1
0
Fork 0
mirror of https://code.forgejo.org/infrastructure/documentation synced 2024-11-12 16:27:35 +00:00

split the README into separate files for clarity

This commit is contained in:
Earl Warren 2024-10-20 11:24:52 +02:00
parent ebb3667a42
commit 731d2931be
No known key found for this signature in database
GPG key ID: 0579CB2928A78A00
6 changed files with 613 additions and 664 deletions

685
README.md
View file

@ -1,493 +1,25 @@
The resources used by the infrastructure are in the https://code.forgejo.org/infrastructure/ organization.
There is a [dedicated chatroom](https://matrix.to/#/#forgejo-ci:matrix.org). A mirror of this repository is available at https://git.pub.solar/forgejo/infrastructure-documentation.
## LXC Hosts
## Table of content
All LXC hosts are setup with [lxc-helpers](https://code.forgejo.org/forgejo/lxc-helpers/).
- Setting up a new [K8S/DRBD/NFS k8s node](k8s.md)
- Setting up a new [LXC/DRBD Host](lxc.md)
- Managing services with a [LXC/DRBD/nginx stack](drbd-nginx-lxc.md)
- Installing a [Forgejo runner in an LXC container](runner-lxc.md)
- Managing the [Octopuce host](octopuce.md)
```sh
name=forgejo-host
lxc-helpers.sh lxc_container_run $name -- sudo --user debian bash
```
See https://github.com/mikesart/inotify-info. Running multiple LXC
containers will quickly use the default limit (128 on bookworm).
```sh
echo fs.inotify.max_user_instances=8192 | sudo tee -a /etc/sysctl.conf
sudo sysctl -p
```
### Unprivileged
```sh
name=forgejo-host
lxc-helpers.sh lxc_container_create --config "unprivileged" $name
echo "lxc.start.auto = 1" | sudo tee -a /var/lib/lxc/$name/config
lxc-helpers.sh lxc_container_start $name
lxc-helpers.sh lxc_container_user_install $name $(id -u) $USER
```
### Docker enabled
```sh
name=forgejo-host
lxc-helpers.sh lxc_container_create --config "docker" $name
echo "lxc.start.auto = 1" | sudo tee -a /var/lib/lxc/$name/config
lxc-helpers.sh lxc_container_start $name
lxc-helpers.sh lxc_install_docker $name
lxc-helpers.sh lxc_container_user_install $name $(id -u) $USER
```
### K8S enabled
```sh
name=forgejo-host
lxc-helpers.sh lxc_container_create --config "k8s" $name
echo "lxc.start.auto = 1" | sudo tee -a /var/lib/lxc/$name/config
lxc-helpers.sh lxc_container_start $name
lxc-helpers.sh lxc_container_user_install $name $(id -u) $USER
```
### Docker and LXC enabled
```sh
name=forgejo-host
ipv4=10.85.12
ipv6=fc33
lxc-helpers.sh lxc_container_create --config "docker lxc" $name
echo "lxc.start.auto = 1" | sudo tee -a /var/lib/lxc/$name/config
lxc-helpers.sh lxc_container_start $name
lxc-helpers.sh lxc_install_docker $name
lxc-helpers.sh lxc_install_lxc $name $ipv4 $ipv6
lxc-helpers.sh lxc_container_user_install $name $(id -u) $USER
```
## firewall
```sh
sudo apt-get install ufw
```
```sh
sudo ufw default allow incoming
sudo ufw default allow outgoing
sudo ufw default allow routed
interface=enp5s0
function internode() {
for from in $@ ; do
for to in $@ ; do
if test $from != $to ; then
sudo ufw allow in on $interface from $from to $to
fi
done
done
}
ipv4="65.108.204.171 88.198.58.177"
internode $ipv4
ipv6="2a01:4f9:1a:a082::2 2a01:4f8:222:507::2"
internode $ipv6
for host_ip in $ipv4 $ipv6 ; do
sudo ufw allow in on $interface to $host_ip port 22 proto tcp
sudo ufw deny in on $interface log-all to $host_ip
done
failover="188.40.16.47 2a01:4f8:fff2:48::2"
for public_ip in $failover ; do
sudo ufw allow in on $interface to $public_ip port 22,80,443,2000:3000 proto tcp
sudo ufw deny in on $interface log-all to $public_ip
done
```
```sh
sudo systemctl enable ufw
sudo ufw enable
```
```sh
sudo ufw status verbose
```
## nftables
```sh
sudo nft list ruleset
```
## Host reverse proxy
The reverse proxy on a host forwards to the designated LXC container with
something like the following examples in
`/etc/nginx/sites-available/example.com`, where A.B.C.D is the
IP allocated to the LXC container running the web service.
And symlink:
```sh
ln -s /etc/nginx/sites-available/example.com /etc/nginx/sites-enabled/example.com
```
The certificate is obtained once and automatically renewed with:
```
sudo apt-get install certbot python3-certbot-nginx
sudo certbot -n --agree-tos --email contact@forgejo.org -d example.com --nginx
```
When removing a configuration, the certificate can also be removed with:
```
sudo certbot delete --cert-name example.com
```
Forwarding TCP streams (useful for ssh) requires installing the module:
```sh
sudo apt-get install libnginx-mod-stream
```
Rate limiting crawlers is done by adding the following to `/etc/nginx/conf.d/limit.conf`:
```
# http://nginx.org/en/docs/http/ngx_http_limit_req_module.html
# https://blog.nginx.org/blog/rate-limiting-nginx
map $http_user_agent $isbot_ua {
default 0;
~*(GoogleBot|GoogleOther|bingbot|YandexBot) 1;
}
map $isbot_ua $limit_bot {
0 "";
1 $binary_remote_addr;
}
limit_req_zone $limit_bot zone=bots:10m rate=1r/m;
limit_req_status 429;
```
and the following in the location to be rate limited:
```
location / {
limit_req zone=bots burst=2 nodelay;
...
