Market Data Replication System Development on Kafka

Your trading bot runs on servers close to the exchange. Risk management needs the same data in a data center across the ocean. Simple file transfer doesn't work: latency grows to seconds, and data loss reaches 8% of ticks over an unstable channel. We encountered a case where a client lost up to 5% of ticks due to an unstable link between New York and Tokyo. You need distributed real-time replication that can handle hundreds of thousands of messages per second without losing a single tick. Replication based on Apache Kafka reduces losses to zero and ensures consistency across all nodes. Order replication that won't fail.

We design and deploy market data replication systems turnkey. Our stack is Apache Kafka as a reliable backbone, MirrorMaker 2 for cross-regional replication, and Confluent Schema Registry for format evolution. Over the past 10+ years, we have delivered over 30 projects for crypto funds, market makers, and prop trading firms. Result: reduce infrastructure costs by up to 40% and cut downtime by 80%. Clients save an average of $15,000 per month on infrastructure by consolidating streams.

In this article, we'll break down how to build a replication system, which topologies to choose, how to guarantee consistency, and how to avoid typical pitfalls.

Why Replication Is Needed

A trading system consists of several components running in different environments:

• Production trading — co-location near the exchange, minimal latency.
• Research/backtesting — data center with large storage volumes.
• Risk management — isolated network with restricted access.
• Analytics dashboards — accessible to a wide audience.

Each environment must receive an identical data stream without overloading the source. Replication solves this by ensuring consistency, fault tolerance, and scalability.

Replication Topologies

For production, we recommend Pub-Sub via Kafka — it's a flexible option that scales easily. Apache Kafka provides durable, scalable, and fault-tolerant message exchange.

How to Ensure Consistency During Replication

Consistency is the key challenge in distributed replication. We solve it with a combination of idempotent consumers and atomic transactions. Consumers deduplicate messages by keys such as trade_id or update_id. For critical streams (risk management), we use exactly-once delivery via Kafka Transactions.

from confluent_kafka import Producer

producer = Producer({
    'bootstrap.servers': 'kafka:9092',
    'enable.idempotence': True,
    'transactional.id': 'market-data-producer-1',
    'acks': 'all'
})

producer.init_transactions()

def publish_trade_batch(trades: list[Trade]):
    producer.begin_transaction()
    try:
        for trade in trades:
            producer.produce(
                topic=f'market.trades.{trade.exchange}.{trade.symbol}',
                key=trade.symbol.encode(),
                value=serialize(trade)
            )
        producer.commit_transaction()
    except Exception as e:
        producer.abort_transaction()
        raise

Why Kafka Is the Standard for Market Data Replication

Apache Kafka provides all necessary properties: durability (data stored on disk), scalability (horizontal partitioning), and independence of consumer groups. We configure topics with names like {data_type}.{exchange}.{symbol}.{interval}, making it easy to filter data.

Topic: market.trades.binance.BTCUSDT
  Partition 0: trades (all, ordered by time)

Topic: market.orderbook.binance.BTCUSDT
  Partition 0: snapshots + diffs (ordered by update_id)

Topic: market.candles.binance.BTCUSDT.1m
  Partition 0: 1-minute OHLCV (ordered by candle time)

Delivery Guarantees and Cross-Datacenter Replication

In market data systems, at-least-once delivery is most common: better to get a duplicate than lose data. Consumers are idempotent — deduplication by trade_id or update_id. For risk management and position accounting, we enable exactly-once via Kafka Transactions.

Kafka MirrorMaker 2 replicates topics between clusters. Example MirrorMaker 2 configuration:

# mirrormaker2.properties
clusters = us-east, eu-west
us-east.bootstrap.servers = kafka-us:9092
eu-west.bootstrap.servers = kafka-eu:9092

us-east->eu-west.enabled = true
us-east->eu-west.topics = market\.*
us-east->eu-west.replication.factor = 2

The EU cluster receives a replica of all market.* topics with a delay of 50–200 ms for transatlantic replication. This is sufficient for most analytics and risk systems.

Retention Management and Monitoring

Market data accumulates quickly. Retention policies:

• For tick data: 7 days, then delete.
• For daily OHLCV: infinite, with a size limit of 10 GB per partition.
• For order book: log compaction — keep only the latest state per price level.

zstd compression reduces data by 40–70% without noticeable CPU load.

Key monitoring metrics:

Consumer lag > 5 minutes for a trading bot is a critical alert. For an analytics system, it's a warning.

Schema Registry and Format Compatibility

To avoid breaking consumers when the schema evolves, we use Confluent Schema Registry and Avro. New optional fields with default null are backward-compatible changes.

