> ## Documentation Index
> Fetch the complete documentation index at: https://openops-ecb4f397-mintlify-add-install-command-explanation-5.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.
# GCP Compute Engine Deployment
> How to deploy OpenOps on a GCP Compute Engine instance
export const target_1 = "VM"
export const target_0 = "VM"
This guide is for testing and evaluation purposes only and is not intended for production deployments. Please reach out to us at [support@openops.com](mailto:support@openops.com) if you'd like to learn how to set up OpenOps in a production environment.
This guide explains how to install the OpenOps Docker Compose release on a newly created GCP Compute Engine VM instance.
It assumes you have appropriate permissions on an existing Google Cloud Platform (GCP) project.
## Initial deployment
### Create a new VM instance
1. In the Google Cloud Console, navigate to **Compute Engine** → **VM instances**.
2. Click **Create Instance**.
3. Under **Machine configuration**, configure the following:
* **Name**: Enter a name for your virtual machine.
* **Region**: Choose a region close to your users (e.g., `us-east1`).
* **Zone**: Choose a zone close to your users (e.g., `us-east1-b`), or leave the default to let Google choose one for you.
* **Series**: Choose any recommended series (e.g., **E2**) or another family you prefer.
* **Machine type**: Choose a machine size similar to **e2-standard-2** or larger. Avoid very small machines, as OpenOps may need additional CPU/RAM for running Docker containers smoothly.
4. Under **OS and storage**:
1. Click **Change**.
2. Select **Ubuntu** as the operating system (Ubuntu 24.04 LTS if available, or a close alternative).
3. Increase **Size** to at least 50GB to accommodate Docker images and databases.
4. Click **Select**.
5. Under **Networking**, select **Allow HTTP traffic**. This will automatically create a firewall rule to open port 80.
6. Keep defaults for other settings (or adjust according to your preferences), then click **Create** to launch the VM.
### Configure firewall for SSH and HTTP
Depending on your project settings, a default firewall rule might already allow SSH access. By checking **Allow HTTP traffic** while creating the instance, HTTP (port 80) is also open. If you need to adjust or review these rules:
1. In the **VPC network** menu, go to **Firewall**.
2. Locate or create firewall rules for:
* **SSH** (TCP/22) from your IP or a limited source range.
* **HTTP** (TCP/80) for public access or your desired source range.
### Connect to the instance
1. Once your VM is running, go to the **VM instances** list.
2. Click **SSH** next to your instance, then click **Open in browser window** to launch an in-browser terminal.
### Install OpenOps
Run the following command in your terminal to install, update, configure and run OpenOps:
```shell theme={null}
curl -fsS https://openops.sh/install | sh
```
Alternatively, you can follow the manual installation steps:
1. Install the `unzip` utility:
```shell theme={null}
sudo apt install unzip
```
2. Create a new directory and download the OpenOps release files:
```shell theme={null}
mkdir -p openops && cd openops && \
wget https://github.com/openops-cloud/openops/releases/download/0.6.18/openops-dc-0.6.18.zip && \
unzip -o openops-dc-0.6.18.zip && cp -n .env.defaults .env
```
3. Update the application URL to use the instance's external IP address:
```shell theme={null}
EXTERNAL_IP=$(curl -4 -s ifconfig.co)
sed -i "s|http://localhost|http://$EXTERNAL_IP|g" .env
```
4. Open the `.env` file in the OpenOps installation folder. Change the values of the following variables that represent credentials. Do it now, as you won't be able to change these values after the initial deployment:
* `OPS_OPENOPS_ADMIN_EMAIL`: the email of your OpenOps installation's root admin account.
* `OPS_OPENOPS_ADMIN_PASSWORD`: the password of your OpenOps installation's root admin account.
* `OPS_POSTGRES_USERNAME`: the username of the Postgres database that OpenOps uses.
* `OPS_POSTGRES_PASSWORD`: the password of the Postgres database that OpenOps uses.
* `OPS_ANALYTICS_ADMIN_PASSWORD`: the password of the [OpenOps Analytics](/reporting-analytics/data-visualization/) admin account (the username is hardcoded to `admin`).
5. Install Docker using Snap:
```shell theme={null}
sudo snap install docker
```
6. Pull the images and start the Docker containers:
```shell theme={null}
sudo COMPOSE_PARALLEL_LIMIT=4 docker compose pull -q && sudo docker compose up -d
```
If you encounter a rate limit, rerun the pull command. Note that pulling images may take several minutes.
If you did not update credentials in the `.env` file, the default username is `admin@openops.com` and the password is `please-change-this-password-1`.
You can now access the OpenOps application by navigating to the public IP address of your {target_0} (e.g., `http://20.20.20.20`).
## Network ports used by OpenOps
By default, OpenOps does not require any inbound ports other than the application port `80`.
