Active-Active Media Stack
Stream-level failover with strong isolation and lower cost
Executive Summary
Active–active media stacks operate independently, with each recorder ingesting streams directly from cameras and serving clients simultaneously. Clients are redundancy-aware and can dynamically select stacks, enabling stream-level failover, load balancing, role-based access, geo-aware distribution, and application specialization. Because each stack records from source and restores gaps through edge recording, consistency is maintained without inter-stack replication. Cybersecurity is stronger than in passive models, since corruption in one stack does not propagate to others, and independent VMS deployments further isolate risk. Efficiency is achieved through a many-to-many approach, allowing fewer standby nodes to support multiple primaries, thereby reducing infrastructure costs while delivering the highest levels of availability and cyber resilience.
Failover
Clients are redundancy-aware and can dynamically select one or more stacks to serve their requests. For example, a client may source some streams from one stack and simultaneously source other streams from the second.
This enables:
Stream Level Failover
Live and Playback content unavailable on one stack are sourced from another.
Role-based Access
Different user groups can be directed to different stacks—for example, security teams may use Stack A for real-time monitoring, while compliance teams access Stack B for archived content. This separation minimizes contention and balances load between stacks.
Geo-Proximity Choice
European clients connect to the EU stack, while U.S. users connect to the U.S. stack to reduce latency.
Hybrid Media Access
A client may source low-latency live streams from Stack A while fetching high-resolution playback from Stack B.
Application Specialization
Stack A may serve Milestone Smart Clients for daily monitoring, while Stack B supports analytics or PSIM integrations.
Consistency
Since each stack directly ingests media from cameras, no content is lost during failover. Gaps are filled through edge recording restoration at the source, maintaining consistency without creating replication pathways that could spread corruption. Optional copying of missing media between stacks is possible, though it reduces cybersecurity.
Cybersecurity
Recording directly from cameras ensures that attacks remain contained to the affected node—corruption or ransomware on one stack does not impact others. Copying of missing media between stacks is possible and if deployed, reduces cybersecurity.
Media stacks can be deployed as parts of independent XProtect instances. Here, isolation is even stronger, further reducing the blast radius of a compromise.
Efficiency
Active–active systems achieve M:N efficiency: N standby stacks can cover M simultaneous primary failures, with N < M. This flexibility lowers infrastructure cost compared to one-to-one standby designs, while delivering superior availability and resilience.
Last updated