{
  "type": "record",
  "name": "Trade",
  "namespace": "com.exchange.market",
  "fields": [
    {"name": "exchange", "type": "string"},
    {"name": "symbol", "type": "string"},
    {"name": "timestamp", "type": "long"},
    {"name": "price", "type": {"type": "bytes", "logicalType": "decimal", "precision": 24, "scale": 8}},
    {"name": "quantity", "type": {"type": "bytes", "logicalType": "decimal", "precision": 24, "scale": 8}},
    {"name": "side", "type": {"type": "enum", "name": "Side", "symbols": ["BUY", "SELL"]}},
    {"name": "is_maker", "type": ["null", "boolean"], "default": null}
  ]
}

What's Included in the Work

• Architectural documentation: topology description, data flow diagram, topic specification.
• Infrastructure access: setup of Kafka clusters, MirrorMaker, Schema Registry.
• Team training: workshop on operations and monitoring.
• Post-launch support: assistance during the first 2 weeks of production operation.

How We Deploy Replication: Step-by-Step Plan

1. Requirements analysis: data volume (up to 100,000 messages/s), latency, delivery guarantees, number of data centers.
2. Topology design: choose between Hub-and-Spoke, Chain, or Pub-Sub.
3. Deploy Kafka cluster (3 to 7 brokers) with MirrorMaker 2 and Schema Registry.
4. Configure topics and retention policies.
5. Integrate consumers with idempotency and deduplication.
6. Set up monitoring (consumer lag, latency) and alerts.
7. Test with synthetic data and real loads.
8. Document and hand over to the client's team.

During the process, we verify: producer idempotency, transaction settings, retention consistency across clusters, deduplication functionality, and absence of duplicates.

Timelines and Cost

A basic implementation takes 7 to 14 days: analytics, cluster deployment, MirrorMaker 2 and Schema Registry setup, monitoring. A full solution integrated into your infrastructure takes up to 4 weeks. Cost is calculated individually: depends on data volume, number of data centers, and required delivery guarantees. Contact us to evaluate your project and get a consultation. Order the development of a replication system for your tasks — and we will ensure reliable data delivery.

Why exchange development requires deep domain expertise

We develop exchanges — not 'chart sites,' but matching engines that process thousands of orders per second without delay, route liquidity between pools, and guarantee that no user gains access to others' funds. Teams that start with the UI and postpone the engine 'for later' end up rewriting everything in six months in 90% of cases.

Order Book vs AMM: where most projects break

Centralized exchanges (CEX) are built around an order book + matching engine. Decentralized exchanges (DEX) either also use an order book (dYdX on StarkEx, Serum/OpenBook on Solana) or an AMM with concentrated liquidity (Uniswap v3/v4, Curve, Balancer). A classic mistake when developing a CEX is implementing the matching engine on top of a relational database with transactions for each match. PostgreSQL handles ~500 RPS without special effort, but at peak loads of 5,000–10,000 orders per second, it turns into a deadlock nightmare. The correct architecture: in-memory order book (Redis Sorted Sets or custom C++/Rust structure), asynchronous writing of matches to PostgreSQL via a queue (Kafka/RabbitMQ), and a separate settlement service that finally updates balances.

For DEX, the most painful problem is sandwich attacks and MEV. A pool with a plain xy=k AMM without slippage protection becomes a target for MEV bots within hours of launch. Uniswap v2 lost hundreds of millions of dollars in user liquidity. Solutions: integration with Flashbots Protect, a commit-reveal scheme for orders, or switching to TWAMM (Time-Weighted AMM) for large trades.

Concentrated liquidity and impermanent loss

Uniswap v3 introduced concentrated liquidity – LPs choose a price range in which to provide liquidity. Capital efficiency increased 4,000x compared to v2 for stable pairs. But implementing this mechanism correctly is non-trivial. The Uniswap v3 liquidity contract uses tick-based accounting: the price space is divided into discrete ticks (tick = log₁.0001(price)), each tick stores accumulated fee growth and liquidity delta. When creating a position, the lower and upper ticks are computed, and the contract recalculates all active positions at each swap. Storage layout is critical here – incorrect variable packing in slots easily adds 40–60% to swap gas cost.

We implemented a Uniswap v3 fork for a client on Polygon with a custom fee tier system. The initial version consumed 180k gas for a swap across 2 ticks. After slot packing of variables in Tick.Info and inlining several internal calls, it dropped to 112k gas. This reduced gas costs by 38% and saved the client substantial costs on fees monthly. The techniques applied are described in the Uniswap v3 Whitepaper and confirmed by our audit experience.

