Optimizing Object Storage for Media & Entertainment James Norman - - PowerPoint PPT Presentation

Optimizing Object Storage for Media & Entertainment

James Norman — Storage Made Easy james@storagemadeeasy.com

Media Data Feeds

Data Feeds

Filming 8-bit RGB @ 24 FPS in UHD = 33GB per minute For a 1 hour episode 33GB X 60mins = 1.6TB per single feed With 5 cameras filming 1.6TB X 5 = 10TB of footage per episode

Simple studio example

• Perfect for handling the large volumes of data
• Can easily handle TB / PB / EB of footage
• Performance scales as cluster scales eliminating hot spots
• Perfect for unstructured data
• Performant: Large file support with DLOs, MPUs & Range Reads
• Cost Saving, Built in DR, Geo Deployment

Why do people choose Object Storage for this type of data?

• Storage needs to scale and be always be available
• Need to pull data down, or up in a quickest possible way
• Need to secure the data and enforce security policies
• Need to share the data with internal & external parties
• Access to storage is crucial to workflows, for example in DAMs or MAMs,

Genome labs, research institutes.

• Ease of Use

What do firms who require Big Data need?

File Fabric

Fabric Services: ACLs, Sharing, Policies, Locking, Search Provider Interfaces (60+) Ingest: Indexing, Discovery, Antivirus, Streaming Drives FTP SFTP REST API Plugins Desktop Mobile WebDAV Web App Legacy Applications Users – local/remote Audit Object Storage File Systems Cloud Services Identity Provider New Applications

• Web – All major browsers - no plugins required
• Desktop, presented as a Network share, or Standalone applications
• Dedicated applications for iOS and Android
• Legacy Protocol Gateway supporting FTP, SFTP and WebDav
• All of these mean it works with your existing toolset

Integrated into every point of access

Access From The Desktop

No more command line required

• Drag and Drop data movement
• Using SME’s M-Stream technology
• Transfers file in multiple parallel streams
• Can transfer TB sized objects
• Uses standard technology, like DLOs, MPUs and Range Reads
• Accelerates:
• X-Cloud transfers
• End-user to storage transfers
• Storage to end-user transfers

Data Movement

M-Stream Upload to Swift

M-Stream X-Cloud Transfer

• View Previews of files
• Media: Movies, Images
• Medical: Dicom & Genome
• Generic: Word, Excel Powerpoint and more
• No need to download entire objects to see a preview

Object Previews

Streaming Previews

• Retire those insecure FTP servers currently in use
• You can’t track access
• You can’t expire access to files
• You can’t individually protect files
• Often using shared credentials
• Replace with a secure file sharing platform like the File Fabric
• All access is logged
• Inbound and outbound sharing
• Individually protect files with Passwords, Time-based Expiry and Limits on

access

• Real time reporting of access

Secure Content Distribution

Secure Content Distribution

Person 1: Creating a shared link

Secure Content Distribution

Person 2: Accessing the shared link

• All of these features + many more, whilst remaining storage

agnostic

• The File Fabric does not write data in any proprietary formats,

and uses storage provided technology to deliver its features like DLOs, SLOs, Range Reads and Versioning.

• No lock-in with the File Fabric.

Storage agnostic & nothing proprietary

• Integrate the File Fabric into Intelligence platforms like Google

Vision and Amazon Media Services & Elemental

• Automatic tagging & classifications (labels, faces, etc)
• Automatic transcription
• All searchable at your fingertips

Intelligence

• Storage needs to scale and be always be available
• Need to pull data down, or up in a quickest possible way
• Need to secure the data and enforce security policies
• Need to share the data with internal & external parties
• Access to storage is crucial to workflows, for example in DAMs or MAMs
• Ease of Use

Revisit: What do firms who require Big Data need?

Contact Me: james@storagemadeeasy.com

StorageMadeEasy.com

Contact Sales: sales@storagemadeeasy.com