debugging cloud-init and user_data on cloud platforms

README.md

Debugging cloud-init and user_data

cloud init

generic cloud init help and docs

• https://cloudinit.readthedocs.io/en/latest/topics/datasources.html

AWS

treats user data and cloud init as cloud-init

• logs to
  • /var/log/cloud-init.log
  • /var/log/cloud-init-output.log

Azure

** There is a known bug with support of cloud-init on ertain distrubutions of linux**

• https://learn.microsoft.com/en-us/azure/virtual-machines/linux/cloud-init-troubleshooting
• https://learn.microsoft.com/en-us/azure/virtual-machines/custom-data
• https://cloudinit.readthedocs.io/en/latest/topics/datasources/azure.html

GCP

**GCP DOES NOT support CLOUD-INIT (even under RHEL) without help generally if your setting CLOUD init you need a custom image **

VMWARE

• https://cloudinit.readthedocs.io/en/latest/topics/datasources/vmware.html

Gists

• debug tipps https://gist.github.com/RagedUnicorn/a70f8540c68e0a41e3e097a2e29130f1
• debug script https://gist.github.com/smoser/dd45b6324be937b67b80bac7aa1a6c47
• helpers https://gist.github.com/smoser/6391b854e6a80475aac473bba4ef0310

user_data

AWS

# If you add the -v --verbose flag
# Or the -i --include flag
curl -v  http://169.254.169.254/latest/
curl -i  http://169.254.169.254/latest/
# You'll see HTTP/1.1 401 Unauthorized
export TOKEN=$(curl -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-metadata-token-ttl-seconds: 21600")

echo $TOKEN

curl -H "X-aws-ec2-metadata-token: $TOKEN" -i http://169.254.169.254/latest/
# ^-- now says HTTP/1.1 200 OK

# A nice shorthand hack to make v2 feel like v1
# (that's probably best done in the active shell only)
# and not in a .bashrc or .zshrc file
alias curl='curl -H "X-aws-ec2-metadata-token: $TOKEN"'
curl -i http://169.254.169.254/latest/
# ^-- now says HTTP/1.1 200 OK

linux

• https://aws.amazon.com/premiumsupport/knowledge-center/ec2-linux-log-user-data/
• logs to
  • /var/log/cloud-init.log
  • /var/log/cloud-init-output.log
• debug reruns https://aws.amazon.com/premiumsupport/knowledge-center/execute-user-data-ec2/
• access the metadata curl http://instance-data/latest/user-data > user-data.sh
• https://dev.to/denolfe/easily-rerun-ec2-userdata-3k18

Windows

• https://aws.amazon.com/premiumsupport/knowledge-center/ec2-windows-troubleshoot-user-data/
• https://docs.aws.amazon.com/AWSEC2/latest/WindowsGuide/ec2-windows-user-data.html
• metadata for helpful reasons https://docs.aws.amazon.com/AWSEC2/latest/WindowsGuide/instancedata-data-categories.html

Azure

• https://learn.microsoft.com/en-us/azure/virtual-machines/user-data

GCP

GCP uses a go app called google_metadata_scrpt_runner **GCP DOES NOT support CLOUD-INIT (even under RHEL) without help generally if your setting CLOUD init you need a custom image **

• https://cloud.google.com/compute/docs/instances/startup-scripts/linux
• https://pkg.go.dev/github.com/GoogleCloudPlatform/guest-agent/google_metadata_script_runner#section-sourcefiles

The script can be run locally from the instance

• sudo google_metadata_script_runner startup
• and you can view logs with sudo journalctl -u google-startup-scripts.service
• https://cloud.google.com/compute/docs/instances/startup-scripts/linux#rerunning
• under LFS teh scripts can be found at
  • /usr/share/cloud
  • /var/lib/cloud/instances

VMWare

cloud-init-status.sh

#!/usr/bin/env bash
set -euo pipefail

## script to deal with cloud init status, and make user_data script wait for cloud_init to complete when both in use 
## Terraform based usage issue https://github.com/hashicorp/terraform/issues/4668 

LOGFILE="cloud_init_status.log"
touch $LOGFILE

function log {
    local level="$1"
    local message="$2"
    local timestamp=$(date +"%Y-%m-%d %H:%M:%S")
    local log_entry="$timestamp [$level] - $message"
    
    echo "$log_entry" | tee -a "$LOGFILE"
}
function ci_status {
    if ! [ -x "$(command -v cloud-init)" ]; then
        log "info" "cloud-init is not installed"
        exit 0
    else
        CI_STATUS=$(cloud-init status)
        if [ $CI_STATUS = "done" ] || [ $CI_STATUS = "disabled" ]; then
            log "INFO" "cloud-init ${CI_STATUS}, continuing."
        else
            cloud-init status --wait
            log "INFO" "cloud-init ${CI_STATUS}, waiting!."
        fi
    fi
}
ci_status