```
## Host wakeup-on-logs
https://code.forgejo.org/infrastructure/wakeup-on-logs
### K8S wakeup-on-logs script
```
$ cat /etc/wakeup-on-logs/forgejo-v8
#!/bin/bash
set -x
self="${BASH_SOURCE[0]}"
name=$(basename $self)
# keep it lower than https://code.forgejo.org/infrastructure/wakeup-on-logs
# otherwise it will get killed by it
timeout=4m
function lxc_run() {
lxc-attach $name -- sudo --user debian KUBECONFIG=/etc/rancher/k3s/k3s.yaml "$@" |& tee -a /var/log/$name.log
}
image=codeberg.org/forgejo-experimental/forgejo
major=${name##*v}
digest=$(skopeo inspect --format "{{.Digest}}" docker://$image:$major-rootless)
values=https://code.forgejo.org/infrastructure/k8s/raw/branch/main/forgejo-v$major/values.yml
lxc_run helm upgrade forgejo -f $values -f /home/debian/secrets.yml oci://code.forgejo.org/forgejo-helm/forgejo --atomic --wait --timeout $timeout --install --set image.digest=$digest
```
### Forgejo example
```
server {
listen 80;
listen [::]:80;
server_name example.com;
location / {
proxy_pass http://A.B.C.D:8080;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto https;
client_max_body_size 2G;
}
}
```
### GitLab example
```nginx
server {
listen 80;
listen [::]:80;
server_name example.com;
location / {
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection "upgrade";
proxy_set_header Host $http_host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
proxy_set_header X-Frame-Options SAMEORIGIN;
client_body_timeout 60;
client_max_body_size 200M;
send_timeout 1200;
lingering_timeout 5;
proxy_buffering off;
proxy_connect_timeout 90;
proxy_send_timeout 300;
proxy_read_timeout 600s;
proxy_pass http://example.com;
proxy_http_version 1.1;
}
}
```
### Vanila example
```nginx
server {
listen 80;
listen [::]:80;
server_name example.com;
location / {
proxy_pass http://A.B.C.D;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto https;
}
}
```
### 302 redirection
```nginx
server {
listen 80;
listen [::]:80;
server_name example.com;
return 302 https://other.example.com$request_uri;
}
```
## Forgejo runners
The LXC container in which the runner is installed must have capabilities that support the backend.
- docker:// needs a Docker enabled container
- lxc:// needs a Docker and LXC enabled container
The runners it contains are not started at boot, it must be done manually. The bash history has the command line to do so.
### Installation
```shell
version=3.5.0
sudo wget -O /usr/local/bin/forgejo-runner-$version https://code.forgejo.org/forgejo/runner/releases/download/v$version/forgejo-runner-$version-linux-amd64
sudo chmod +x /usr/local/bin/forgejo-runner-$version
echo 'export TERM=xterm-256color' >> .bashrc
```
### Creating a runner
Multiple runners can co-exist on the same machine. To keep things
organized they are located in a directory that is the same as the URL
from which the token is obtained. For instance
DIR=codeberg.org/forgejo-integration means that the token was obtained from the
https://codeberg.org/forgejo-integration organization.
If a runner only provides unprivileged docker containers, the labels
in `config.yml` should be:
`labels: ['docker:docker://node:20-bookworm']`.
If a runner provides LXC containers and unprivileged docker
containers, the labels in `config.yml` should be
`labels: ['self-hosted:lxc://debian:bookworm', 'docker:docker://node:20-bookworm']`.
```shell
name=myrunner
mkdir -p $DIR ; cd $DIR
forgejo-runner generate-config > config-$name.yml
## edit config-$name.yml and adjust the `labels:`
## Obtain a $TOKEN from https://$DIR
forgejo-runner-$version register --no-interactive --token $TOKEN --name runner --instance https://codeberg.org
forgejo-runner-$version --config config-$name.yml daemon |& cat -v > runner.log &
```
## Octopuce
[Octopuce provides hardware](https://codeberg.org/forgejo/sustainability) managed by [the devops team](https://codeberg.org/forgejo/governance/src/branch/main/TEAMS.md#devops). It can only be accessed via SSH.
To access the services hosted on the LXC containers, ssh port forwarding to the private IPs can be used. For instance:
```sh
echo 127.0.0.1 private.forgejo.org >> /etc/hosts
sudo ssh -i ~/.ssh/id_rsa -L 80:10.77.0.128:80 debian@forgejo01.octopuce.fr
firefox http://private.forgejo.org
```
### Containers
- `forgejo-host`
Dedicated to http://private.forgejo.org
- Docker enabled
- upgrades checklist:
```sh
emacs /home/debian/run-forgejo.sh # change the `image=`
docker stop forgejo
sudo rsync -av --numeric-ids --delete --progress /srv/forgejo/ /root/forgejo-backup/
docker rm forgejo
bash -x /home/debian/run-forgejo.sh
docker logs -n 200 -f forgejo
```
- `forgejo-runner-host`
Has runners installed to run against private.forgejo.org
- Docker and LXC enabled 10.85.12 fc33
## Hetzner
All hardware machines are running Debian GNU/linux bookworm. They are LXC hosts
setup with [lxc-helpers](https://code.forgejo.org/forgejo/lxc-helpers/).