The following ports are used by bundled services and, in most cases, should not be exposed:
* `5432`: The PostgreSQL database used by OpenOps. Expose this port only if you need direct database access, such as for [connecting OpenOps to external analytics tools](/reporting-analytics/connecting-to-external-tools/). Restrict access to a VPN or a trusted IP range.
* `6379`: OpenOps' internal Redis service. Expose this port only if required for debugging or monitoring purposes. Restrict access to a VPN or a trusted IP range.
## External databases
To use external PostgreSQL or Redis databases, modify the relevant variables in the `.env` file. You can disable the corresponding containers by adding a profile in the `docker-compose.yml` file:
```yaml theme={null}
services:
postgres:
profiles: ["db"]
```
If you remove or disable the `db` profile in `.env` or in Docker Compose, that container won't start.
After making any changes to the `.env` file, restart the OpenOps containers:
```shell theme={null}
sudo docker compose down
sudo docker compose up -d
```
## Enabling TLS
For production usage, it's recommended to enable TLS (HTTPS). In addition to the security aspect, this also ensures that [workflow templates](/workflow-management/workflow-templates/) load properly in all browsers.
### Automatically
The easiest way to enable TLS is to use an OpenOps script that requests and sets up a TLS certificate from Let's Encrypt. Before running the script, make sure you have a domain name that points to your {target_1}'s external IP address. If you're configuring DNS right before running the script, you may need to wait for the DNS change to propagate.
1. Run the following command in your terminal:
```shell theme={null}
curl -fsS https://openops.sh/tls | sh
```
2. When prompted, enter a domain name that points to the external IP address of your {target_1}.
3. When prompted, enter an email address to receive certificate-related notifications from Let's Encrypt.
The script will use the Certbot library to request a certificate for your domain from Let's Encrypt. It receives and saves the certificate, updates the OpenOps configuration file accordingly, and restarts OpenOps.
By default, the certificate expires in 3 months. See [https://certbot.org/renewal-setup](https://certbot.org/renewal-setup) if you want to configure auto-renewal.
### Manually
Alternatively, you can create a TLS certificate yourself. This lets you use DNS validation from Let's Encrypt (rather than the HTTP validation the automatic script performs) or request a certificate from a different provider.
To set up TLS manually:
1. Obtain certificate and private key files from your certificate provider.
2. Upload the certificate files to your OpenOps installation under the `tls` directory:
* `/tls/cert.pem` (certificate file)
* `/tls/key.pem` (private key file)
3. Open the `.env` file in your OpenOps installation directory and update the following variables:
* Set `OPS_NGINX_CONFIG_FILE` to `nginx.gateway.tls.conf`
* Set `OPS_PUBLIC_URL` to the secure URL of your OpenOps installation (e.g. `https://openops.example.com`)
After making any changes to the `.env` file, restart the OpenOps containers:
```shell theme={null}
sudo docker compose down
sudo docker compose up -d
```
## Disabling host validation
By default, OpenOps does not allow workflows to call internal network addresses such as `127.0.0.1` or `192.168.0.0`. This affects HTTP and SMTP actions, as well as webhook triggers. Host validation protects users from creating workflows that could accidentally or maliciously access internal services, scan networks, or escalate privileges.
You may need to disable this check in certain circumstances, such as in non-production deployments or when workflows intentionally interact with internal-only infrastructure.
To disable host validation, open the `.env` file in your installation folder and set the `OPS_ENABLE_HOST_VALIDATION` environment variable to `false`.
After making any changes to the `.env` file, restart the OpenOps containers:
```shell theme={null}
sudo docker compose down
sudo docker compose up -d
```
Disabling this check removes an important safety guard and may allow workflows to access internal infrastructure. Use caution and avoid disabling it in production.
## Configuring allowed SMTP ports
By default, OpenOps allows SMTP connections on ports 25, 465, 587, and 2525. Administrators can customize which SMTP ports are available by setting the `OPS_SMTP_ALLOWED_PORTS` environment variable in the `.env` file.
To configure allowed SMTP ports, open the `.env` file in your installation folder and set the `OPS_SMTP_ALLOWED_PORTS` variable to a comma-separated list of port numbers:
```shell theme={null}
OPS_SMTP_ALLOWED_PORTS=25,465,587,2525
```
Port numbers must be valid TCP ports in the range 1-65535. At least one valid port must be configured. Invalid port numbers will be filtered out.
After making any changes to the `.env` file, restart the OpenOps containers:
```shell theme={null}
sudo docker compose down
sudo docker compose up -d
```
## Updating OpenOps to a newer version
See [Updating OpenOps](/getting-started/updating-openops).
## Support
Feel free to join our [Slack community](https://slack.openops.com) if you have any questions or need help with your installation.