System Design: Designing an Ad Click Aggregator

Ad click aggregation is a massive scale data problem. When billions of users click on ads across the web, those clicks must be aggregated, deduplicated, and stored for both real-time analytics (advertiser dashboards) and accurate billing.

1. Core Requirements

graph LR
    Producer[Producer Service] -->|Publish Event| Kafka[Kafka / Event Bus]
    Kafka -->|Consume| Consumer1[Consumer Group A]
    Kafka -->|Consume| Consumer2[Consumer Group B]
    Consumer1 --> DB1[(Primary DB)]
    Consumer2 --> Cache[(Redis)]

• High Throughput: Handling billions of clicks per day (tens of thousands per second).
• Accuracy: Billing requires exactly-once processing. We cannot charge an advertiser twice for the same click (Deduplication).
• Latency: Real-time dashboards should update within seconds.
• Resilience: No click should ever be lost, even if a data center goes down.

2. The Data Path

1. Click Event: A user clicks an ad. The browser sends a request to our Click Tracking Server.
2. Raw Log Ingestion: The Tracking Server immediately pushes the raw click event into Apache Kafka.
   • Why Kafka? It acts as a high-performance buffer and persistent log.
3. Aggregation Engine:Apache Flink or Spark Streaming consumes from Kafka.
   • Deduplication: Uses a Redis cache or a stateful Flink map to filter out duplicate clicks (based on click_id and user_id).
   • Windowing: Clicks are aggregated in 1-minute tumbling windows.
4. Storage:
   • Real-time: Aggregated counts are stored in Cassandra or Redis for the advertiser dashboard.
   • Historical: Raw clicks are stored in Amazon S3 (Parquet) for long-term auditing and fraud detection.

3. Dealing with "Exactly-Once" Semantics

Billing systems cannot tolerate duplicates.

• Kafka Idempotency: Producers are configured with enable.idempotence=true.
• Checkpointing: Flink uses distributed snapshots (checkpoints) to ensure that if a worker fails, it resumes from the exact point in the log where it left off, ensuring no event is processed twice or missed.

4. Scaling the Write Volume

The biggest bottleneck is the write volume to the database.

• Pre-aggregation: Never write every single click to the database. Aggregate them in RAM (in Flink) and write only the summary (e.g., "Ad 123 got 500 clicks in the last minute") once to the database.
• Sharding: Shard the database by ad_id to distribute the aggregation load across multiple nodes.

5. Fraud Detection

Ad fraud (bots clicking ads) is a major concern.

• Real-time Filter: Use ML models or rule-based filters (e.g., "more than 10 clicks from same IP in 1 second") to flag and filter fraudulent clicks before they reach the billing layer.

Summary

The engineering of an ad click aggregator is a battle of Write Throughput and Data Integrity. By using Kafka for ingestion and a stateful stream processor like Flink for pre-aggregation and deduplication, you can build a system that processes billions of events with perfect accuracy and sub-second latency.

Engineering Standard: The "Staff" Perspective

In high-throughput distributed systems, the code we write is often the easiest part. The difficulty lies in how that code interacts with other components in the stack.

1. Data Integrity and The "P" in CAP

Whenever you are dealing with state (Databases, Caches, or In-memory stores), you must account for Network Partitions. In a standard Java microservice, we often choose Availability (AP) by using Eventual Consistency patterns. However, for financial ledgers, we must enforce Strong Consistency (CP), which usually involves distributed locks (Redis Redlock or Zookeeper) or a strictly linearizable sequence.

2. The Observability Pillar

Writing logic without observability is like flying a plane without a dashboard. Every production service must implement:

• Tracing (OpenTelemetry): Track a single request across 50 microservices.
• Metrics (Prometheus): Monitor Heap usage, Thread saturation, and P99 latencies.
• Structured Logging (ELK/Splunk): Never log raw strings; use JSON so you can query logs like a database.

3. Production Incident Prevention

To survive a 3:00 AM incident, we use:

• Circuit Breakers: Stop the bleeding if a downstream service is down.
• Bulkheads: Isolate thread pools so one failing endpoint doesn't crash the entire app.
• Retries with Exponential Backoff: Avoid the "Thundering Herd" problem when a service comes back online.

