🐞 Known Issues During First Launch (EC2 from AWS Marketplace)

This document contains a list of known issues that may occur during the initial launch of the EC2 instance when deploying BookStack via AWS Marketplace. These behaviors are considered normal and self-recovering in most cases.

 

⚠️ 1. Transient DNF Cache Failures

Symptoms:

• Cloud-init fails with messages like:

[Errno 2] No such file or directory: '/var/cache/dnf/.../*.rpm'
Failed to run module scripts-user

Cause:

• This issue is caused by a temporary inconsistency between metadata and actual RPM packages in Amazon Linux 2023 mirrors.

• It may also be triggered by race conditions within cloud-init (e.g., DNF downloads package, then cache becomes unavailable before install completes).

Workaround:

• The Auto Scaling Group will automatically replace the failed instance.

• On the second attempt, the packages are downloaded again and the process typically succeeds.

Recommendation:

• Add retry logic and DNF cache clearing in your user-data, such as:

dnf clean all
dnf makecache
retry dnf install -y ...

 

⚠️ 2. cloud-init Failure After SELinux Reboot

Symptoms:

⚠️ On Amazon Linux 2023, the system may reboot automatically on first launch to apply SELinux settings.

This may cause the initial cloud-init status to show as failed, even though the environment is ready.

Log shows:

Rebooting machine to apply SELinux kernel commandline setting
Failed to run module scripts-user
Ran 20 modules with 1 failures

Cause:

• On Amazon Linux 2023, the first boot may trigger a mandatory reboot to apply SELinux or kernel settings.

• This interrupt happens during cloud-init execution, marking it as failed — even if everything worked.

Workaround:

• The system reboots automatically and continues working fine.

• No action is required — this is an expected behavior.

 

⚠️ 3. Duplicate Welcome Message in Shell

Symptoms:

sh-5.2$ sudo su - ec2-user
Last login: Tue Apr  8 08:03:38 UTC 2025 on pts/1
 Welcome to your EC2 instance!
-----------------------------
 Operating System: GNU/Linux
 Hostname: ip-172-31-7-183.ec2.internal
 Current Date and Time: Tue Apr  8 08:14:25 UTC 2025
-----------------------------
 Welcome to your EC2 instance!
-----------------------------
 Operating System: GNU/Linux
 Hostname: ip-172-31-7-183.ec2.internal
 Current Date and Time: Tue Apr  8 08:14:25 UTC 2025

Cause:

• The base AMI used by AWS during scanning includes pre-installed user-data scripts or messages, and this overlaps with your current user-data.

Resolution:

• Ignore — has no effect on functionality.

• Optional: Switch back to dynamically generated AMI if you want to avoid overlap - /aws/service/ami-amazon-linux-latest/al2023-ami-kernel-default-arm64.

 

⚠️ 4. EC2 Instance Logs Are Not Persisted to File

Symptoms:

• File /var/log/journal/log-export.log is missing after instance launch.

Cause:

• The logging unit contains a misconfigured ExecStart command in the file: /etc/systemd/system/journald-export.service

Incorrect line:

ExecStart=/bin/bash -c "journalctl -f -o short-iso > tee -a /var/log/journal/log-export.log"

Correct line should be:

ExecStart=/bin/bash -c "journalctl -f -o short-iso > /var/log/journal/log-export.log"

Resolution:

sudo systemctl daemon-reload
sudo systemctl enable --now journald-export.service
sudo systemctl start journald-export.service

 

✅ Summary

These issues do not indicate a failure of the product, but are a result of how cloud-init, Amazon Linux 2023, and Marketplace validation behave. 

🔧 The duplicate welcome screen and missing EC2 instance logs are known issues and will be addressed in a future release.