Troubleshooting AWS CloudShell

While using AWS CloudShell, you might encounter issues, such as when you launch CloudShell or perform key tasks using the shell command line interface. The information that's covered in this chapter covers how to troubleshoot some of the common issues that you might encounter.

For answers to a variety of questions about CloudShell, see the AWS CloudShell FAQs. You can also search for answers and post questions in the AWS CloudShell Discussion Forum. When you enter this forum, you might be required to sign in to AWS. You can also contact us directly.

Troubleshooting errors

When you come across any of the following indexed errors, you can use the following solutions to resolve these errors.

Unable to start the environment. To retry, refresh the browser or restart by selecting Actions, Restart AWS CloudShell

Issue: When you attempt to launch AWS CloudShell from the AWS Management Console, you're denied access even after you have required permissions from your IAM administrator and you have refreshed your browser or restarted CloudShell.

Solution: Contact AWS Support.

(back to top)

Unable to start the environment. You don't have required permissions. Ask your IAM administrator to grant access to AWS CloudShell

Issue: When you attempt to launch AWS CloudShell from the AWS Management Console, you're denied access and notified that you don't have required permissions.

Cause: The IAM identity that you're using to access AWS CloudShell lacks the necessary IAM permissions.

Solution: Request your IAM administrator to provide you with the necessary permissions. They can do this either through adding an attached AWS managed policy (AWSCloudShellFullAccess) or an embedded inline policy. For more information, see Managing AWS CloudShell access and usage with IAM policies.

(back to top)

Unable to access AWS CloudShell command line

Issue: After modifying a file that the compute environment uses, you can't access the command line in AWS CloudShell.

Solution: If you lose access after incorrectly modifying .bashrc or any other file, you can return AWS CloudShell to its default settings by deleting your home directory.

(back to top)

Unable to ping external IP addresses

Issue: When you run a ping command from the command line (for example, ping amazon.com), you receive the following message.

ping: socket: Operation not permitted

Cause: The ping utility uses Internet Control Message Protocol (ICMP) to send echo requests packets to a target host. It waits for an echo to reply from the target. Because the ICMP protocol isn't enabled in AWS CloudShell, the ping utility doesn't operate in the shell's compute environment.

Solution: Due to the fact that ICMP is not supported in AWS CloudShell, you can run the following command to install Netcat. Netcat is a is a computer networking utility for reading from, and writing to, network connections using TCP or UDP.

sudo yum install nc
nc -zv www.amazon.com 443

(back to top)

There were some issues preparing your terminal

Issue: When trying to access AWS CloudShell using the Microsoft Edge browser, you can't start a shell session, and the browser displays an error message.

Cause: AWS CloudShell isn't compatible with earlier versions of Microsoft Edge. You can access AWS CloudShell using the latest four major versions of supported browsers.

Solution: Install an updated version of Edge browser from the Microsoft site.

(back to top)

Arrow keys not working correctly in PowerShell

Issue: In normal operation, you can use arrow keys to navigate the command line interface and scan backwards and forwards through your command history. But, when you press arrow keys in certain versions of PowerShell on AWS CloudShell, letters might be incorrectly outputted.

Cause: The situation where arrow keys incorrectly output letters is a known issue with PowerShell 7.2.x versions running on Linux.

Solution: To strip out escape sequences that modify the behavior of arrow keys, edit the PowerShell profile file and set the $PSStyle variable to PlainText.

1. In the AWS CloudShell command line, enter the following command to open the profile file.

   vim ~/.config/powershell/Microsoft.PowerShell_profile.ps1 

   Note

   If you're already in PowerShell, you can also open the profile file in the editor with the following command.

   vim $PROFILE

2. In the editor, go to the end of the file's existing text, press i to enter Insert mode, and then add the following statement.

   $PSStyle.OutputRendering = 'PlainText'

3. After you make the edit, press Esc to enter the command mode. Next, enter the following command to save the file and exit the editor.

   :wq

Note

Your changes take effect the next time you start PowerShell.

(back to top)

Unsupported Web Sockets cause a failure to start CloudShell sessions

Issue: When trying to start AWS CloudShell, you repeatedly receive the following message: Failed to open sessions : Timed out while opening the session.

Cause: CloudShell depends on the WebSocket protocol, which allows for two-way interactive communication between your web browser and AWS CloudShell. If you're using a browser in a private network, secure access to the internet is probably facilitated by proxy servers and firewalls. WebSocket communication can usually traverse proxy servers without a problem. But, in some cases, proxy servers prevent WebSockets from working correctly. If this issue occurs, CloudShell can't start a shell session and the attempt to connect eventually times out.

Solution: A connection timeout might be caused by an issue other than unsupported WebSockets. If this is the case, first refresh the browser window where the CloudShell command line interface is located.

If you're still getting timeout errors after the refresh, see the documentation for your proxy server. And, make sure that your proxy server is configured to allow Web Sockets. Alternatively, contact your network's system administrator.

Note

Say that you want to define granular permissions by allowlisting specific URLs. You can add part of the URL that the AWS Systems Manager session uses to open a WebSocket connection for sending input and receiving outputs. Your AWS CloudShell commands are sent to that Systems Manager session.