Critical Interview Nuance

When an interviewer asks you about this topic, don't just explain the code. Explain the Trade-offs. A Staff Engineer is someone who knows that every architectural decision is a choice between two "bad" outcomes. You are picking the one that aligns with the business goal.

Performance Checklist for High-Load Systems:

1. Minimize Object Creation: Use primitive arrays and reusable buffers.
2. Batching: Group 1,000 small writes into 1 large batch to save I/O cycles.
3. Async Processing: If the user doesn't need the result immediately, move it to a Message Queue (Kafka/SQS).

Advanced Architectural Blueprint: The Staff Perspective

In modern high-scale engineering, the primary differentiator between a Senior and a Staff Engineer is the ability to see beyond the local code and understand the Global System Impact. This section provides the exhaustive architectural context required to operate this component at a "MANG" (Meta, Amazon, Netflix, Google) scale.

1. High-Availability and Disaster Recovery (DR)

Every component in a production system must be designed for failure. If this component resides in a single availability zone, it is a liability.

• Multi-Region Active-Active: To achieve "Five Nines" (99.999%) availability, we replicate state across geographical regions using asynchronous replication or global consensus (Paxos/Raft).
• Chaos Engineering: We regularly inject "latency spikes" and "node kills" using tools like Chaos Mesh to ensure the system gracefully degrades without a total outage.

2. The Data Integrity Pillar (Consistency Models)

When managing state, we must choose our position on the CAP theorem spectrum.

3. Observability and "Day 2" Operations

Writing the code is only 10% of the lifecycle. The remaining 90% is spent monitoring and maintaining it.

• Tracing (OpenTelemetry): We use distributed tracing to map the request flow. This is critical when a P99 latency spike occurs in a mesh of 100+ microservices.
• Structured Logging: We avoid unstructured text. Every log line is a JSON object containing correlationId, tenantId, and latencyMs.
• Custom Metrics: We export business-level metrics (e.g., "Orders processed per second") to Prometheus to set up intelligent alerting with PagerDuty.

4. Production Readiness Checklist for Staff Engineers

• Capacity Planning: Have we performed load testing to find the "Breaking Point" of the service?
• Security Hardening: Is all communication encrypted using mTLS (Mutual TLS)?
• Backpressure Propagation: Does the service correctly return HTTP 429 or 503 when its internal thread pools are saturated?
• Idempotency: Can the same request be retried 10 times without side effects? (Critical for Payment systems).

Critical Interview Reflection

When an interviewer asks "How would you improve this?", they are looking for your ability to identify Bottlenecks. Focus on the network I/O, the database locking strategy, or the memory allocation patterns of the JVM. Explain the trade-offs between "Throughput" and "Latency." A Staff Engineer knows that you can never have both at their theoretical maximums.

Optimization Summary:

1. Reduce Context Switching: Use non-blocking I/O (Netty/Project Loom).
2. Minimize GC Pressure: Prefer primitive specialized collections over standard Generics.
3. Data Sharding: Use Consistent Hashing to avoid "Hot Shards."

Technical Trade-offs: Messaging Systems

Key Takeaways

• High Throughput: Handling billions of clicks per day (tens of thousands per second).
• Accuracy: Billing requires exactly-once processing. We cannot charge an advertiser twice for the same click (Deduplication).
• Latency: Real-time dashboards should update within seconds.

Read Next

Mental Model

Connecting isolated components into a resilient, scalable, and observable distributed web.

• Speculative Retries: The Google Approach to Solving Tail Latency
• Building Production Observability with OpenTelemetry and Grafana Stack
• Introduction to High-Level Design

Verbal Interview Script

Interviewer: "How would you ensure high availability and fault tolerance for this specific architecture?"

Candidate: "To achieve 'Five Nines' (99.999%) availability, we must eliminate all Single Points of Failure (SPOF). I would deploy the API Gateway and stateless microservices across multiple Availability Zones (AZs) behind an active-active load balancer. For the data layer, I would use asynchronous replication to a read-replica in a different region for disaster recovery. Furthermore, it's not enough to just deploy redundantly; we must protect the system from cascading failures. I would implement strict timeouts, retry mechanisms with exponential backoff and jitter, and Circuit Breakers (using a library like Resilience4j) on all synchronous network calls between microservices."