[Helm] Updated helm install docs based on new issues

CHANGELOG.md

@@ -7,10 +7,12 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## \[Unreleased]
 ### Added
-
+- Multi-line text attributes supported (<https://github.com/opencv/cvat/pull/6458>)
 - \[SDK\] A `DeferredTqdmProgressReporter` class, which doesn't have glitchy output
  like `TqdmProgressReporter` in certain circumstances
  (<https://github.com/opencv/cvat/pull/6556>)
+- Updated helm deploy documentation (<https://github.com/opencv/cvat/pull/6476>)
+
 
 ### Changed
 

site/content/en/docs/administration/advanced/k8s_deployment_with_helm.md

@@ -65,6 +65,13 @@ helm dependency update
  echo "$(minikube ip) cvat.local" | sudo tee -a /etc/hosts
  ```
 
+ Note for Microk8s use:
+ The external ip should be the ip from the cluster you are running cvat from, you can obtain it using ```microk8s config```. Make sure to add it to the hosts file as mentioned above.
+ ```sh
+ echo "<your microk8s cluster ip> cvat.local" | sudo tee -a /etc/hosts
+ ```
+
+
 ## Configuration
 1. Create `values.override.yaml` file inside `helm-chart` directory.
 1. Fill `values.override.yaml` with new parameters for chart.
@@ -165,7 +172,7 @@ Execute following command from repo root directory
 HELM_RELEASE_NAMESPACE="<desired_namespace>" &&\
 HELM_RELEASE_NAME="<release_name>" &&\
 BACKEND_POD_NAME=$(kubectl get pod --namespace $HELM_RELEASE_NAMESPACE -l tier=backend,app.kubernetes.io/instance=$HELM_RELEASE_NAME -o jsonpath='{.items[0].metadata.name}') &&\
-kubectl exec -it --namespace $HELM_RELEASE_NAMESPACE $BACKEND_POD_NAME -c cvat-backend-app-container -- python manage.py createsuperuser
+kubectl exec -it --namespace $HELM_RELEASE_NAMESPACE $BACKEND_POD_NAME -c cvat-app-backend-server-container -- python manage.py createsuperuser
 ```
 ## FAQ
 
@@ -319,3 +326,32 @@ cvat:
  claimName: my-claim-name
 
 ```
+
+### Im receiving service unavailable errors
+There could be multiple explanations for that, a hint can often be found by inspecting the logs of the running pods, some possible explanations are
+1. Pod multi attach errors:
+
+ This can happen when running cvat using helm chart in a production environment. Pods can get scheduled to run on different nodes, however the same volume using ReadWriteOnce can only be accessed from one node at a time. You can either change the volume access mode from ReadWriteOnce to ReadWriteMany if your volume supports it, or add pod affinity rules to all the backend pods in the cvat values.yaml file to make sure they are all scheduled on the same node like so:
+ ```yaml
+ affinity:
+ podAffinity:
+ requiredDuringSchedulingIgnoredDuringExecution:
+ - labelSelector:
+ matchExpressions:
+ - key: tier
+ operator: In
+ values:
+ - backend
+ topologyKey: "kubernetes.io/hostname"
+ ```
+ 2. your DB is not responding, ( database errors like health_check not found etc)
+
+ It is possible that the migrations did not run automatically, you may need to run them manually:
+ ```sh
+ HELM_RELEASE_NAMESPACE="<desired_namespace>" &&\
+ HELM_RELEASE_NAME="<release_name>" &&\
+ BACKEND_POD_NAME=$(kubectl get pod --namespace $HELM_RELEASE_NAMESPACE -l tier=backend,app.kubernetes.io/instance=$HELM_RELEASE_NAME -o jsonpath='{.items[0].metadata.name}') &&\
+ kubectl exec -it --namespace $HELM_RELEASE_NAMESPACE $BACKEND_POD_NAME -c cvat-app-backend-server-container -- python manage.py migrate &&\
+ kubectl exec -it --namespace $HELM_RELEASE_NAMESPACE $BACKEND_POD_NAME -c cvat-app-backend-server-container -- python manage.py health_check
+ ```
+