An updated version of the Open Source Brain platform
The OSB deployment is built on top of CloudHarness. The deployment process is based on Python 3.7+ scripts. It is recommended to setup a virtual environment first.
With conda:
conda create --name osb python=3.9
conda activate osb
To install CloudHarness:
git clone https://github.com/MetaCell/cloud-harness.git
cd cloud-harness
pip install -r requirements.txt
skaffold is needed to build the images and run them on minikube. Get it here.
You can install helm from here.
CloudHarness scripts script automate the deployment process.
To manually create the helm chart to use on any Kubernetes deployment, run:
harness-deployment cloud-harness .
Kubernetes 1.19+ is supported (v1 spec)
helm repo add ingress-nginx https://kubernetes.github.io/ingress-nginx
helm repo update
helm install ingress-nginx ingress-nginx/ingress-nginx
The cert-manager must be installed in order to use letsencrypt generated certificates
To check if cert-manager is installed, run:
kubectl get pods -n cert-manager
If cert manager is installed, the command will return 3 lines.
To install the cert manager on a new cluster, run:
kubectl apply --validate=false -f https://github.com/jetstack/cert-manager/releases/download/v0.15.1/cert-manager-legacy.yaml
See also https://cert-manager.io/docs/installation/kubernetes/.
On google cloud, the Compute Engine persistent disk CSI Driver must be enabled in order for the volume cloning to work.
kubectl create ns osb2
helm install osb2 deployment/helm --namespace osb2
to install.kubectl create rolebinding osb-admin-default --clusterrole=admin --serviceaccount=osb2:default -n osb2
to allow workflows to run on namespace osb2To upgrade an existing deployment, use:
helm upgrade osb2 deployment/helm --namespace osb2 --install --reset-values [--force]
Minikube is recommended to setup locally. The procedure is different depending on where Minikube is installed. The simplest procedure is with Minikube hosted in the same machine where running the commands.
At least 6GB of ram and 4 processors are needed to run MNP
To create a new cluster, run
minikube start --memory="6000mb" --cpus=4 --disk-size=60000mb
Enable the ingress addon:
minikube addons enable ingress
Create the namespace kubectl create ns osblocal
Connect your docker registry with minikube with:
eval $(minikube docker-env)
Then run:
harness-deployment cloud-harness . -l -n osblocal -d osb.local -u -dtls -e local -i osb-portal
You do not need to run the port-forwarding commands on the local deployment.
Here, you can modify the argument of the -e
option to select what environment you want to deploy.
These correspond to the files from the osb-portal/deploy
directory.
So, selecting the environment will load specific overriding configuration files (like [APP_NAME]/deploy/values-[ENV].yaml
) specific to the environment.
If you only want to run the back-end in the minikube deployment, change the osb-portal
to workspaces
.
You can then use npm start:minikube
to point npm
to the local minikube back-end.
Note that the domain in package.json
for the start:minikube
command should match the namespace used for minikube.
Finally, run skaffold to build and run the images on minikube:
skaffold dev
On making local changes, you can re-run the harness-deployment
command to update the deployment.
With the registry on localhost:5000 run:
harness-deployment cloud-harness . -l -n osblocal -d osb.local -u -dtls -e local -i osb-portal -r registry:5000
See below to learn how to configure Minikube and forward the registry.
If Minikube is installed in a different machine, the following procedure will allow to connect kubectl.
~/.minikube
from the client to the server (skip cache and machines)~/.kube/config
from the Minikube server to the client machine (make a backup of the previous version) and adjust paths to match the home folder on the client machineIf you don't want to replace the whole content of the configuration you can copy only
the relevant entries in ~/.kube/config
from the server to the client on clusters
, context
Examples:
On clusters
- cluster:
certificate-authority: /home/user/.minikube/ca.crt
server: https://192.168.99.106:8443
name: minikube
On context
- context:
cluster: minikube
user: minikube
name: minikube
On users
- name: minikube
user:
client-certificate: /home/user/.minikube/client.crt
client-key: /home/user/.minikube/client.key
Set default context:
current-context: minikube
In the case we are not building from the same machine as the cluster (which will always happen without Minikube), we need a way to share the registry.
Procedure to share registry:5000 from a kube cluster
In the minikube installation:
minikube addons enable registry
In order to use the registry address add the following entry to the hosts file
[MINIKUBE_ADDRESS] registry
To also add the name to minikube:
% minikube ssh
$ sudo su
$ echo "127.0.0.1 registry" >> /etc/hosts
Also may need to add the host to the insecure registry on your docker configuration.
To use localhost, on the machine running the infrastructure-generate script, run
kubectl port-forward --namespace kube-system $(kubectl get po -n kube-system --field-selector=status.phase=Running | grep registry | grep -v proxy | \awk '{print $1;}') 5000:5000