How a matching engine delivers performance

A production-ready matching engine is built according to the following scheme:

• Order ingestion layer – WebSocket gateway (Go or Rust), accepts orders, validates signature, checks balance via Redis, queues them. Latency at this level must be <1ms.
• Matching core – single-threaded event loop (eliminates race conditions without mutexes). In memory, we hold two Sorted Sets for each trading instrument: bids and asks. FIFO matching for limit orders, immediate-or-cancel for market orders. Throughput with a proper Rust implementation – 500k–1M matches per second on a single core.
• Settlement service – reads matches from Kafka, atomically updates balances in PostgreSQL (UPDATE accounts SET balance = balance - $1 WHERE id = $2 AND balance >= $1). Optimistic locking via row versioning.
• Withdrawal pipeline – separate service with cold/hot wallet architecture. The hot wallet holds 5–10% of total deposits, the rest is cold storage with multi-sig (Gnosis Safe or custom HSM). Automatic withdrawals only from hot wallet, large amounts require manual authorization.

How our exchange development process ensures reliability

Smart contracts and gas optimization

For EVM-based DEX (Ethereum, Arbitrum, Optimism, Polygon), the entire critical path lives in Solidity. Main contracts: Pool, Factory, Router, PositionManager (for v3-like), and Quoter for off-chain calculations. Typical mistakes we see in audits:

Reentrancy via callback. Uniswap v3 uses flash swap with a callback (uniswapV3SwapCallback). If your router lacks a nonReentrant guard and you don't check msg.sender == pool, the contract gets drained via a nested call. This is not hypothetical – several v3 forks lost funds this way.

Oracle manipulation in AMM. If your contract uses the spot price from the pool for collateral calculation, it is front-runnable. Correct: TWAP over 30+ minutes (Uniswap v3 OracleLib) or an external oracle (Chainlink).

Unbounded loops in liquidity range. If a swap crosses many ticks in a row (price impact 80%+), gas may exceed the block limit. Need MAX_TICKS_CROSSED with partial fill and returning the remainder.

For Solana DEX (Anchor framework, Rust), the architecture is fundamentally different: account-based model, Program Derived Addresses (PDA) instead of storage, Cross-Program Invocations instead of internal calls. Solana's throughput (~3,000–4,000 TPS vs 15–30 on Ethereum mainnet) allows building on-chain order books – exactly what Phoenix DEX does.

Liquidity bootstrapping and aggregator integration

Launching a pool is not enough – you need to ensure liquidity at launch. Practical mechanisms:

• Liquidity Bootstrapping Pool (LBP) – initial price is high, asset weights dynamically shift, creating selling pressure and even token distribution. Implemented in Balancer v2.
• Initial Liquidity Offering via Uniswap v3 – adding liquidity in a narrow range around the initial price, then gradually expanding as volume grows. Requires active liquidity management or integration with Arrakis/Gamma.
• Integration with 1inch, Paraswap, Li.Fi – aggregators bring traffic but require standard compliance: the pool must have correct getAmountsOut, support ERC-20 approval/permit, and not have custom transfer hooks that break the aggregator's routing.

Development process and deliverables

Analytics and design begin with choosing the architectural model: CEX with custodial storage, non-custodial DEX, or hybrid (off-chain order book + on-chain settlement, like dYdX v3). This decision determines everything – regulatory load, tech stack, team.

Development proceeds in layers: first smart contracts with full Foundry coverage (fuzzing, invariant testing), then backend services, then integration layer, and finally frontend. Testing includes fork testing on mainnet via Foundry – we reproduce real liquidity conditions, not synthetic ones.

Audit is mandatory before mainnet deployment. For DEX contracts, minimally one firm with manual review (Trail of Bits, Spearbit, Code4rena contest). For CEX custody, audit of key storage processes. We guarantee all contracts undergo formal verification and fuzzing testing (Echidna, Foundry invariant).

Estimated timelines

Cost is calculated individually after a technical briefing: chain selection, throughput requirements, custodial model. Our certified engineers with 10+ years of experience will help you choose the optimal architecture and avoid common pitfalls. Contact our team for a detailed proposal.

Pitfalls to avoid at launch

• Forgetting the price oracle in AMM. Spot price can be manipulated with a flash loan in one transaction. If your lending protocol uses the spot price from its own pool, that's a bug.
• Hot wallet without limits. A CEX without daily limits on automatic withdrawals is an invitation for attackers. Compromising one key should lose at most 10% of total funds.
• Absence of circuit breaker. A 40% price drop in 5 minutes should halt automatic liquidations or withdrawals until manual review. Without this, a cascading liquidation spiral destroys all TVL.
• Incorrect decimal handling. USDC uses 6 decimals, WBTC – 8, most tokens – 18. Mixing without normalization leads to either precision loss or overflow. Solidity has no float; we work with fixed-point using FullMath (mulDiv with overflow protection).

Want to avoid these problems? Get a consultation — we will select the architecture for your project and provide exact timelines. Order exchange development with quality guarantee and ongoing support.