Concert Ticket Marketplace on Ethereum Layer-2

A fully decentralized ticket marketplace where concert organizers mint NFT tickets on Polygon zkEVM — no middlemen, no fakes, no scalping.

Stack

Solidity · Next.js · Hardhat · ethers.js · Polygon zkEVM · IPFS · MetaMask · ERC-721

What It Does

Three core smart contract functions drive the whole lifecycle:

• createEvent — organizer submits event details (name, date, capacity, price); contract stores it immutably and returns a unique event ID
• purchaseTicket — buyer sends ETH, contract atomically mints an ERC-721 NFT to their wallet; metadata stored on IPFS
• useTicket — event staff enters the Token ID; contract verifies ownership, marks ticket as used, then burns the NFT to prevent reuse

Everything is on-chain. No backend, no trusted third party.

Why Polygon zkEVM

Ethereum mainnet is too expensive for high-volume ticket sales. Polygon zkEVM solves this with ZK Rollups — transactions execute off-chain, a single ZK-SNARK proof is submitted to Ethereum L1 for verification. The result: same security guarantees, drastically lower cost.

Tested across three testnets to find the right tradeoff:

Cardona wins on cost (~375× cheaper than Sepolia per operation). Amoy wins on speed. For a marketplace that needs to scale, cost efficiency matters more. Arch overview

Performance Testing

Ran four load test scenarios on Cardona. Key findings:

Ramp-Up Purchase Test (2→10 concurrent transactions)

• 100% success rate across all 150 transactions
• TPS scaled linearly from 0.99 → 4.65 as concurrency increased
• Average response time stayed under 50ms even at peak load
• Peak burst TPS: 285.7

Sustained Load Test (8 concurrent, 80 transactions, 20s)

• Still 100% success rate
• The network got faster over time — TPS improved from 175 → 292 as it warmed up
• Final batch: 25ms duration, 23ms response time
• 66.8% TPS improvement from first batch to last

The warm-up effect was consistent across all tests: first-batch gas usage runs ~20% higher (cold start overhead), then stabilizes at ~332,619 gas for subsequent purchases.

What I Learned

• ZK-SNARK verification adds compute overhead that explains Cardona's slower confirmation time — it's the price of cryptographic security without a trusted party
• The first NFT mint always costs more gas than subsequent ones; state initialization is expensive, subsequent operations reuse cached state
• Polygon zkEVM's self-optimizing behavior under sustained load was surprising — it wasn't designed behavior I explicitly coded, it's the network adapting
• Building on L2 forces you to think about finality differently: zkEVM gives instant finality once the proof is verified, vs. optimistic rollups that need a challenge window
• IPFS for metadata + on-chain ownership is the right split — heavy data off-chain, proof of ownership immutably on-chain

Results

• ✅ Full ticket lifecycle working end-to-end on Polygon zkEVM Cardona testnet
• ✅ 100% transaction success rate under all load scenarios
• ✅ Ticket purchase cost: 0.00000393 ETH (~$0.01) vs. comparable L1 costs orders of magnitude higher
• ✅ Peak throughput: 320 TPS under sustained load
• ✅ Zero double-spend — burn mechanism prevents any ticket from being used twice

UI overview