Running a Standalone Container
Nx Cloud can be deployed in two ways:
- Using Kubernetes (several containers working together)
- Using a single standalone container (NOT RECOMMENDED)
The flags and the capabilities are the same between the two, but the Kubernetes setup is more robust and better documented. This document covers the latter version.
Nx Cloud consists of 3 parts:
- The stateless Nx Cloud service
- MongoDB database
- File server
By default, the container created by the nxprivatecloud/nxcloud
image will create all three: the service, the database, and the file server. Using a single container is the easiest way to set it up, but it isn't the most robust way to run Nx Cloud.
When running everything together, you won't be able to run more than one instance of the Nx Cloud container. So even though it is not required, we recommend you to run the MongoDB separately (see below how to do it).
The instructions will go through running everything together first, and then, at the end, will talk about running the database and the file server separately.
Running Nx Cloud
Step 1: Pull the Image
> docker pull nxprivatecloud/nxcloud
To update the version of Nx Cloud, pull the new version of the image and run it against the same mount (see below).
Step 2: Create a Container
Depending on how your infrastructure is set up, you can either run Nx Cloud using HTTPS or HTTP. If you have a proxy/load-balancer in front of Nx Cloud, you will likely want to run Nx Cloud using HTTP (the proxy/load-balancer will handle TLS). Otherwise, you will likely want to run Nx Cloud using HTTPS.
To create a container:
- You will need to create a directory on the host machine where data will be stored. (This is not necessary if you are running mongo yourself. See below.)
- You will need to know the URL that the Nx Cloud installation can be accessed by (see
NX_CLOUD_APP_URL
below).NX_CLOUD_APP_URL
should be accessible from your CI and dev machines.NX_CLOUD_APP_URL
can be set with an HTTP or HTTPS url. In a case where you are using a proxy/load-balancer, you can still put HTTPS (the url will be resolved by the proxy before hitting the app).NX_CLOUD_APP_URL
is likely to be an external IP/domain of the load balancer.
- If you are running Nx Cloud using HTTPS, you need to generate or obtain an SSL certificate and an SSL private key.
Once you obtain all the needed information, you can run the following:
Using HTTPS
> docker create --name cloud \
-p 443:8081 \
-e CERT_KEY="$(cat ./tools/certs/key.pem)" \
-e CERT="$(cat ./tools/certs/cert.pem)" \
-e NX_CLOUD_APP_URL="https://cloud.myorg.com" \
-e ADMIN_PASSWORD=admin \
-v /data/private-cloud:/data nxprivatecloud/nxcloud:latest
Using HTTP (no proxy)
> docker create --name cloud \
-p 80:8081 \
-e NX_CLOUD_APP_URL="http://cloud.myorg.com" \
-e ADMIN_PASSWORD=admin \
-v /data/private-cloud:/data nxprivatecloud/nxcloud:latest
Using HTTPS via proxy
> docker create --name cloud \
-p 80:8081 \
-e NX_CLOUD_APP_URL="https://cloud.myorg.com" \
-e ADMIN_PASSWORD=admin \
-v /data/private-cloud:/data nxprivatecloud/nxcloud:latest
Let's see what those options mean:
443:8081
maps the internal port 8081 to 443, so it can be accessed in the browser without specifying the port. 80: 8081 works the same way when you use HTTP instead of HTTPS.CERT_KEY
andCERT
contain the values of private key and cert. The file extensions of the cert and key files can be different, but as long as they are in the PEM format (which is the case if you use, for instance, OpenSSL), the command will work.NX_CLOUD_APP_URL
is the URL the cloud can be accessed by (e.g.,https://nxcloud.privateurl.com
). Important: Unless you are experimenting, it won't be localhost. It has to be the URL that your CI and your developer machine can reach. Also note, there is no trailing slash in the URL.ADMIN_PASSWORD
contains the password of the admin user. The admin user is created the first time you run cloud, you can remove this env variable after that. Instead of an admin password, you can also follow the instructions here to set up GitHub auth.-v /data/private-cloud:/data
sets up the volume where the data is stored./data/private-cloud
refers to a folder on your machine,/data
is the shareable folder from the Docker image.
Step 3: Run the Container
Once you create the container, you can start it using:
> docker start cloud
Imagine NX_CLOUD_APP_URL
is set to https://nxcloud.privateurl.com
.
Now, go to https://nxcloud.privateurl.com to see cloud running. You can log into the account using admin/ADMIN_PASSWORD
.
Step 4: Connect Your Workspace
Run NX_CLOUD_API=https://nxcloud.privateurl.com
nx g @nrwl/nx-cloud:init. Click on the link to connect the workspace to your admin account.
Optional step 5: set up GitHub auth
Follow the instructions here to set up GitHub OAuth authentication so you can invite other members in your team to the workspace.
Optional step 6: set up GitHub Pull Request integration
You can optionally configure Nx Cloud to post build stats directly on your GitHub pull requests.
Optional step 7: Setting Up Proxy
If your container cannot access api.nrwl.io
directly and has to talk via a proxy, you can add -e HTTPS_PROXY="https://myproxy.myorg.com"
to the container creation command.
