Notification to all parties in LLD - System Design Exercise

Design: Notification to All Parties System

Design covers notification sending, scheduling, retry, and status tracking. Does not cover user management or content creation for notifications.

Functional Requirements

FR1: Send notifications to multiple parties (users, admins, external systems) simultaneously

FR2: Support multiple notification channels: email, SMS, push notifications

FR3: Allow scheduling notifications for future delivery

FR4: Provide delivery status tracking for each notification

FR5: Support retry mechanism for failed notifications

FR6: Allow priority levels for notifications (high, medium, low)

FR7: Handle up to 100,000 notifications per hour

FR8: Ensure notifications are delivered within 5 seconds for real-time messages

Non-Functional Requirements

NFR1: System must be highly available with 99.9% uptime

NFR2: Latency for sending notifications should be under 5 seconds for real-time notifications

NFR3: Support horizontal scaling to handle increased load

NFR4: Ensure data privacy and security for user contact information

NFR5: Design must be modular to add new notification channels easily

Think Before You Design

Questions to Ask

❓ Question 1

❓ Question 2

❓ Question 3

❓ Question 4

❓ Question 5

❓ Question 6

❓ Question 7

Key Components

API Gateway or Notification Request Handler

Notification Scheduler

Message Queue for decoupling

Notification Dispatcher for each channel

Delivery Status Tracker

Database for storing notification metadata and status

Retry Mechanism

Monitoring and Alerting

Design Patterns

Publish-Subscribe pattern for distributing notifications

Queue-based asynchronous processing

Circuit Breaker for external channel failures

Bulkhead pattern to isolate failures

Priority Queue for handling notification priorities

Reference Architecture

  +-------------------+       +------------------+       +---------------------+
  | Notification      |       | Notification     |       | Notification        |
  | Request Handler   | ----> | Scheduler        | ----> | Message Queue       |
  +-------------------+       +------------------+       +---------------------+
                                                                |
                                                                v
  +---------------------+    +---------------------+    +---------------------+
  | Email Dispatcher    |    | SMS Dispatcher      |    | Push Notification   |
  | (SMTP or API)       |    | (SMS Gateway API)   |    | Dispatcher          |
  +---------------------+    +---------------------+    +---------------------+
            |                        |                          |
            v                        v                          v
  +---------------------+    +---------------------+    +---------------------+
  | Delivery Status     |    | Delivery Status     |    | Delivery Status     |
  | Tracker DB          |    | Tracker DB          |    | Tracker DB          |
  +---------------------+    +---------------------+    +---------------------+

  Retry Mechanism monitors failures and resubmits messages.

  Monitoring & Alerting watches system health and failures.

Components

Notification Request Handler

REST API Server (Node.js/Express or Python Flask)

Receives notification requests from clients and validates them

Notification Scheduler

Scheduler Service (Cron jobs or Quartz Scheduler)

Schedules notifications for immediate or future delivery

Message Queue

RabbitMQ or Apache Kafka

Decouples request handling from notification dispatching for scalability

Notification Dispatchers

Separate microservices for Email, SMS, Push (using SMTP, Twilio API, Firebase Cloud Messaging)

Send notifications through respective channels

Delivery Status Tracker

Relational DB (PostgreSQL) or NoSQL (MongoDB)

Stores notification metadata and delivery status updates

Retry Mechanism

Background worker or scheduled job

Retries failed notifications based on configurable policies

Monitoring and Alerting

Prometheus + Grafana or Cloud monitoring tools

Tracks system health, failures, and performance metrics

Request Flow

1. Client sends notification request to Notification Request Handler via REST API.

2. Request Handler validates and stores notification metadata in DB.

3. If scheduled for future, Notification Scheduler queues the notification at the right time.

4. Notification Scheduler pushes notification message to Message Queue.

5. Notification Dispatchers consume messages from the queue based on channel type.

6. Dispatchers send notifications via external services (SMTP, SMS gateway, push service).

7. Dispatchers update Delivery Status Tracker with success or failure.

8. Retry Mechanism monitors failures and resubmits messages to the queue.

9. Monitoring system collects metrics and alerts on failures or delays.

Database Schema

Entities: - Notification: id (PK), content, channel, priority, scheduled_time, created_at - Recipient: id (PK), notification_id (FK), party_id, contact_info - DeliveryStatus: id (PK), recipient_id (FK), status (pending, sent, failed), last_attempt_time, retry_count Relationships: - Notification 1:N Recipient - Recipient 1:1 DeliveryStatus

Scaling Discussion

Bottlenecks

Message Queue throughput limits under high load

Notification Dispatchers overwhelmed by large volume

Database write/read bottlenecks for delivery status tracking

Retry mechanism causing duplicate or delayed notifications

External channel rate limits or failures

Solutions

Partition or shard message queues; use multiple queues per channel or priority

Scale dispatchers horizontally with load balancing

Use database sharding or caching for delivery status; consider NoSQL for high write throughput

Implement idempotency and exponential backoff in retry logic

Use circuit breakers and fallback channels; monitor external service limits

Interview Tips

Time: Spend 10 minutes understanding requirements and clarifying scope, 20 minutes designing components and data flow, 10 minutes discussing scaling and trade-offs, 5 minutes summarizing.

Clarify notification channels and delivery guarantees

Explain asynchronous processing with message queues

Discuss retry and failure handling strategies

Highlight modular design for adding new channels

Address scaling challenges and solutions

Mention security and privacy considerations