Find the Cloud Waste Contest

CloudQuery News

Find the Cloud Waste Contest

•

TL;DR Contest alert 🚨 Win CloudQuery swag (t-shirts, sweatshirts, mugs) by finding wasteful AWS resources in our public S3 bucket. Sync the data with CloudQuery, write SQL queries to identify unattached volumes, dead weight storage, and public IPs. Post your results, claim your prizes. Bonus: Use your own AWS data and show your boss the savings for double swag!

We've loaded AWS infrastructure data into a public S3 bucket. Sync it with CloudQuery into your own destination, run SQL queries, and hunt down wasteful resources.

Bonus: Got your own AWS account? Sync your own data, share your queries, and we'll send you swag too. Show your boss how much money you're saving and win double.

The Challenges #

Level 1: Unattached Volumes → CloudQuery T-shirt #

Find all EBS volumes sitting in 'available' state that aren't attached to any EC2 instance. These are burning storage costs for nobody.

Hint: You're looking for volumes where the state shows they're available (not in-use). Check for volumes that don't appear in any instance's block device mappings. Think about using a LEFT JOIN between volumes and instances, then filtering for NULL matches.

Level 2: Dead Weight Storage → CloudQuery Mug #

Find volumes attached to EC2 instances that aren't running. Stopped or terminated instances don't need their volumes mounted, but they're still racking up charges.

Hint: Look for volumes that show 'in-use' as their state, but the instance they're attached to has a state other than 'running' or 'pending'. Instance state is stored as JSON - you'll need to extract it properly. Consider states like 'stopped', 'stopping', 'terminated', etc.

Level 3: Public IPs → CloudQuery Sweatshirt #

Find ALL cloud resources with public IPs. This includes EC2 instances with public IPs, Elastic IPs (especially unassociated ones costing you money), internet-facing load balancers, and publicly accessible S3 buckets.

Hint: This is a multi-table challenge. You'll need to query EC2 instances with public_ip_address fields, Elastic IPs (check their association status), load balancers (both v1 Classic and v2 ALB/NLB) with scheme='internet-facing', and S3 buckets with public access (check bucket policies for Principal: "*"). Consider using UNION queries to combine results from different resource types.

How It Works #

We've loaded AWS data into our public S3 bucket
You configure CloudQuery to use our S3 bucket as your SOURCE
You pick your own destination - SQLite, DuckDB, PostgreSQL, BigQuery, Snowflake, CSV files, whatever you want
Run a CloudQuery sync to pull the data from S3 to your destination
Query your local data to solve the challenges
Post your results and claim your swag

S3 Bucket Details:

Bucket: cloudquery-contest-readonly-2025
Region: us-east-1
Format: Parquet files
Public read-only access (no AWS credentials needed)

Tables available:

aws_ec2_instances
aws_ec2_ebs_volumes
aws_ec2_eips
aws_elbv1_load_balancers
aws_elbv2_load_balancers
aws_s3_buckets
aws_s3_bucket_policies
aws_ec2_network_interfaces

Getting Started: Syncing from S3 #

Step 1: Install CloudQuery CLI

Download from CloudQuery.io/download

Step 2: Create your config file

Initialize a new CloudQuery project:

cloudquery init

This creates a starter config file. You need TWO things: an S3 SOURCE (our bucket) and a DESTINATION (your choice).

Edit your config file to sync from our S3 bucket to SQLite locally:

kind: source
spec:
  name: s3
  path: cloudquery/s3
  registry: cloudquery
  version: "v1.8.17"
  tables: ["*"]
  destinations:
    - sqlite
  spec:
    bucket: "cloudquery-contest-readonly-2025"
    region: "us-east-1"
    path: "cloudquery-contest/{{TABLE}}/*.parquet"
    format: "parquet"

---
kind: destination
spec:
  name: sqlite
  path: cloudquery/sqlite
  registry: cloudquery
  version: "v2.13.1"
  write_mode: "overwrite-delete-stale"
  spec:
    connection_string: "./cloudquery-contest.db"

Want to use a different destination? Check out all available options:

S3 Source Plugin: hub.CloudQuery.io/plugins/source/CloudQuery/s3
All Destination Plugins: hub.CloudQuery.io/plugins/destination

Popular choices include DuckDB, BigQuery, Snowflake, CSV files, and 50+ others. The setup is the same - just swap the destination spec.

