Merge pull request #55 from bbkz/dev

Release 0.2.1 * fixes #54 Database migration fails * fix celery redis password * update development setup
wger-project · Sep 17, 2024 · e06c70a · e06c70a
2 parents 1cb7ac0 + 796ba85
commit e06c70a
Show file tree

Hide file tree

Showing 5 changed files with 99 additions and 143 deletions.
diff --git a/DEVEL.md b/DEVEL.md
@@ -4,7 +4,9 @@ The following is done on a Fedora Desktop to run a minikube rootless setup. For
 
 ## Prepare the system
 
-First install the required network component `slirp4netns`
+The network namespace of the Node components has to have a non-loopback interface, which can be for example configured with slirp4netns, VPNKit, or lxc-user-nic(1).
+
+Let's install the network component `slirp4netns`
 
 ```bash
 sudo dnf install slirp4netns
@@ -27,6 +29,8 @@ So this was the only parts where root priviledges are needed.
 
 Now install and setup minikube with the calico network driver. Assuming you have `~/bin` in your `$PATH` environment variable.
 
+Check your `~/.kube` folder if you have a old minikube config and (re)move it.
+
 ```bash
 wget https://storage.googleapis.com/minikube/releases/latest/minikube-linux-amd64 -O ~/bin/minikube
 chmod 755 ~/bin/minikube
@@ -38,132 +42,62 @@ minikube config set container-runtime containerd
 minikube start --cni calico
 ```
 
-Now you have a running cluster on your machine.
-
-Minikube comes with a integrated `kubectl` command. So you can run `kubectl` commands, without downloaded `kubectl` binary:
-
-```bash
-minikube kubectl -- get pods -A
-```
-
-But for using `helm` and our convenience, we install `kubectl` alongside `minikube`:
+Download `kubectl`, into `~/bin`:
 
 ```bash
 wget "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/linux/amd64/kubectl" -O ~/bin/kubectl
 ```
 
-Finally we install `helm`, into `~/bin`:
+Download `helm`, into `~/bin`:
 
 ```bash
 export HELM_INSTALL_DIR=~/bin; export USE_SUDO=false; curl https://raw.githubusercontent.com/helm/helm/main/scripts/get-helm-3 | bash
 ```
 
-## Setup wger
+In your `~/.kube/config` set the IP of your host instead of `127.0.0.1`, to make helm work (may not be neccessary).
 
-You can install wger without any changes to the `values.yaml`, this will run wger in development mode.
+## Setup wger
 
-First clone the `wger-helm-charts` repository and optionally create `your_values.yaml` file:
+First clone the `wger-helm-charts` repository and optionally create your own values file based on `example/devel.yaml`:
 
 ```bash
 git clone https://github.com/wger-project/helm-charts.git
 cd helm-charts
-vi your_values.yaml
-```
-
-The following is a example of `your_values.yaml`:
-
-```yaml
-app:
-  environment:
-      # x-real-ip - remote ip - x-forward-for  -
-    - name: GUNICORN_CMD_ARGS
-      value: "--timeout 240 --workers=2 --access-logformat '%({x-real-ip}i)s %(l)s %(h)s %(l)s %({x-forwarded-for}i)s %(l)s %(t)s \"%(r)s\" %(s)s %(b)s \"%(f)s\" \"%(a)s\"' --access-logfile - --error-logfile -"
-  nginx:
-    enabled: true
-  axes:
-    enabled: true
-celery:
-  enabled: true
-  flower:
-    enabled: true
 ```
 
-Deploy the helm chart from the cloned git repo. Omit `-f ../../your_values.yaml` when you don't have the file:
+Deploy the helm chart from your local files:
 
 ```bash
-cd helm-charts/charts/wger
+cd charts/wger
 helm dependency update
-helm upgrade --install wger . -n wger --create-namespace -f ../../your_values.yaml
+helm upgrade --install wger . -n wger --create-namespace -f ../../example/devel.yaml
 ```
 
-To access the webinterface, you can port forward `8000` from the wger app to a port on your machine, be aware you need a high port number, which doesn't require root priviledges.
+To access the webinterface, you can port forward `8080` from the wger app to a port on your machine, be aware you need a high port number, which doesn't require root priviledges.
+
+Also note you need to connect to the container port directly not the service port.
 
