Winco | Product | Promise | Solution | AV Post Product Solution

Overview

Reliable, high-performance RAID storage is at the center of digital film and video postproduction(see Figure 1). As digital video data-rates grow, the critical storage characteristics are capacity and performance/ bandwidth. Innovations delivered by the PROMISE VTrak E-Class RAID subsystem solve many of the storage challenges of postproduction (post). The capacity and performance metrics are exceptional and the modular design enables projects of all sizes.

The transition to 100% digital workflow has changed the way productions are shot, edited and delivered. Storage technology is one of the enabling factors. From the production perspective, digital cameras let directors get nearly instant feedback on set. Solidstate camera media enables non-linear access to any point in time, eliminating the need to queue up for the next shot, like you would with tape formats. This allows crews to shoot quickly with higher shooting ratios and removes time-consuming tasks like on-set film handling or rush processing. Ingesting footage into post has many fewer steps, so rough cutting can start sooner.

From the post perspective 100% digital means all of the material will be stored and transformed on disk. Post techniques of the past used expensive singlepurpose edit stations with small islands of directattached storage.These are now replaced with centralized/shared storage and consumer-off-the-shelf computers running powerful software. Centralized storage drastically improves agility; it also allows proven enterprise data management techniques to be used to assure data availability.


Capacity	Performance/Bandwidth
Requirements influence by: Video format Resolution/compression Duration/duplication VFX/compositing/audio Output format	Considerations for shared storage: Number of hosts Concurrent streams required Read/write ratio Reserved bandwidth/isolation
Figure 1: Capacity and performance/bandwidth

High quality digital video files are large (see Table 1). Video files take up a lot of space on disk and need fast networks to support multi-stream playback without dropping frames. The table below shows some popular formats that are used in film and video production. The popularity of HD and shift to digital has blurred the lines between capture resolution and release resolution, pushing more productions toward HD capture even if the project will be used for lower resolution web delivery.

Once the files are stored on disk, multiple copies are inevitably made as the files are processed. Tracking and managing the files is critical to project efficiency and crucial for larger post facilities where multiple projects use the same capacity and bandwidth resources.

Video format	Disk space for 2-hours of video	Data rate MB/second
SD – ProRes-422 HQ 486i @ 30 fps	60 GB	7.88
HD – ProRes-422 HQ 1080p @ 25 fps	170 GB	23
HD – Uncompressed 10-bit 1080i @ 30 fps	1.2 TB	166
2K – 10bit RGB/full frame DPX	2.2TB	305
4K – Uncompressed 10-bit DPX	8.7TB	1210
Table 1: Popular video codecs

VTrak E-Class Storage

Workflow environments are only as solid as the foundation they’re built upon. The VTrak E-Class RAID subsystem sets the highest standard for reliability, price, and performance. VTrak is performance tuned for the film and video industry’s best large-file sequential throughput. With VTrak E-Class you don’t have to sacrifice performance for rock solid fault tolerance and reliability.
Data Integrity Features:

Redundant battery backup unit for cache (up to 72 hours)
APC UPS support – Changes to cache to write-through during power outage
Self-healing with Predictive Data Migration (PDM) on the drives
Dynamic priority control for background tasks like rebuild or PDM

Advanced Video Codecs

New digital video formats allow images to be captured at virtually film resolutions and moved from the camera/media to protected RAID storage faster than real-time. Postproduction software can manipulate many of the files in the native camera format. This allows initial work to begin the moment the files are on disk.
By streamlining constraints at ingest, hours or even days can be cut from the postproduction schedules of the past.

FCP Native Codec Support

DVCPRO HD (P2) – Panasonic
XDCAM EX – Sony
AVC-Intra
AVCHD
RED (RED plug-in required)

Figure 2: VTrak E-Class front view. LED indicators display current system health and activity

Figure 3: VTrak E-Class rear view. All components hot-swap. LED indicators display component status

Redundant and Active-Active

VTrak E-Class has a No Single Point of Failure (NSPOF) design with redundant, active-active RAID controllers (see Figure 3). This assures access through any component failure, even an optical cable break. Active-active means both controllers work simultaneously. Under normal operation performance is optimal with this configuration. If a controller were to fail, access from the failed controller moves to the surviving controller. All files remain accessible without interruption. High availability is critical to support multi-user environments where downtime impacts the whole team.
In some video applications, maximum write performance is more important than zero-downtime. VTrak controllers can be set to operate independently by disabling cache mirroring. With this configuration the user data is securely protected by RAID as well as a battery backed cache. If a controller were to fail, targets can easily be assigned to the surviving controller since all drives are accessible to both controllers. Advanced features, like Logical Unit Number (LUN) masking, are provided at no additional charge. LUN masking allows shared storage targets and private storage targets to coexist securely across a Storage Area Network (SAN). LUN masking can also be used for smaller installations to provision dedicated storage for multiple hosts with a single VTrak E-Class subsystem.

Fibre Channel Topology

For AV applications, Fibre Channel host connectivity offers the flexibility to go from a small direct-attached solution to a Storage Area Network (SAN) using the same VTrak E-Class hardware. Fibre Channel is a mature, high-performance industry standard, used in enterprise computing. Like Ethernet networking, the greatest benefit of FC is that it can provide diverse connectivity relationships that can improve availability and can effectively scale to meet your growing needs.
The use of SAN file systems like Xsan can further expand Fibre Channel’s scaling capability, by enabling the storage to be pooled across physical hardware, providing both aggregated performance and transparent expansion to multiple SAN attached hosts simultaneously.

RAID Expansion

Capacity Planning
　

Before production begins, project planners should be able to estimate capacity requirements for all stages of the production. The use of a media asset manager like Final Cut Server™ from Apple can improve work efficiency and capacity utilization by providing revision control. Target capacity utilization should not exceed 70-80% in AV environments.

Estimation

Program duration
Capture format
Shooting ratio
Ingest ratio
Visual FX
Audio production
Music production
2D & 3D Graphics
Edit reviews & revisions
Final delivery formats

VTrak E-Class SAS expansion ports enable the user to dynamically expand the RAID subsystem with VTrak J-Class JBOD expansion chassis (see Figure 4). Expansion can be done incrementally as capacity requirements grow or upon initial deployment to support larger configurations.
For the large-file sequential performance needed for video, the 32-drive E+J configuration provides optimal performance. When additional capacity is needed deploying another E or E+J will allow for linear scaling of performance and capacity.
The VTrak E-Class architecture supports attaching up to four JBOD subsystems for capacity applications like archive and backup. But the documented multi-user AV performance 32-drive configurations should be used in AV environments.

Figure 4: VTrak E-Class expansion

Download AV Postproduction with the VTrak E-Class RAID Subsystem Guide