Running the Mongo Database Separately (Recommended)
Nx Cloud uses MongoDB to store its metadata. By default, Nx Cloud is going to start a MongoDB instance and store its data in the provided volume. But you can also tell Nx Cloud to use a different MongoDB instance (e.g., if you are using MongoDB Atlas or Cosmos DB). To do this, provision the NX_CLOUD_MONGO_SERVER_ENDPOINT
env variable when creating a container, like so:
-e NX_CLOUD_MONGO_SERVER_ENDPOINT="mongodb://domain-with-mongo:27017/nrwl-api"
By default, Nx Cloud requires Mongo 4.2+. If you are using an older version of Mongo (for instance, if you are using Cosmos DB), please add
-e NX_CLOUD_USE_MONGO42=false
Using MongoDB Kubernetes Operator
The MongoDB team maintains the open source MongoDB Kubernetes Operator. You can use it to set up your own deployment of MongoDB.
Using CosmosDB
If you are deploying to Azure, you might have access to CosmosDB. See here for more information.
Using Mongo Atlas
Mongo Atlas is a great option for deploying MongoDB.
Using External File Storage
By default, Nx Cloud is going to start a file server and store the cached artifacts in the provided volume. But you can also configure Nx Cloud to use an external file storage. At the moment, only S3 and Azure Blob are supported.
Using S3/Minio
To configure S3 as a file storage, provision the AWS_S3_ACCESS_KEY_ID
, AWS_S3_SECRET_ACCESS_KEY
, and AWS_S3_BUCKET
env variables when creating the Nx Cloud docker container, like so:
-e AWS_S3_ACCESS_KEY_ID="SOMEKEY"
-e AWS_S3_SECRET_ACCESS_KEY="SOMESECRETKEY"
-e AWS_S3_BUCKET="nx-cache-bucket-name"
If you are using an accelerated bucket, add: -e AWS_S3_ACCELERATED=true
If you are using a local S3 installation (e.g., Minio), you can set the endpoint as follows:
-e AWS_S3_ENDPOINT="https://local-installation.myorg.com"
-e AWS_S3_ACCESS_KEY_ID="SOMEKEY"
-e AWS_S3_SECRET_ACCESS_KEY="SOMESECRETKEY"
-e AWS_S3_BUCKET="nx-cache-bucket-name"
Using Azure
To configure Azure Blob as a file storage, provision the AZURE_CONNECTION_STRING
, AZURE_CONTAINER
env variables when creating the Nx Cloud docker container, like so:
-e AZURE_CONNECTION_STRING="SOME-CONNECTION-STRING"
-e AZURE_CONTAINER="files"
To obtain the AZURE_CONNECTION_STRING
value go to your "Storage Account" and click on "Access Keys". You will also need to create a container in your storage account before starting the Nx Cloud container.
If you use an external file storage and an external MongoDB instance, you don't have to provision the volume.
Configure Memory Limits
By default, the Nx Cloud container is configured to run on an instance with 8GB of RAM.
If you have a container with 4GB of RAM, you can decrease the memory limits by setting the following env variables:
NX_CLOUD_FILE_SERVER_MEMORY_LIMIT=500
NX_CLOUD_API_MEMORY_LIMIT=800
NX_CLOUD_DATABASE_MEMORY_LIMIT=1
Example:
> docker create --name cloud \
-p 80:8081 \
-e NX_CLOUD_APP_URL="https://cloud.myorg.com" \
-e ADMIN_PASSWORD=admin \
-e NX_CLOUD_FILE_SERVER_MEMORY_LIMIT=500 \
-e NX_CLOUD_API_MEMORY_LIMIT=800 \
-e NX_CLOUD_DATABASE_MEMORY_LIMIT=1 \
-v /data/private-cloud:/data nxprivatecloud/nxcloud:latest
The right amount of RAM depends heavily on how you run Nx Cloud.
- The
NX_CLOUD_FILE_SERVER_MEMORY_LIMIT
value is only relevant if you use the built-in file server. - The
NX_CLOUD_DATABASE_MEMORY_LIMIT
value is only relevant if you use the built-in database.
For instance, if you use S3 to store the cached artifacts and you host Mongo DB yourself, even 2GB might be sufficient. You can set the following limit:
NX_CLOUD_API_MEMORY_LIMIT=800
If you run everything in the Nx Cloud container, then 8GB is much preferred.
Configure Artifact Expiration When Using Built-in File Server
By default, the Nx Cloud container is going to remove cached artifacts after two weeks. You can change it by setting NX_CACHE_EXPIRATION_PERIOD_IN_DAYS
when starting the container.
Example:
> docker create --name cloud \
-p 80:8081 \
-e NX_CLOUD_APP_URL="https://cloud.myorg.com" \
-e ADMIN_PASSWORD=admin \
-e NX_CACHE_EXPIRATION_PERIOD_IN_DAYS=5 \
-v /data/private-cloud:/data nxprivatecloud/nxcloud:latest
Self-Signed Certificates
If you have a self-signed certificate, you will have to provision NODE_EXTRA_CA_CERTS
. The env variable should point to a PEM file with either your certificate, or the root certificate your certificate was created from. Though this can be accomplished with a CLI command like NODE_EXTRA_CA_CERTS=./tools/certs/cert.crt nx test myapp
, you will most likely want to configure it as a global env variable (for instance in your .bashrc
file).
A self-sign certificate registered in your OS won't be picked up by Node. Node requires you to provision NODE_EXTRA_CA_CERTS
.