StreamingStaff Level

Global Video Transcoding & Streaming Pipeline

Q: What is Adaptive Bitrate Streaming (HLS/DASH)?

The video player dynamically monitors device bandwidth and switches between 1080p, 720p, or 480p chunks mid-stream without interrupting playback.

Design a YouTube / Netflix video platform with multi-bitrate HLS/DASH streaming and edge CDN delivery.

Practice in ClawPad Desktop ← All System Design Problems

Target Scale

Engineering Scale & Performance SLAs

Target production parameters expected in a senior or staff interview round.

Peak Throughput

50,000 Video Starts/sec

Sustained peak request volume during high-traffic events.

Active Users

200 Million DAU

Daily active users generating read and write operations.

Storage Ingestion

500 TB uploaded / day

Projected data ingestion and replication storage capacity.

Latency Budget

Playback start < 800ms

Strict end-to-end percentile latency SLA constraint.

Stage 01

Functional & Non-Functional Requirements

Establish clear problem boundaries before proposing architectural components.

Functional Scope

Core System Capabilities

Upload high-resolution source videos (4K/1080p).
Asynchronously transcode video into multi-bitrate chunks (HLS/DASH).
Stream video adaptively based on client network bandwidth.

Non-Functional Scope

Reliability & Latency SLAs

Smooth playback with minimal buffering under 800ms start time.
High availability and global distribution via multi-CDN edge caching.
Resilient, cost-effective cold video storage archival.

Stage 02

Capacity Estimation Math

Step-by-step arithmetic conversions for QPS, storage, and bandwidth.

Dimension	Calculation Formula	Estimated Result
Daily Ingestion Volume	500,000 hours uploaded/day × 1 GB/hour	~500 Terabytes new raw video / day
CDN Egress Bandwidth	200M DAU × 2 hours/day × 2 Mbps average bitrate	~400 Terabits/sec peak global egress

Stage 03

Multi-Tier Architecture & Component Topology

How requests navigate ingress gateways, application logic, caching, and persistence.

Ingestion Tier

Direct S3 / GCS Multipart Upload · Pre-Signed Upload Tokens

Allows client apps to stream raw video files directly to cloud object storage.

Transcoding Pipeline

Distributed Worker Cluster (FFmpeg) · DAG Task Orchestrator

Splits videos into 6-second chunks and transcodes in parallel across 1080p, 720p, 480p, and 360p.

Streaming & CDN Delivery

Cloudflare / Akamai Multi-CDN · Origin Shield Cache

Caches video segments at global POPs closest to viewers, enabling sub-second start latency.

Stage 04

Database Schemas & Partitioning Strategy

Entity models, indexing, and primary key partitioning.

Table: video_metadata

PK: video_id

video_id (UUID)
uploader_id (UUID)
title (VARCHAR)
manifest_url (VARCHAR)
status (ENUM)

Indexed by uploader_id and release timestamp.

Stage 05

Critical Architectural Trade-Offs

How to defend engineering compromises when challenged by interviewers.

Decision Point

Pre-Transcoding vs On-Demand Transcoding

Option A: Pre-transcode popular resolutions on upload

Option B: Just-In-Time (JIT) dynamic transcoding on first play

Rationale: Pre-transcode standard profiles (360p-1080p) to guarantee zero start lag for viewer playback.

Technical FAQ

Frequently Asked Questions: Global Video Transcoding & Streaming Pipeline

Key interview questions and conceptual defenses.

What is Adaptive Bitrate Streaming (HLS/DASH)?