Media Stack High Availability
For the Storage and Delivery of Media
Applicability
Executive Summary
Media High Availability (HA) refers to uninterrupted access to video for both live monitoring and playback, and all associated metadata, delivered through the Recording Service and its Media Database—the Media Stack. This article compares three architectural approaches to Media HA—Active-Passive, Active-Semi-Active, and Active-Active—using four evaluation dimensions: failover, consistency, cybersecurity, and efficiency.
Active-Passive Media Stacks rely on standby nodes that take over when the primary fails. Solutions such as Milestone Recorder Failover and Clustering with Media Sync provide basic stack-level failover but not stream-level continuity. They introduce gaps during failover, limited access to pre-failover media, and, in the case of media sync, a risk that ransomware or corruption spreads across nodes. Efficiency is strictly one-to-one, requiring a dedicated standby for each primary.
Active-Semiactive Media Stacks run a clone recorder in parallel, capturing media directly from the source but detached from client services until the primary fails. This improves consistency—no media is lost during failover—but introduces risks, because synchronization of missing media and configuration can propagate cyber threats between nodes. Like Active-Passive, efficiency remains one-to-one.
Active-Active Media Stacks operate multiple independent stacks simultaneously, allowing clients to select the best source dynamically. This allows stream-level failover, load balancing, geo-aware access, and specialized roles across stacks. Consistency is preserved via direct source recording and edge restoration. Cybersecurity is strengthened because corruption in one stack does not affect the other, and separate deployments further contain risk. Efficiency is many-to-many, enabling fewer standby nodes to cover multiple primary failures.
Our Take
While Active-Passive and Semi-Active models offer familiar mechanisms, they are hindered by failover delays and/or, limited resilience against cyber threats, and poor efficiency. Active-Active architectures demand smarter client logic but deliver the strongest availability, cyber resilience, and cost efficiency, making them the most future-proof choice for video management systems such as Milestone XProtect.
Read the subsequent sections for a deep dive into these techniques.
Last updated