Troubleshoot instance issues
Topics
- Tags must be set correctly
- Amazon CodeDeploy agent must be installed and running on instances
- Deployments do not fail for up to an hour when an instance is terminated during a deployment
- Analyzing log files to investigate deployment failures on instances
- Create a new CodeDeploy log file if it was accidentally deleted
- Troubleshooting “InvalidSignatureException – Signature expired: [time] is now earlier than [time]” deployment errors
Tags must be set correctly
Use the list-deployment-instances command to confirm the instances used for a deployment are tagged correctly. If an EC2 instance is missing in the output, use the EC2 console to confirm the tags have been set on the instance. For more information, see Working with tags in the console in the Amazon EC2 User Guide for Linux Instances.
Note
If you tag an instance and immediately use CodeDeploy to deploy an application to it, the instance might not be included in the deployment. This is because it can take several minutes before CodeDeploy can read the tags. We recommend that you wait at least five minutes between the time you tag an instance and attempt to deploy to it.
Amazon CodeDeploy agent must be installed and running on instances
To verify the CodeDeploy agent is installed and running on an instance, see Verify the CodeDeploy agent is running.
To install, uninstall, or reinstall the CodeDeploy agent, see Install the CodeDeploy agent.
Deployments do not fail for up to an hour when an instance is terminated during a deployment
CodeDeploy provides a one-hour window for each deployment lifecycle event to run to completion. This provides ample time for long-running scripts.
If the scripts don't run to completion while a lifecycle event is in progress (for example, if an instance is terminated or the CodeDeploy agent is shut down), it might take up to an hour for the status of the deployment to be displayed as Failed. This is true even if the timeout period specified in the script is shorter than an hour. This is because when the instance is terminated, the CodeDeploy agent shuts down and cannot process more scripts.
If an instance is terminated between lifecycle events or before the first lifecycle event step starts, the timeout occurs after just five minutes.
Analyzing log files to investigate deployment failures on instances
If the status of an instance in the deployment is anything other than
Succeeded, you can review the deployment log file data to help identify the
problem. For information about accessing deployment log data, see View log data for CodeDeploy EC2/On-Premises
deployments.
Create a new CodeDeploy log file if it was accidentally deleted
If you accidentally delete the deployment log file on an instance, CodeDeploy does not create a replacement log file. To create a new log file, sign in to the instance, and then run these commands:
For an Amazon Linux, Ubuntu Server, or RHEL instance, run these commands in this order, one at a time:
systemctl stop codedeploy-agent
systemctl start codedeploy-agent
For a Windows Server instance:
powershell.exe -Command Restart-Service -Name codedeployagent
Troubleshooting “InvalidSignatureException – Signature expired: [time] is now earlier than [time]” deployment errors
CodeDeploy requires accurate time references to perform its operations. If the date and time on your instance are not set correctly, they might not match the signature date of your deployment request, which CodeDeploy rejects.
To avoid deployment failures related to incorrect time settings, see the following topics: