Troubleshooting deployments

If a deployment fails, check your instances or deployments to identify the problem. Go to the Deployment Manager to see detailed messages that indicate what caused the deployment to fail.

A deployment might fail to complete successfully for one of the following reasons:

  • The deployment manager fails to create the necessary resources, such as an instance, disk, or load-balancing forwarding rule.
  • The software fails to install and configure.
  • The deployment does not configure the firewall rules correctly on the Compute Engine network.

Contents

Accessing your deployment
Deployment manager failure
Software installation failure
Network and firewall issues
Sending email from software deployments

Accessing your deployment

You can return to your solutions by revisiting the Cloud Launcher. Go to the solution page where you'll see a dropdown that shows your previous deployments. For services, you can update the service options directly from that page.

Otherwise, if your deployment is running a VM instance, you can go to the Deployment Manager in the Google Cloud Platform console. From there, you can manage your deployment and view of all the Google Cloud Platform resources that make up the deployment.

Deployment manager failure

Sometimes your project might not have enough CPU quota to create the deployment resources. Check the remaining quota for your project at Compute Engine Quotas, and request additional quota if necessary. Then try to deploy the software package again.

Software installation failure

Most solutions provide logging for your instances so you can troubleshoot software installation failures. If your deployment creates all of the requested resources but you see a deployment error message on the Deployment Manager, it's possible that the software failed to install or configure successfully. Connect to the instances in the deployment and check the logs to identify the issue.

Each entry in the log corresponds to the sequenced command that was executed on that instance. Execution terminates on the first command to exit with a non-zero exit code. The specific cause of failure will be typically be found at the end of the error log.

Network and firewall issues

If your software deployment hosts content on an external IP address, the software deployment automatically configures firewall rules on the network where your deployment runs. For example, Wordpress deployments open ports so that you can access the web console and display content over HTTP and HTTPS. If the firewall configuration fails, you might receive a HTTP 404 error when you try to connect to the web console on the external IP address.

If you need to configure the firewall manually or open additional ports for other services that run on your instances, add firewall rules to the Compute Engine network that your deployments run on.

Sending email from software deployments

Some software packages try to send email over specific ports. However, Google Compute Engine does not allow outbound connections on ports 25, 465, and 587. To send email from the instances that host your software packages, you must install and enable a partner service on your instances.

Monitor your resources on the go

Get the Google Cloud Console app to help you manage your projects.

Send feedback about...

Cloud Launcher