> **NOTE:** only use [EX101 with a ASRockRack W680D4U-1L motherboard](https://forum.hetzner.com/index.php?thread/31135-all-ex101-with-asustek-w680-crash-on-sequential-read/).
### vSwitch
A vSwitch is assigned via the Robot console on all servers for backend communications
and [configured](https://docs.hetzner.com/robot/dedicated-server/network/vswitch#example-debian-configuration)
in /etc/network/interfaces for each of them with something like:
```
auto enp5s0.4000
iface enp5s0.4000 inet static
address 10.53.100.2
netmask 255.255.255.0
vlan-raw-device enp5s0
mtu 1400
```
The IP address ends with the same number as the hardware (hetzner02 => .2).
#### vSwitch DRBD
The vSwitch on VLAN 4000 is for DRBD exclusively
#### vSwitch NFS
The vSwitch on VLAN 4001 is for NFS
#### vSwitch k8s
The vSwitch on VLAN 4002 is for the k8s control plane
### DRBD
DRBD is [configured](https://linbit.com/drbd-user-guide/drbd-guide-9_0-en/#p-work) like in the following example with hetzner02 as the primary and hetzner03 as the secondary:
```sh
$ apt-get install drbd-utils
$ cat /etc/drbd.d/r0.res
resource r0 {
net {
# A : write completion is determined when data is written to the local disk and the local TCP transmission buffer
# B : write completion is determined when data is written to the local disk and remote buffer cache
# C : write completion is determined when data is written to both the local disk and the remote disk
protocol C;
cram-hmac-alg sha1;
# any secret key for authentication among nodes
shared-secret "***";
}
disk {
resync-rate 1000M;
}
on hetzner02 {
address 10.53.100.2:7788;
volume 0 {
# device name
device /dev/drbd0;
# specify disk to be used for device above
disk /dev/nvme0n1p5;
# where to create metadata
# specify the block device name when using a different disk
meta-disk internal;
}
}
on hetzner03 {
address 10.53.100.3:7788;
volume 0 {
device /dev/drbd0;
disk /dev/nvme1n1p5;
meta-disk internal;
}
}
}
$ sudo drbdadm create-md r0
$ sudo systemctl enable drbd
$ sudo systemctl start drbd
```
On hetzner02 (the primary), [pretend all is in sync](https://linbit.com/drbd-user-guide/drbd-guide-9_0-en/#s-skip-initial-resync) to save the initial bitmap sync since
there is actually no data at all.
```sh
sudo drbdadm new-current-uuid --clear-bitmap r0/0
```
The DRBD device is mounted on `/var/lib/lxc` in `/etc/fstab` there is a noauto line:
```
/dev/drbd0 /var/lib/lxc ext4 noauto,defaults 0 0
```
To prevent split brain situations a manual step is required at boot
time, on the machine that is going to be the primary.
```sh
sudo drbdadm primary r0
sudo drbdsetup status
sudo mount /var/lib/lxc
sudo lxc-autostart start
sudo lxc-ls -f
sudo drbdsetup status
```
### hetzner{01,04}
## hetzner{01,04}
https://hetzner{01,04}.forgejo.org run on [EX101](https://www.hetzner.com/dedicated-rootserver/ex101) Hetzner hardware.
#### LXC
### LXC
```sh
lxc-helpers.sh lxc_install_lxc_inside 10.41.13 fc29
```
#### Disk partitioning
### Disk partitioning
- First disk
- OS
@ -495,14 +27,14 @@ lxc-helpers.sh lxc_install_lxc_inside 10.41.13 fc29
- Second disk
- configured with DRBD for precious data.
#### Root filesystem backups
### Root filesystem backups
- `hetzner01:/etc/cron.daily/backup-hetzner04`
`rsync -aHS --delete-excluded --delete --numeric-ids --exclude /proc --exclude /dev --exclude /sys --exclude /precious --exclude /srv --exclude /var/lib/lxc 10.53.100.4:/ /srv/backups/hetzner04/ >& /var/log/$(basename $0).log`
- `hetzner04:/etc/cron.daily/backup-hetzner01`
`rsync -aHS --delete-excluded --delete --numeric-ids --exclude /proc --exclude /dev --exclude /sys --exclude /precious --exclude /srv --exclude /var/lib/lxc 10.53.100.1:/ /srv/backups/hetzner01/ >& /var/log/$(basename $0).log`
#### LXC containers
### LXC containers
- `runner-lxc-helpers` (hetzner01)
@ -588,17 +120,17 @@ lxc-helpers.sh lxc_install_lxc_inside 10.41.13 fc29
- Docker enabled
### hetzner{02,03}
## hetzner{02,03}
https://hetzner02.forgejo.org & https://hetzner03.forgejo.org run on [EX44](https://www.hetzner.com/dedicated-rootserver/ex44) Hetzner hardware.
#### LXC
### LXC
```sh
lxc-helpers.sh lxc_install_lxc_inside 10.6.83 fc16
```
#### Disk partitioning
### Disk partitioning
- First disk
- OS
@ -606,14 +138,14 @@ lxc-helpers.sh lxc_install_lxc_inside 10.6.83 fc16
- Second disk
- non precious data such as the LXC containers with runners.
