add a load-balancer and bind an additional floating IP for
240+
public access (e.g. for an ingress controller like ingress-nginx), that will add
241+
10.00 (load-balancer) + 3.34 (IP) = CHF 13.34/month.
242+
243+
See their technical [documentation](https://docs.infomaniak.cloud) and
244+
[pricing](https://www.infomaniak.com/fr/hebergement/public-cloud/tarifs).
245+
246+
## More on RKE2 & OpenStack
247+
248+
[RKE2 cheat sheet](https://gist.github.com/superseb/3b78f47989e0dbc1295486c186e944bf)
249+
250+
```
251+
# alias already set on the nodes
252+
crictl
253+
kubectl (server only)
254+
255+
# logs
256+
sudo systemctl status rke2-server.service
257+
journalctl -f -u rke2-server
258+
259+
sudo systemctl status rke2-agent.service
260+
journalctl -f -u rke2-agent
261+
262+
less /var/lib/rancher/rke2/agent/logs/kubelet.log
263+
less /var/lib/rancher/rke2/agent/containerd/containerd.log
264+
less /var/log/cloud-init-output.log
265+
266+
# check san
267+
openssl s_client -connect 192.168.42.3:10250 </dev/null 2>/dev/null | openssl x509 -inform pem -text
268+
269+
# defrag etcd
270+
kubectl -n kube-system exec $(kubectl -n kube-system get pod -l component=etcd --no-headers -o custom-columns=NAME:.metadata.name | head -1) -- sh -c "ETCDCTL_ENDPOINTS='https://127.0.0.1:2379' ETCDCTL_CACERT='/var/lib/rancher/rke2/server/tls/etcd/server-ca.crt' ETCDCTL_CERT='/var/lib/rancher/rke2/server/tls/etcd/server-client.crt' ETCDCTL_KEY='/var/lib/rancher/rke2/server/tls/etcd/server-client.key' ETCDCTL_API=3 etcdctl defrag --cluster"
271+
272+
# increase volume size
273+
# shutdown instance
274+
# detach volumne
275+
# expand volume
276+
# recreate node
277+
terraform apply -target='module.rke2.module.servers["server"]' -replace='module.rke2.module.servers["server"].openstack_compute_instance_v2.instance[0]'
278+
```

‎outputs.tf‎

Lines changed: 7 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -11,10 +11,16 @@ output "lb_subnet_id" {
1111
}
1212

1313
output "restore_cmd" {
14-
value = "sudo systemctl stop rke2-server && sudo rke2 server --cluster-reset --etcd-s3 --etcd-s3-bucket=${local.s3.bucket} --etcd-s3-access-key=${local.s3.access_key} --etcd-s3-secret-key=${local.s3.access_secret} --cluster-reset-restore-path=[filename]"
14+
value = "sudo rke2 server --cluster-reset --etcd-s3 --etcd-s3-bucket=${local.s3.bucket} --etcd-s3-access-key=${local.s3.access_key} --etcd-s3-secret-key=${local.s3.access_secret} --cluster-reset-restore-path=[filename]"
1515
sensitive = true
1616
}
1717

18+
output "ssh_cmd" {
19+
value = "ssh -o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=accept-new -o ForwardAgent=yes ubuntu@${local.external_ip}"
20+
sensitive = true
21+
}
22+
23+
1824
output "ssh_config" {
1925
value = <<EOF
2026
Host ${var.name}

0 commit comments

Comments
 (0)