We’re Tracking the Future of Security

Tag: Video Storage

  • Edge Recording and Video Backfill: Closing Gaps After Network Outages

    Edge Recording and Video Backfill: Closing Gaps After Network Outages

    A network video system can lose its connection to the recorder without losing the event itself. Cameras with supported edge storage can continue recording locally and later return missing footage to a video management system. That recovery process is often called video backfill, failover recording or edge-storage synchronization.

    The idea is simple, but dependable operation requires coordinated camera, storage, time, network and VMS behavior. An SD card in a camera does not automatically guarantee that footage will be discovered, transferred and presented correctly after an outage.

    What Edge Recording Protects Against

    Local recording is useful when the camera remains powered but temporarily loses its route to the central recorder. Common causes include switch failure, uplink interruption, recorder maintenance or a WAN outage at a remote site. It does not protect against every failure: loss of camera power, physical destruction, storage-card failure or theft can still remove the local evidence.

    Designers should define the failure modes they intend to cover and avoid describing edge storage as a universal backup. Power resilience and physical protection remain separate requirements.

    How Backfill Works

    During normal operation, the VMS receives live streams and records them centrally. If the connection fails, the camera records to approved local media. When communication returns, the VMS identifies the missing time range, requests the stored recording and merges or indexes it in the evidence timeline.

    ONVIF Profile G defines standardized functions for configuring, requesting and controlling recording from conformant devices. Conformance can improve interoperability, but buyers still need to verify the exact camera, firmware, storage medium and VMS combination. Optional features and vendor-specific behavior can affect the result.

    Time Synchronization Determines Evidence Quality

    Backfill depends on accurate timestamps. If a camera clock drifts during isolation, returned footage may overlap, leave apparent gaps or appear out of order. Cameras, recorders and supporting infrastructure should use a monitored time source, with alerting for excessive drift.

    Commissioning should include a deliberate network interruption long enough to create a visible recording gap, followed by reconnection and timeline review. Operators should confirm that search, export and evidentiary timestamps remain coherent.

    Control Recovery Bandwidth

    When many cameras reconnect together, simultaneous uploads can compete with live video and other operational traffic. Recovery policies should limit concurrency or bandwidth, prioritize critical cameras and consider site uplink capacity. A remote location with a narrow WAN path may require staged transfer over hours rather than immediate synchronization.

    The retention window at the edge must be long enough for the expected outage plus recovery delay. High resolution, high frame rate and complex scenes consume local storage faster, so capacity should be calculated from measured or conservatively estimated bitrates.

    Storage Media Is a Maintained Component

    Use media approved for the camera and expected write workload. Continuous video recording can wear consumer-grade cards quickly, particularly in hot enclosures. Systems should report missing, read-only or failed storage where supported, and maintenance teams should verify those alarms reach a monitored workflow.

    Encryption, card removal and evidence handling also deserve attention. Locally stored footage may contain the same sensitive material as the central recorder, so access control and disposal procedures should cover edge media.

    Commissioning and Operations Checklist

    • Confirm the supported camera, firmware, storage and VMS combination.
    • Measure usable edge retention under representative video settings.
    • Interrupt the network and verify continued local recording.
    • Reconnect and confirm backfill, timestamps, search and export.
    • Test simultaneous recovery without degrading live video.
    • Alert on storage failure, clock drift and prolonged disconnection.
    • Repeat the test after major firmware or VMS upgrades.

    Edge storage should be part of a documented continuity plan, not an unchecked feature box. NIST contingency-planning guidance similarly emphasizes defined recovery requirements, testing and maintenance. For the wider camera, VMS and storage architecture, see SectechMedia’s video surveillance systems guide.

    References

  • Video Surveillance Storage: How Much Storage Do You Really Need?

    Video Surveillance Storage: How Much Storage Do You Really Need?

    Storage is one of the largest cost components in a video surveillance system, yet it is often estimated with oversimplified assumptions. Real requirements depend on bitrate, frame rate, resolution, compression, scene complexity, recording mode, redundancy and retention policy.

    How the Technology Works

    The basic calculation is straightforward. A camera producing an average bitrate of 4 megabits per second generates roughly 43 gigabytes per day before overhead. Multiply that by the number of cameras and retention days, and storage grows quickly.

    Average bitrate is more useful than resolution alone. A 4K camera does not always use four times the storage of a lower-resolution camera because modern codecs, frame rate and scene activity have major effects. A quiet corridor may compress extremely well, while a tree-filled outdoor scene with rain and movement can require much more bandwidth.

    Operational Considerations

    Variable bitrate is common because it allocates more data to complex scenes and less to static ones. This improves efficiency but makes capacity planning dependent on realistic average and peak values. Integrators should use field measurements where possible rather than rely only on nominal manufacturer figures.

    Recording policy can reduce storage dramatically. Continuous recording is appropriate for many critical environments, but some cameras may use motion-based or event-based recording. Pre-event and post-event buffers preserve context while avoiding continuous high-bitrate storage in low-risk areas.

    Retention should be driven by operational need and regulation. Keeping every camera for 90 days because storage is available may be unnecessary. Different camera groups can have different retention periods. Critical entrances may need longer retention than low-risk internal spaces.

    Deployment and Risk

    Redundancy also consumes capacity. RAID, replication, failover recording and backup must be included in the design. Usable storage is always lower than raw disk capacity.

    Cloud storage introduces additional variables such as upload bandwidth, egress charges and subscription tiers. Hybrid systems may keep recent high-resolution video locally and archive selected evidence to the cloud.

    A good storage design includes a safety margin and monitoring. Actual bitrate should be reviewed after commissioning, and capacity alerts should warn administrators before retention drops below policy.

    Conclusion

    The objective is not to buy the largest array possible. It is to create a documented storage model that matches camera behavior, evidence requirements and resilience goals. Accurate calculation can save substantial cost while ensuring that critical video is available when an investigation begins.