#### Root filesystem backups
### Root filesystem backups
- `hetzner03:/etc/cron.daily/backup-hetzner02`
`rsync -aHS --delete-excluded --delete --numeric-ids --exclude /proc --exclude /dev --exclude /sys --exclude /srv --exclude /var/lib/lxc 10.53.100.2:/ /srv/backups/hetzner02/`
- `hetzner02:/etc/cron.daily/backup-hetzner03`
`rsync -aHS --delete-excluded --delete --numeric-ids --exclude /proc --exclude /dev --exclude /sys --exclude /srv --exclude /var/lib/lxc 10.53.100.3:/ /srv/backups/hetzner03/`
#### Public IP addresses
### Public IP addresses
The public IP addresses attached to the hosts are not failover IPs that can be moved from one host to the next.
The DNS entry needs to be updated if the primary hosts changes.
@ -653,7 +185,7 @@ ListenAddress 65.21.67.73
#ListenAddress ::
```
#### Port forwarding
### Port forwarding
Forwarding a port to an LXC container can be done with [nginx streeam](https://nginx.org/en/docs/stream/ngx_stream_core_module.html) for the public IP of code.forgejo.org (65.21.67.71 & 2a01:4f9:3081:51ec::102) to the private IP (10.6.83.195) of the `code` LXC container in `/etc/nginx/modules-enabled/ssh.conf`:
@ -677,13 +209,13 @@ stream {
}
```
#### 302 redirects
### 302 redirects
- On hetzner02
- try.next.forgejo.org redirects to v(latest stable).next.forgejo.org
- dev.next.forgejo.org redirects to v(latest dev).next.forgejo.org
#### Containers
### Containers
- `forgejo-code` on hetzner02
@ -785,186 +317,11 @@ stream {
Dedicated to https://codeberg.org/forgejo-contrib/forgejo-helm and running from an ephemeral disk
### hetzner{05,06}
## hetzner{05,06}
https://hetzner05.forgejo.org & https://hetzner06.forgejo.org run on [EX44](https://www.hetzner.com/dedicated-rootserver/ex44) Hetzner hardware.
#### Imaging
Using installimage from the rescue instance.
- `wipefs -fa /dev/nvme*n1`
- `installimage -r no -n hetzner0?`
- Debian bookworm
- `PART / ext4 100G`
- `PART /srv ext4 all`
- ESC 0 + yes
- reboot
Partitioning.
- First disk
- OS
- non precious data such as the LXC containers with runners.
- Second disk
- a partition configured with DRBD
Debian user.
- `ssh root@hetzner0?.forgejo.org`
- `useradd --shell /bin/bash --create-home --groups sudo debian`
- `mkdir -p /home/debian/.ssh ; cp -a .ssh/authorized_keys /home/debian/.ssh ; chown -R debian /home/debian/.ssh`
- in `/etc/sudoers` edit `%sudo ALL=(ALL:ALL) NOPASSWD:ALL`
#### Install helpers
Each node is identifed by the last digit of the hostname.
```sh
sudo apt-get install git etckeeper
git clone https://code.forgejo.org/infrastructure/documentation
cd documentation/k3s-host
cp variables.sh.example variables.sh
cp secrets.sh.example secrets.sh
```
Variables that must be set depending on the role of the node.
- first server node
- secrets.sh: node_drbd_shared_secret
- other server node
- secrets.sh: node_drbd_shared_secret
- secrets.sh: node_k8s_token: content of /var/lib/rancher/k3s/server/token on the first node
- variables.sh: node_k8s_existing: identifier of the first node (e.g. 5)
- etcd node
- secrets.sh: node_k8s_token: content of /var/lib/rancher/k3s/server/token on the first node
- variables.sh: node_k8s_existing: identifier of the first node (e.g. 5)
- variables.sh: node_k8s_etcd: identifier of the node whose role is just etcd (e.g. 3)
The other variables depend on the setup.
#### Firewall
`./setup.sh setup_ufw`
#### DRBD
DRBD is [configured](https://linbit.com/drbd-user-guide/drbd-guide-9_0-en/#p-work) with:
`./setup.sh setup_drbd`
Once two nodes have DRBD setup for the first time, it can be initialized by [pretending all is in sync](https://linbit.com/drbd-user-guide/drbd-guide-9_0-en/#s-skip-initial-resync) to save the initial bitmap sync since there is actually no data at all.
```sh
sudo drbdadm primary r1
sudo drbdadm new-current-uuid --clear-bitmap r1/0
sudo mount /precious
```
#### NFS
`./setup.sh setup_nfs`
On the node that has the DRBD volume `/precious` mounted, set the IP of the NFS server to be used by k8s:
```sh
sudo ip addr add 10.53.101.100/24 dev enp5s0.4001
```
#### K8S
For the first node `./setup.sh setup_k8s`. For nodes joining the cluster `./setup.sh setup_k8s 6` where `hetzner06` is an existing node.
- [metallb](https://metallb.universe.tf) instead of the default load balancer because it does not allow for a public IP different from the `k8s` node IP.
`./setup.sh setup_k8s_metallb`
- [traefik](https://traefik.io/) requests with [annotations](https://github.com/traefik/traefik-helm-chart/blob/7a13fc8a61a6ad30fcec32eec497dab9d8aea686/traefik/values.yaml#L736) specific IPs from `metalldb`.
`./setup.sh setup_k8s_traefik`
- [cert-manager](https://cert-manager.io/).