 ```bash
 export POD=$(kubectl get pods -n wger -l "app.kubernetes.io/name=wger-app" -o jsonpath="{.items[0].metadata.name}")
-echo "wger runs on: http://localhost:10001"; kubectl -n wger port-forward ${POD} 10001:8000
+echo "wger runs on: http://localhost:10001"; kubectl -n wger port-forward ${POD} 10001:8080
 ```
 
 Go to http://localhost:10001 and login as `admin` `adminadmin` ;-)
 
-## Advanced Setup
-
-When you activated `nginx` persistent storage will be automatically activated as a requirement. You can see the volumes (pv) and it's claims (pvc):
-
-```bash
-kubectl get pv
-kubectl get pvc -n wger
-```
-
-**@todo sorry but, mounting with rootless podman and minikube doesn't work yet**
-
-There is a special claim `code` which will not be created but will overload the wger django code, this can be used to mount your local development code into the setup.
-
-First checkout the code to in the example i use `$HOME/test/wger`.
-
-As minikube is running in a VM we first need to mount the local files into the minikube VM to make it available for the kubernetes cluster. You can login to the minikube VM with `minikube ssh`.
-
-Now mount the folder into the minikube system, i use `/wger-code` here.
-
-```bash
-minikube stop
-minikube start --cni calico --mount-string="$HOME/test/wger:/wger-code"
-# or
-minikube mount $HOME/test/wger:/wger-code
-```
-
-Add the following to `your_values.yaml`.
-
-```yaml
-app:
-  persistence:
-    existingClaim:
-      code: wger-code
-```
-
-Manually create a volume and claim for your local wger code. For this add a new file `wger-code.yaml` and apply it to the cluster:
-
-```yaml
-TBD
-```
-
-```bash
-kubectl create ns wger
-kubectl apply -n wger -f ../../wger-code-volume.yaml
-```
+## Uninstall wger
 
-Activate the new values with the `wger-code` volume in the containers:
+To uninstall:
 
 ```bash
-helm upgrade --install wger . -n wger --create-namespace -f ../../your_values.yaml
+helm -n wger uninstall wger
+kubectl delete ns wger
 ```
 
-## Uninstall wger
+## Stop and remove minikube setup
 
-To uninstall:
+You can delete the whole cluster, including your config settings:
 