Step 3: Run the sync

cloudquery sync config.yml

This pulls all the AWS data from our S3 bucket into your destination. No AWS credentials needed - the bucket is public.

Step 4: Query your local data

Now write your SQL queries against your destination to solve the challenges!

# If using SQLite
sqlite3 cloudquery-contest.db

# If using DuckDB
duckdb cloudquery-contest.duckdb

# If using your own Postgres
psql your_connection_string

Using Your Own AWS Data #

Got your own AWS account? You can sync your own data instead:

Configure CloudQuery with the AWS source plugin
Sync your AWS resources to any destination
Run the same challenge queries against your data
Share your queries and results with us for swag

This way you can show your boss the actual cost savings in your infrastructure.

Claiming Your Prize #

Post your work anywhere and tag us, or send it directly:

X (Twitter): Tag @cloudqueryio with your results
LinkedIn: Tag @cloudqueryio
Threads: Tag @cloudqueryio
Community forum: Post in General community.CloudQuery.io
Email: [email protected]

What to include:

Screenshot showing your CloudQuery sync completion
The queries you wrote
Query results showing your answers

Need help? #

Quickstart guide: CloudQuery.io/docs/quickstart
S3 source plugin: hub.CloudQuery.io/plugins/source/CloudQuery/s3
AWS source plugin: hub.CloudQuery.io/plugins/source/CloudQuery/aws
All destination plugins: hub.CloudQuery.io/plugins/destination
Questions? Ask in the community forum or hit us up on social. Let's see what you find.

FAQ #

Q: Do I need AWS credentials to participate? A: No. The S3 bucket is public and read-only. You just need CloudQuery CLI and a destination to sync to.

Q: Can I download the files directly from S3 without syncing? A: Technically yes, but that defeats the purpose. We want you to experience CloudQuery's sync workflow. Plus, you need to show proof of sync for the swag.

Q: What if I get the wrong answer? A: That's fine - we want to see your process. Submit what you found, show your queries, and we'll work with you. This is about learning CloudQuery, not being perfect.

Q: Do I need to complete all three challenges? A: No. Complete what you can. Each challenge you finish gets you swag. Challenge 1 = T-shirt. Challenge 2 = Mug. Challenge 3 = Sweatshirt.

Q: Which destination should I use? A: Whatever you're comfortable with. SQLite requires zero setup. DuckDB is great for analytics. Your own Postgres if you have one running. CSV files if you want to load into Excel. Pick what works for you.

Q: Can I use AI tools to help write queries? A: Yes. We're not testing your SQL memorization skills. We want to see you use CloudQuery and find the answers. Use whatever tools help you learn.

Q: What if my sync fails? A: Make sure you have write permissions for your destination. Run with --log-level debug to see detailed errors. Ask in our community forum - we're here to help.

Q: Is this real AWS data? A: It's based on real AWS resource structures, but sanitized and anonymized for the contest. All account IDs, resource names, and sensitive info have been changed.

Q: Can I share my queries with others? A: Yes, but where's the fun in that? The learning happens when you figure it out yourself. Share your approach, not your answers.

Q: Do I need to know AWS to participate? A: Basic understanding helps, but the hints tell you what to look for. This is more about SQL and CloudQuery than deep AWS knowledge.

Q: Can I really use my own AWS data? A: Absolutely. Just share your queries and results (sanitized if needed) and we'll send you swag. We're excited to see how you analyze your own infrastructure.

CloudQuery