Media Workflow Disaster Recovery (DR) – Replication Strategies

Need a disaster recovery strategy for your mission critical workflows?  Worried your revenue-generating productions or projects will halt because of a local outage?

Object Matrix provides solutions and consultancy for Disaster Recovery strategies within:

Click the edit button to change this text.

  • Media repositories for video on demand and content distribution platforms
  • Post-Production workflows (protecting work in progress)
  • MAMs & PAMs (Production Asset Management) including proven and trusted Avid Interplay and Cantemo based Adobe workflows

Reliable Data Replication is a Key Building Block for Disaster Recovery

Replication enables organisations to keep their objects (media assets) in MatrixStore clusters situated in geographically dispersed locations. What goes into a MatrixStore vault in one location gets replicated over to the second location. Automatically.

Ten Great Attributes to Look for in a Replication Solution

  1. Simple to set-up and use = less human or technical errors
  2. Not tied to the underlying hardware = flexibility and serviceable in the future
  3. Neither O/S specific nor data type specific
  4. Works over standard transport protocols or VPNs
  5. Secure data transmission options
  6. Scalable performance
  7. Easy to select data that needs replicating and data that doesn’t
  8. Ability to replicate deletes or not to
  9. Facilitate high availability, e.g., by allowing Reads from either the source or target of the replication
  10. Facilitate fast disaster recovery in the case that one site is lost, that work can immediately continue at the replicated site

MatrixStore supports and/or facilitates each of the above attributes.

How does MatrixStore Replication Work?

The local MatrixStore cluster knows exactly which assets have been created or updated and replicates those changes via a secured asynchronous model of transmission to a second cluster. Because the IP connection is secured it can be carried out via a public ISP or via a private connection.

Individual vaults (workspaces) within MatrixStore can be selected to be replicated. Thus, not all the data on a MatrixStore cluster needs to be replicated and the two clusters do not need to be the same size as one another.

Example 1: Basic Replication

A vault/workspace on each location can replicate to the other cluster. Straight forward, simple and automated disaster recovery.

Disaster Recovery

Example 2: Star Replication

A cluster may receive data from several other clusters so that, e.g., a central data repository could be created that receives data from many branch offices.

Example 3: Digital Library (Digital Preservation)

Optionally deletes can be replicated or not. Thus the Replica Vault at the remote location can become a growing digital media library for long-term archiving whilst the source cluster can be kept as a smaller local nearline cache.

Example 4: Universal Libraries

In this example several clusters feed into a single media library.

Other MatrixStore Replication Features: Store and Forward

When a vault/workspace is set to be “single instance” then actually two instances of an asset will be kept at the source cluster until one instance is effectively forwarded to the second cluster. Thus at no time will the user have the danger of a single instance of an asset.

The OM View: Replication is Essential to Disaster Recovery

The main reason for data replication is so that you know that you have a copy of your assets in a geographically separate location from the original location of the data. Because MatrixStore has replication built-in it is easy to use and extremely powerful. Furthermore, Object Matrix makes no extra charge for the functionality.

For more information on replication, object storage or MatrixStore vaults please get in touch and request the Object Matrix MatrixStore white paper.