`./setup.sh setup_k8s_certmanager`
- NFS storage class
`./setup.sh setup_k8s_nfs`
#### Forgejo
[forgejo](https://code.forgejo.org/forgejo-helm/forgejo-helm) configuration in [ingress](https://code.forgejo.org/forgejo-helm/forgejo-helm#ingress) for the reverse proxy (`traefik`) to route the domain and for the ACME issuer (`cert-manager`) to obtain a certificate. And in [service](https://code.forgejo.org/forgejo-helm/forgejo-helm#service) for the `ssh` port to be bound to the desired IPs of the load balancer (`metallb`).
```
ingress:
enabled: true
annotations:
# https://cert-manager.io/docs/usage/ingress/#supported-annotations
# https://github.com/cert-manager/cert-manager/issues/2239
cert-manager.io/cluster-issuer: letsencrypt-http
cert-manager.io/private-key-algorithm: ECDSA
cert-manager.io/private-key-size: 384
kubernetes.io/ingress.class: traefik
traefik.ingress.kubernetes.io/router.entrypoints: websecure
tls:
- hosts:
- t1.forgejo.org
secretName: tls-forgejo-t1-ingress-http
hosts:
- host: t1.forgejo.org
paths:
- path: /
pathType: Prefix
service:
http:
type: ClusterIP
ipFamilyPolicy: PreferDualStack
port: 3000
ssh:
type: LoadBalancer
annotations:
metallb.universe.tf/loadBalancerIPs: 188.40.16.47,2a01:4f8:fff2:48::2
metallb.universe.tf/allow-shared-ip: "key-to-share-failover"
ipFamilyPolicy: PreferDualStack
port: 2222
```
### K8S NFS storage creation
Define the 20GB `forgejo-data` pvc owned by user id 1000.
```sh
./setup.sh setup_k8s_pvc forgejo-data 20Gi 1000
```
[Instruct the forgejo pod](https://code.forgejo.org/forgejo-helm/forgejo-helm#persistence) to use the `forgejo-data` pvc.
```yaml
persistence:
enabled: true
create: false
claimName: forgejo-data
```
## Disaster recovery and maintenance
### When a machine or disk is scheduled for replacement.
* `kubectl drain hetzner05` # evacuate all the pods out of the node to be shutdown
* `kubectl taint nodes hetzner05 key1=value1:NoSchedule` # prevent any pod from being created there (metallb speaker won't be drained, for instance)
* `kubectl delete node hetzner05` # let the cluster know it no longer exists so a new one by the same name can replace it
### Routing the failover IP
When the machine to which the failover IP (failover.forgejo.org) is routed is unavailable or to be shutdown, to the [Hetzner server panel](https://robot.hetzner.com/server), to the IPs tab and change the route of the failover IP to another node. All nodes are configured with the failover IP, there is nothing else to do.
### Manual boot operations
#### On the machine that runs the NFS server
* `sudo drbdadm primary r1` # Switch the DRBD to primary
* `sudo mount /precious` # DRBD volume shared via NFS
* `sudo ip addr add 10.53.101.100/24 dev enp5s0.4001` # add NFS server IP
#### On the other machines
* `sudo ip addr del 10.53.101.100/24 dev enp5s0.4001` # remove NFS server IP
Nodes of [a k8s cluster](k8s.md).
## Uberspace

177
drbd-nginx-lxc.md Normal file
View file

@ -0,0 +1,177 @@
## nftables
```sh
sudo nft list ruleset
```
## Host reverse proxy
The reverse proxy on a host forwards to the designated LXC container with
something like the following examples in
`/etc/nginx/sites-available/example.com`, where A.B.C.D is the
IP allocated to the LXC container running the web service.
And symlink:
```sh
ln -s /etc/nginx/sites-available/example.com /etc/nginx/sites-enabled/example.com
```
The certificate is obtained once and automatically renewed with:
```
sudo apt-get install certbot python3-certbot-nginx
sudo certbot -n --agree-tos --email contact@forgejo.org -d example.com --nginx
```
When removing a configuration, the certificate can also be removed with:
```
sudo certbot delete --cert-name example.com
```
Forwarding TCP streams (useful for ssh) requires installing the module:
```sh
sudo apt-get install libnginx-mod-stream
```
Rate limiting crawlers is done by adding the following to `/etc/nginx/conf.d/limit.conf`:
```
# http://nginx.org/en/docs/http/ngx_http_limit_req_module.html
# https://blog.nginx.org/blog/rate-limiting-nginx
map $http_user_agent $isbot_ua {
default 0;
~*(GoogleBot|GoogleOther|bingbot|YandexBot) 1;
}
map $isbot_ua $limit_bot {
0 "";
1 $binary_remote_addr;
}
limit_req_zone $limit_bot zone=bots:10m rate=1r/m;
limit_req_status 429;
```
and the following in the location to be rate limited:
```
location / {
limit_req zone=bots burst=2 nodelay;
...
