Shared capacity for eligible servers
Bandwidth Pool
Share contracted network capacity across eligible servers instead of sizing each one for a peak that happens at a different time. Request a pool using the port and peak traffic of each member.
Workload illustration, not live server measurements.
Fleet configuration
Choose the shared capacity for the whole fleet, then specify the ports on its members. A server cannot draw more than its own port allows, even when the pool has room.
We confirm hardware, eligible networks, billing and availability in your quote.
Price on quote
[email protected]Add demand across eligible servers at the same point in time. Each member remains limited by its own port.
Unused capacity can help another member during a spike, up to the pool and port limits.
Confirm the metering interval, eligible traffic and treatment of overages before adding a node.
Alerting, upgrade path, overage handling, and eligible server rules are confirmed in the written quote. Check whether the agreement measures a traffic volume or a rate, and over what interval.
Use cases
A shared Gbps limit controls simultaneous traffic; a shared TB allowance counts data over time. For a throughput pool, compare aligned samples from every member. See a two-server pooling example.
Workload: Several servers serving different channels, regions, or customer groups.
Sizing: Traffic may be uneven: one node spikes while another is quiet.
Fit: A pool prevents unused capacity from being trapped in the wrong place commercially.
Limit: Physical port limits still apply on each server.
Workload: Cache nodes added as audience clusters grow.
Sizing: Traffic shape changes by region, cache-hit rate, and event schedule.
Fit: Pooling gives simpler fleet billing while the architecture changes.
Limit: Network locations and test endpoints should be confirmed before launch planning.
Workload: Live sessions spike across channels and relays during prime time.
Sizing: Bitrate, concurrency, and middleware overhead determine actual capacity.
Fit: Pooling can reduce commercial waste when peaks move between nodes.
Limit: Rights to distribute content remain the customer's responsibility.
Provisioning
Pooling is a network and billing arrangement, not an operating system. Choose an OS for each member server and use the same reporting interval across the fleet.
Check installation, driver and licensing requirementsQuestions & answers
It helps when demand shifts between eligible servers. Add their egress at the same timestamps; summing each node’s separate daily maximum can overstate the capacity the fleet needs.
No. Available pool capacity does not raise a server’s configured port speed. Each node is still limited by its port, hardware and network path.
Confirm the eligible servers and networks in the pool order. Membership, traffic directions and billing rules are contract details, not properties implied by the network names.
Model traffic after one node is removed. The surviving nodes need enough individual port capacity as well as enough aggregate pool capacity to carry the redistributed load.
Request the committed capacity, measurement interval, treatment of excess usage and process for adding or removing nodes. Specify whether replication between servers counts toward the pool.
Related paths
Compare adjacent bandwidth, media, and pooled-capacity routes before choosing a configuration.
Debian stable
Ubuntu Server LTS
AlmaLinux
Rocky Linux
Fedora Server
Windows Server