Your Browser is Out of Date

Nytro.ai uses technology that works best in other browsers.
For a full experience use one of the browsers below
Computer Vision Workflow Enhancement – Image Restoration with ProHawk
Overview Solution overview Partner overview Architecture Sizing and scaling guidance Use cases
Define a balanced VM configuration for ProHawk Vision Server
Measure the improvement of CV results using ProHawk Vision Server
Milestone XProtect Smart Client Plugin functional testing
ProHawk High Availability Assessment
Conclusions and References

ProHawk High Availability Assessment

Home > Workload Solutions > Computer Vision > White Papers > Computer Vision Workflow Enhancement – Image Restoration with ProHawk > ProHawk High Availability Assessment

ProHawk High Availability Assessment

  • In many Enterprise environments it is critical to ensure that all frames of video were safely archived and available for historic analysis or evidence. It is important to assess the impact of deploying ProHawk Vision into these sort of deployments that require High Availability.

    The specific area of focus is related to the real-time restoration of streams. In this scenario, video comes from the camera and is fed in to a ProHawk Vision server and then the restored stream is passed to the Milestone XProtect VMS for playback and archival. We need to assess the impact if/when a ProHawk Vision server fails.

    Results

    The assessment can be broken in to multiple components:

    ProHawk Application HA

    When using the ProHawk Vision server for real-time restoration of video we do not natively have application HA capabilities. This can be a problem if a ProHawk Vision server fails then we lose the video stream going in to into the VMS and video is lost.

    To mitigate this, it is possible to configure a secondary raw stream from the camera to Milestone XProtect (by-passing ProHawk). The restored ProHawk stream would also flow in to into the VMS. The obvious tradeoff is that this doubles the storage requirements for cameras that require ProHawk real-time restoration but it does ensure that the original stream is captured regardless of the uptime of the ProHawk Vision server.

    Infrastructure HA

    By default, the Dell Validated Design for Computer Vision has implicit HA capabilities. Based on the VxRail system we can actively manage planned and unplanned downtime of a VM. In an outage scenario, a VM can be reinstantiated on another node in the cluster if a particular node has a problem. It is important to note that not all cases will trigger the HA process. If a ProHawk VM is operating fine but the ProHawk Vision application is not working as expected this will not automatically trigger a HA event.

    To avail of this HA capability the ProHawk Vision server also needs to be configured with NVIDIA vGPU capabilities. This allows automatic redistribution of GPU workloads across the VxRail cluster. If passthrough GPUs are used, then manual intervention is required in any outage scenario.

    Findings

    • High Availability needs to be carefully planned to ensure no data loss in the event of an issue with a ProHawk Vision server.
    • If vGPUs are used, then a level of automated HA can be relied upon. Otherwise manual intervention is required to limit the loss of video.