```
## Host wakeup-on-logs
https://code.forgejo.org/infrastructure/wakeup-on-logs
### K8S wakeup-on-logs script
```
$ cat /etc/wakeup-on-logs/forgejo-v8
#!/bin/bash
set -x
self="${BASH_SOURCE[0]}"
name=$(basename $self)
# keep it lower than https://code.forgejo.org/infrastructure/wakeup-on-logs
# otherwise it will get killed by it
timeout=4m
function lxc_run() {
lxc-attach $name -- sudo --user debian KUBECONFIG=/etc/rancher/k3s/k3s.yaml "$@" |& tee -a /var/log/$name.log
}
image=codeberg.org/forgejo-experimental/forgejo
major=${name##*v}
digest=$(skopeo inspect --format "{{.Digest}}" docker://$image:$major-rootless)
values=https://code.forgejo.org/infrastructure/k8s/raw/branch/main/forgejo-v$major/values.yml
lxc_run helm upgrade forgejo -f $values -f /home/debian/secrets.yml oci://code.forgejo.org/forgejo-helm/forgejo --atomic --wait --timeout $timeout --install --set image.digest=$digest
```
### Forgejo example
```
server {
listen 80;
listen [::]:80;
server_name example.com;
location / {
proxy_pass http://A.B.C.D:8080;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto https;
client_max_body_size 2G;
}
}
```
### GitLab example
```nginx
server {
listen 80;
listen [::]:80;
server_name example.com;
location / {
proxy_set_header Upgrade $http_upgrade;
proxy_set_header Connection "upgrade";
proxy_set_header Host $http_host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto $scheme;
proxy_set_header X-Frame-Options SAMEORIGIN;
client_body_timeout 60;
client_max_body_size 200M;
send_timeout 1200;
lingering_timeout 5;
proxy_buffering off;
proxy_connect_timeout 90;
proxy_send_timeout 300;
proxy_read_timeout 600s;
proxy_pass http://example.com;
proxy_http_version 1.1;
}
}
```
### Vanila example
```nginx
server {
listen 80;
listen [::]:80;
server_name example.com;
location / {
proxy_pass http://A.B.C.D;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
proxy_set_header X-Forwarded-Proto https;
}
}
```
### 302 redirection
```nginx
server {
listen 80;
listen [::]:80;
server_name example.com;
return 302 https://other.example.com$request_uri;
}
```

176
k8s.md Normal file
View file

@ -0,0 +1,176 @@
#### Imaging
Using installimage from the rescue instance.
- `wipefs -fa /dev/nvme*n1`
- `installimage -r no -n hetzner0?`
- Debian bookworm
- `PART / ext4 100G`
- `PART /srv ext4 all`
- ESC 0 + yes
- reboot
Partitioning.
- First disk
- OS
- non precious data such as the LXC containers with runners.
- Second disk
- a partition configured with DRBD
Debian user.
- `ssh root@hetzner0?.forgejo.org`
- `useradd --shell /bin/bash --create-home --groups sudo debian`
- `mkdir -p /home/debian/.ssh ; cp -a .ssh/authorized_keys /home/debian/.ssh ; chown -R debian /home/debian/.ssh`
- in `/etc/sudoers` edit `%sudo ALL=(ALL:ALL) NOPASSWD:ALL`
#### Install helpers
Each node is identifed by the last digit of the hostname.
```sh
sudo apt-get install git etckeeper
git clone https://code.forgejo.org/infrastructure/documentation
cd documentation/k3s-host
cp variables.sh.example variables.sh
cp secrets.sh.example secrets.sh
```
Variables that must be set depending on the role of the node.
- first server node
- secrets.sh: node_drbd_shared_secret
- other server node
- secrets.sh: node_drbd_shared_secret
- secrets.sh: node_k8s_token: content of /var/lib/rancher/k3s/server/token on the first node
- variables.sh: node_k8s_existing: identifier of the first node (e.g. 5)
- etcd node
- secrets.sh: node_k8s_token: content of /var/lib/rancher/k3s/server/token on the first node
- variables.sh: node_k8s_existing: identifier of the first node (e.g. 5)
- variables.sh: node_k8s_etcd: identifier of the node whose role is just etcd (e.g. 3)
The other variables depend on the setup.
#### Firewall
`./setup.sh setup_ufw`
#### DRBD
DRBD is [configured](https://linbit.com/drbd-user-guide/drbd-guide-9_0-en/#p-work) with:
`./setup.sh setup_drbd`
Once two nodes have DRBD setup for the first time, it can be initialized by [pretending all is in sync](https://linbit.com/drbd-user-guide/drbd-guide-9_0-en/#s-skip-initial-resync) to save the initial bitmap sync since there is actually no data at all.
```sh
sudo drbdadm primary r1
sudo drbdadm new-current-uuid --clear-bitmap r1/0
sudo mount /precious
```
#### NFS
`./setup.sh setup_nfs`
On the node that has the DRBD volume `/precious` mounted, set the IP of the NFS server to be used by k8s:
```sh
sudo ip addr add 10.53.101.100/24 dev enp5s0.4001
```
#### K8S
For the first node `./setup.sh setup_k8s`. For nodes joining the cluster `./setup.sh setup_k8s 6` where `hetzner06` is an existing node.
- [metallb](https://metallb.universe.tf) instead of the default load balancer because it does not allow for a public IP different from the `k8s` node IP.
`./setup.sh setup_k8s_metallb`
- [traefik](https://traefik.io/) requests with [annotations](https://github.com/traefik/traefik-helm-chart/blob/7a13fc8a61a6ad30fcec32eec497dab9d8aea686/traefik/values.yaml#L736) specific IPs from `metalldb`.
`./setup.sh setup_k8s_traefik`
- [cert-manager](https://cert-manager.io/).
`./setup.sh setup_k8s_certmanager`
- NFS storage class
`./setup.sh setup_k8s_nfs`
#### Forgejo
[forgejo](https://code.forgejo.org/forgejo-helm/forgejo-helm) configuration in [ingress](https://code.forgejo.org/forgejo-helm/forgejo-helm#ingress) for the reverse proxy (`traefik`) to route the domain and for the ACME issuer (`cert-manager`) to obtain a certificate. And in [service](https://code.forgejo.org/forgejo-helm/forgejo-helm#service) for the `ssh` port to be bound to the desired IPs of the load balancer (`metallb`).
