The attributes behind your decision
- Decision input
- Measured workload
- Primary variables
- Capacity, control, operations
- Validation
- Representative testing
- Next step
- Swiss infrastructure
Calculate audience demand from bitrate
Multiply delivered bitrate by concurrent viewing time and add protocol overhead. Model several quality levels and peak concurrency rather than monthly viewers alone.
Separate origin, transcoding and delivery
Storage, encoding and delivery produce different CPU, disk and network demand. A CDN can distribute content while the origin still needs reliable ingest and cache-fill capacity.
Plan storage and content lifecycle
Estimate source media, encoded renditions, thumbnails, logs and retention. Choose storage according to active delivery and archive requirements.
Measure quality from viewer locations
Test startup time, buffering and delivery paths from actual markets. Monitor peak port use, transfer and failed segments, then revise capacity before events.
Separate origin delivery from transcoding
An origin serves media files or segments, while transcoding creates different formats or bitrates. Those jobs have different storage, compute and network requirements. Establish which work happens on your server and which is handled by a distribution layer before selecting a plan.
Estimate peak concurrent viewers and bitrate, then total the expected viewing hours for transfer planning. Test startup time, buffering and error rates with a representative audience. Allow space for source media, generated variants and retention, and confirm what happens when an event exceeds the expected audience.