 ```bash
-helm -n wger uninstall wger
-kubectl -n wger delete -f ../../wger-code-volume.yaml
-kubectl delete ns wger
+minikube delete
 ```
 
diff --git a/README.md b/README.md
@@ -443,57 +443,7 @@ Generally persistent volumes needs to be configured depending on your setup.
 
 ## Developing locally
 
-The following is a basic instruction, for a more in-depth manual please have a look at [DEVEL.md](DEVEL.md). It also covers mounting the wger django code into the container.
-
-In order to develop locally, you will need [minikube](https://minikube.sigs.k8s.io/docs/) installed.
-It sets a local Kubernetes cluster that you can use for testing the Helm chart.
-
-If this is your first time developing a Helm chart, you'd want to try the following:
-
-```bash
-# start minikube
-$ minikube start
-
-# deploy the helm chart from the cloned git repo
-$ cd charts/wger
-$ helm dependency update
-$ helm upgrade --install wger . -n wger --create-namespace -f ../../your_values.yaml
-
-# observe that the pods start correctly
-$ watch kubectl -n wger get pods
-NAME                          READY   STATUS    RESTARTS   AGE
-wger-app-86c65dcbb9-9ftr6     5/5     Running   0          12h
-wger-postgres-0               1/1     Running   0          39h
-wger-redis-65b686bf87-cphzm   1/1     Running   0          39h
-
-# read the logs from the init container (postgres & redis check)
-$ kubectl -n wger logs -f -l app.kubernetes.io/name=wger-app -c init-container
-
-# read the logs from the wger django app
-$ kubectl -n wger logs -f -l app.kubernetes.io/name=wger-app -c wger
-PostgreSQL started :)
-*** Database does not exist, creating one now
-Operations to perform:
-  Apply all migrations: auth, authtoken, config, contenttypes, core, easy_thumbnails, exercises, gallery, gym, mailer, manager, measurements, nutrition, sessions, sites, weight
-Running migrations:
-  Applying contenttypes.0001_initial... OK
-.....
-
-# if you need to debug something in the pods, you can start a shell
-$ export POD=$(kubectl get pods -n wger -l "app.kubernetes.io/name=wger-app" -o jsonpath="{.items[0].metadata.name}")
-$ kubectl -n wger exec -it $POD -c wger -- bash
-wger@wger-app-86c65dcbb9-9ftr6:~/src$
-
-# start a port forwarding to access the webinterface
-$ echo "wger runs on: http://localhost:10001"
-$ kubectl -n wger port-forward ${POD} 10001:8000
-
-# when you are finished with the testing, stop minikube
-$ minikube stop
-
-# if you'd like to start clean, you can delete the whole cluster
-$ minikube delete
-```
+Please have a look at [DEVEL.md](DEVEL.md).
 
 
 ## Contact

diff --git a/charts/wger/Chart.yaml b/charts/wger/Chart.yaml
@@ -1,6 +1,6 @@
 ---
 apiVersion: v2
-version: 0.2.1-rc.1
+version: 0.2.1
 appVersion: latest
 name: wger
 description: A Helm chart for Wger installation on Kubernetes

diff --git a/charts/wger/templates/configmap.yaml b/charts/wger/templates/configmap.yaml
@@ -79,4 +79,4 @@ metadata:
   name: wger-pg-init
 data:
   40-grantSuperuser.sql: |
-    ALTER USER {{ .Values.postgres.userDatabase.name }} WITH SUPERUSER;
+    ALTER USER {{ .Values.postgres.userDatabase.user.value }} WITH SUPERUSER;
diff --git a/example/devel.yaml b/example/devel.yaml
@@ -0,0 +1,72 @@
+---
+#
+# Development Setup
+# see -> https://https://github.com/wger-project/helm-charts/blob/master/DEVEL.md
+#
+# Have a look at the packaged values.yaml for defaults and more settings:
+#   * https://github.com/wger-project/helm-charts/blob/master/charts/wger/values.yaml
+#
+# App settings
+app:
+  global:
+    replicas: 1
+    # image:
+    #   PullPolicy: IfNotPresent
+  nginx:
+    enabled: true
+  axes:
+    enabled: true
+    failureLimit: 10
+    # in minutes
+    cooloffTime: 30
+    # number of reverse proxies involved
+    ipwareProxyCount: 1
+    # order of magnitude from last proxy for the real ip
+    ipwareMetaPrecedenceOrder: "X_FORWARDED_FOR,REMOTE_ADDR"
+  persistence:
+    enabled: true
+  environment:
+    - name: CSRF_TRUSTED_ORIGINS
+      value: "http://localhost:10001,http://127.0.0.1:10001"
+    - name: GUNICORN_CMD_ARGS
+      value: "--timeout 240 --workers 2 --worker-class gthread --threads 3 --forwarded-allow-ips * --proxy-protocol True --access-logformat='%(h)s %(l)s %({client-ip}i)s %(l)s %({x-real-ip}i)s %(l)s %({x-forwarded-for}i)s %(l)s %(t)s \"%(r)s\" %(s)s %(b)s \"%(f)s\" \"%(a)s\"' --access-logfile - --error-logfile -"
+
+celery:
+    enabled: true
+
+ingress:
+  enabled: false
+
+postgres:
+  enabled: true
+  settings:
+    superuser:
+      value: postgres
+    superuserPassword:
+      value: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
+  userDatabase:
+    name:
+      value: wger
+    user:
+      value: wger
+    password:
+      value: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
+  service:
+    port: 5432
+
+redis:
+  enabled: true
+  auth:
+    enabled: true
+  # Additional environment variables (Redis server and Sentinel)
+  env:
+    - name: REDIS_PASSWORD
+      valueFrom:
+        secretKeyRef:
+          name: redis
+          key: redis-password
+  # Arguments for the container entrypoint process (Redis server)
+  args:
+    - "--requirepass $(REDIS_PASSWORD)"
+  service:
+    serverPort: 6379