```
ingress:
enabled: true
annotations:
# https://cert-manager.io/docs/usage/ingress/#supported-annotations
# https://github.com/cert-manager/cert-manager/issues/2239
cert-manager.io/cluster-issuer: letsencrypt-http
cert-manager.io/private-key-algorithm: ECDSA
cert-manager.io/private-key-size: 384
kubernetes.io/ingress.class: traefik
traefik.ingress.kubernetes.io/router.entrypoints: websecure
tls:
- hosts:
- t1.forgejo.org
secretName: tls-forgejo-t1-ingress-http
hosts:
- host: t1.forgejo.org
paths:
- path: /
pathType: Prefix
service:
http:
type: ClusterIP
ipFamilyPolicy: PreferDualStack
port: 3000
ssh:
type: LoadBalancer
annotations:
metallb.universe.tf/loadBalancerIPs: 188.40.16.47,2a01:4f8:fff2:48::2
metallb.universe.tf/allow-shared-ip: "key-to-share-failover"
ipFamilyPolicy: PreferDualStack
port: 2222
```
### K8S NFS storage creation
Define the 20GB `forgejo-data` pvc owned by user id 1000.
```sh
./setup.sh setup_k8s_pvc forgejo-data 20Gi 1000
```
[Instruct the forgejo pod](https://code.forgejo.org/forgejo-helm/forgejo-helm#persistence) to use the `forgejo-data` pvc.
```yaml
persistence:
enabled: true
create: false
claimName: forgejo-data
```
## Disaster recovery and maintenance
### When a machine or disk is scheduled for replacement.
* `kubectl drain hetzner05` # evacuate all the pods out of the node to be shutdown
* `kubectl taint nodes hetzner05 key1=value1:NoSchedule` # prevent any pod from being created there (metallb speaker won't be drained, for instance)
* `kubectl delete node hetzner05` # let the cluster know it no longer exists so a new one by the same name can replace it
### Routing the failover IP
When the machine to which the failover IP (failover.forgejo.org) is routed is unavailable or to be shutdown, to the [Hetzner server panel](https://robot.hetzner.com/server), to the IPs tab and change the route of the failover IP to another node. All nodes are configured with the failover IP, there is nothing else to do.
### Manual boot operations
#### On the machine that runs the NFS server
* `sudo drbdadm primary r1` # Switch the DRBD to primary
* `sudo mount /precious` # DRBD volume shared via NFS
* `sudo ip addr add 10.53.101.100/24 dev enp5s0.4001` # add NFS server IP
#### On the other machines
* `sudo ip addr del 10.53.101.100/24 dev enp5s0.4001` # remove NFS server IP

160
lxc.md Normal file
View file

@ -0,0 +1,160 @@
## LXC Hosts
All LXC hosts are setup with [lxc-helpers](https://code.forgejo.org/forgejo/lxc-helpers/).
```sh
name=forgejo-host
lxc-helpers.sh lxc_container_run $name -- sudo --user debian bash
```
See https://github.com/mikesart/inotify-info. Running multiple LXC
containers will quickly use the default limit (128 on bookworm).
```sh
echo fs.inotify.max_user_instances=8192 | sudo tee -a /etc/sysctl.conf
sudo sysctl -p
```
### Unprivileged
```sh
name=forgejo-host
lxc-helpers.sh lxc_container_create --config "unprivileged" $name
echo "lxc.start.auto = 1" | sudo tee -a /var/lib/lxc/$name/config
lxc-helpers.sh lxc_container_start $name
lxc-helpers.sh lxc_container_user_install $name $(id -u) $USER
```
### Docker enabled
```sh
name=forgejo-host
lxc-helpers.sh lxc_container_create --config "docker" $name
echo "lxc.start.auto = 1" | sudo tee -a /var/lib/lxc/$name/config
lxc-helpers.sh lxc_container_start $name
lxc-helpers.sh lxc_install_docker $name
lxc-helpers.sh lxc_container_user_install $name $(id -u) $USER
```
### K8S enabled
```sh
name=forgejo-host
lxc-helpers.sh lxc_container_create --config "k8s" $name
echo "lxc.start.auto = 1" | sudo tee -a /var/lib/lxc/$name/config
lxc-helpers.sh lxc_container_start $name
lxc-helpers.sh lxc_container_user_install $name $(id -u) $USER
```
### Docker and LXC enabled
```sh
name=forgejo-host
ipv4=10.85.12
ipv6=fc33
lxc-helpers.sh lxc_container_create --config "docker lxc" $name
echo "lxc.start.auto = 1" | sudo tee -a /var/lib/lxc/$name/config
lxc-helpers.sh lxc_container_start $name
lxc-helpers.sh lxc_install_docker $name
lxc-helpers.sh lxc_install_lxc $name $ipv4 $ipv6
lxc-helpers.sh lxc_container_user_install $name $(id -u) $USER
```
## Hetzner
All hardware machines are running Debian GNU/linux bookworm. They are LXC hosts
setup with [lxc-helpers](https://code.forgejo.org/forgejo/lxc-helpers/).
> **NOTE:** only use [EX101 with a ASRockRack W680D4U-1L motherboard](https://forum.hetzner.com/index.php?thread/31135-all-ex101-with-asustek-w680-crash-on-sequential-read/).