The format for this StreamUrl that's used by Systems Manager is wss://ssmmessages.region.amazonaws.com/v1/data-channel/session-id?stream=(input|output).

The region represents the Region identifier for an AWS Region that's supported by AWS Systems Manager. For example, us-east-2 is the Region identifier for the US East (Ohio) Region.

Because the session-id is created after a particular Systems Manager session is successfully started, you can only specify wss://ssmmessages.region.amazonaws.com when you update your URL allowlist. For more information, see the StartSession operation in the AWS Systems Manager API Reference.

(back to top)

Unable to import the AWSPowerShell.NetCore module

Issue: When you import the AWSPowerShell.NetCore module in PowerShell by Import-Module -Name AWSPowerShell.NetCore, you receive the following error message:

Import-Module: The specified module 'AWSPowerShell.NetCore' was not loaded because no valid module file was found in any module directory.

Cause: The AWSPowerShell.NetCore module is replaced by the per-service AWS.Tools modules in AWS CloudShell.

Solution: Any explicit import statements might no longer be required or need to be changed to the related per-service AWS.Tools module.

• For most cases, as long as no .Net types are used, you don’t need any explicit import statement. The following are examples of import statements.

  • Get-S3Bucket

  • (Get-EC2Instance).Instances

• If .Net types are used, import the service-level module (AWS.Tools.<Service>). The following is example syntax.

  Import-Module -Name AWS.Tools.EC2
  $InstanceTag = [Amazon.EC2.Model.Tag]::new("Environment","Dev")

  Import-Module -Name AWS.Tools.S3 
  $LifecycleRule = [Amazon.S3.Model.LifecycleRule]::new()

For more information, see the version 4 announcement for the AWS Tools for PowerShell.

(back to top)

Docker is not running when using AWS CloudShell

Issue: Docker is not running properly when using AWS CloudShell. You receive the following error message: docker: Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running?.

Solution: Try restarting your environment. This error message can occur when you run Docker in AWS CloudShell in a Region that doesn't support it. Ensure you are running Docker in a supported Region, for information on which Regions support the use of Docker containers with AWS CloudShell, see Docker Regions.

Docker has ran out of disk space

Issue: You are receiving the following error message: ERROR: failed to solve: failed to register layer: write [...]: no space left on device.

Cause: The Dockerfile is exceeding the available disk space in AWS CloudShell. This can be caused due to large individual images or too many pre-existing Docker images.

Solution: Run df -h to find the disk usage. Run sudo du -sh /folder/folder1 to asses the size of certain folders that you feel may be large and consider deleting other files to free up space. One option would be to consider removing unused Docker images by running docker rmi. You should be aware that Docker has limited space in the environment, for more information on Docker, see the Docker Documentation guide.

docker push is timing out and keeps retrying

Issue: When you run docker push it is timing out and continues to retry with no success.

Cause: This can be caused as a result of missing permissions, pushing to the wrong repository or a lack of authentication.

Solution: To try and resolve this issue ensure you are pushing to the correct repository. Run docker login to properly authenticate. Ensure that you have all the required permissions for pushing to an Amazon ECR repository.

Unable to access resources within VPC from my AWS CloudShell VPC environment

Issue: Unable to access resources within the VPC while using my AWS CloudShell VPC environment.

Cause: Your AWS CloudShell VPC environment inherits the network settings of your VPC.

Solution: To resolve this issue ensure that your VPC is set up correctly to access your resources. For more information, see VPC documentation Connect your VPC to other networks and the and the Network Access Analyzer documentation Network Access Analyzer. You can find the IPv4 address that the AWS CloudShell VPC environment is using, by running the command `ip -a` inside your environment in the command line prompt, or on the VPC Console page.

The ENI used by AWS CloudShell for my VPC environment is not cleaned up

Issue: Unable to clean up the ENI used by AWS CloudShell for my VPC environment.

Cause: ec2:DeleteNetworkInterface permission is not enabled for your role.

Solution: To resolve this issue, ensure that ec2:DeleteNetworkInterface permission is enabled for your role as shown in the following sample script:

{
  "Effect": "Allow",
  "Action": [
    "ec2:DeleteNetworkInterface"
  ],
  "Condition": {
    "StringEquals": {
      "aws:ResourceTag/ManagedByCloudShell": ""
    }
  },
  "Resource": "arn:aws:ec2:*:*:network-interface/*"
}

User with CreateEnvironment permission for only VPC environments also has access to public AWS CloudShell environments

Issue: User restricted with CreateEnvironment permission for only VPC environments is also able to access public AWS CloudShell environments.

Cause: When you limit CreateEnvironment permissions for creation of VPC environments only and if you have already created a public environment, you will keep your access to the existing public CloudShell environment until this environment is deleted using the web user interface. But if you have never used CloudShell before, you will not have access to public environments.

Solution: To restrict access to public AWS CloudShell environments, the IAM administrator must first update the IAM policy with the restriction, and then the user must manually delete the existing public environment using the AWS CloudShell web user interface. (Actions → Delete CloudShell environment).