### vSwitch
A vSwitch is assigned via the Robot console on all servers for backend communications
and [configured](https://docs.hetzner.com/robot/dedicated-server/network/vswitch#example-debian-configuration)
in /etc/network/interfaces for each of them with something like:
```
auto enp5s0.4000
iface enp5s0.4000 inet static
address 10.53.100.2
netmask 255.255.255.0
vlan-raw-device enp5s0
mtu 1400
```
The IP address ends with the same number as the hardware (hetzner02 => .2).
#### vSwitch DRBD
The vSwitch on VLAN 4000 is for DRBD exclusively
### DRBD
DRBD is [configured](https://linbit.com/drbd-user-guide/drbd-guide-9_0-en/#p-work) like in the following example with hetzner02 as the primary and hetzner03 as the secondary:
```sh
$ apt-get install drbd-utils
$ cat /etc/drbd.d/r0.res
resource r0 {
net {
# A : write completion is determined when data is written to the local disk and the local TCP transmission buffer
# B : write completion is determined when data is written to the local disk and remote buffer cache
# C : write completion is determined when data is written to both the local disk and the remote disk
protocol C;
cram-hmac-alg sha1;
# any secret key for authentication among nodes
shared-secret "***";
}
disk {
resync-rate 1000M;
}
on hetzner02 {
address 10.53.100.2:7788;
volume 0 {
# device name
device /dev/drbd0;
# specify disk to be used for device above
disk /dev/nvme0n1p5;
# where to create metadata
# specify the block device name when using a different disk
meta-disk internal;
}
}
on hetzner03 {
address 10.53.100.3:7788;
volume 0 {
device /dev/drbd0;
disk /dev/nvme1n1p5;
meta-disk internal;
}
}
}
$ sudo drbdadm create-md r0
$ sudo systemctl enable drbd
$ sudo systemctl start drbd
```
On hetzner02 (the primary), [pretend all is in sync](https://linbit.com/drbd-user-guide/drbd-guide-9_0-en/#s-skip-initial-resync) to save the initial bitmap sync since
there is actually no data at all.
```sh
sudo drbdadm new-current-uuid --clear-bitmap r0/0
```
The DRBD device is mounted on `/var/lib/lxc` in `/etc/fstab` there is a noauto line:
```
/dev/drbd0 /var/lib/lxc ext4 noauto,defaults 0 0
```
To prevent split brain situations a manual step is required at boot
time, on the machine that is going to be the primary.
```sh
sudo drbdadm primary r0
sudo drbdsetup status
sudo mount /var/lib/lxc
sudo lxc-autostart start
sudo lxc-ls -f
sudo drbdsetup status
```

35
octopuce.md Normal file
View file

@ -0,0 +1,35 @@
## Octopuce
[Octopuce provides hardware](https://codeberg.org/forgejo/sustainability) managed by [the devops team](https://codeberg.org/forgejo/governance/src/branch/main/TEAMS.md#devops). It can only be accessed via SSH.
To access the services hosted on the LXC containers, ssh port forwarding to the private IPs can be used. For instance:
```sh
echo 127.0.0.1 private.forgejo.org >> /etc/hosts
sudo ssh -i ~/.ssh/id_rsa -L 80:10.77.0.128:80 debian@forgejo01.octopuce.fr
firefox http://private.forgejo.org
```
### Containers
- `forgejo-host`
Dedicated to http://private.forgejo.org
- Docker enabled
- upgrades checklist:
```sh
emacs /home/debian/run-forgejo.sh # change the `image=`
docker stop forgejo
sudo rsync -av --numeric-ids --delete --progress /srv/forgejo/ /root/forgejo-backup/
docker rm forgejo
bash -x /home/debian/run-forgejo.sh
docker logs -n 200 -f forgejo
```
- `forgejo-runner-host`
Has runners installed to run against private.forgejo.org
- Docker and LXC enabled 10.85.12 fc33

44
runner-lxc.md Normal file
View file

@ -0,0 +1,44 @@
## Forgejo runners
The LXC container in which the runner is installed must have capabilities that support the backend.
- docker:// needs a Docker enabled container
- lxc:// needs a Docker and LXC enabled container
The runners it contains are not started at boot, it must be done manually. The bash history has the command line to do so.
### Installation
```shell
version=3.5.0
sudo wget -O /usr/local/bin/forgejo-runner-$version https://code.forgejo.org/forgejo/runner/releases/download/v$version/forgejo-runner-$version-linux-amd64
sudo chmod +x /usr/local/bin/forgejo-runner-$version
echo 'export TERM=xterm-256color' >> .bashrc
```
### Creating a runner
Multiple runners can co-exist on the same machine. To keep things
organized they are located in a directory that is the same as the URL
from which the token is obtained. For instance
DIR=codeberg.org/forgejo-integration means that the token was obtained from the
https://codeberg.org/forgejo-integration organization.
If a runner only provides unprivileged docker containers, the labels
in `config.yml` should be:
`labels: ['docker:docker://node:20-bookworm']`.
If a runner provides LXC containers and unprivileged docker
containers, the labels in `config.yml` should be
`labels: ['self-hosted:lxc://debian:bookworm', 'docker:docker://node:20-bookworm']`.
```shell
name=myrunner
mkdir -p $DIR ; cd $DIR
forgejo-runner generate-config > config-$name.yml
## edit config-$name.yml and adjust the `labels:`
## Obtain a $TOKEN from https://$DIR
forgejo-runner-$version register --no-interactive --token $TOKEN --name runner --instance https://codeberg.org
forgejo-runner-$version --config config-$name.yml daemon |& cat -v > runner.log &
```