Introducing Global Snapshot Clones
In VAST 4.6 we are introducing a new capability to support multi-site workflows that we call Global Snapshot Clones. Until now, our immutable VAST Snapshots have provided ultra-efficient, point-in-time read-only views of data available on the local cluster or on a replication target. Now, Global Snapshot Clones allow users to access a VAST Snapshot from a remote VAST cluster as an instantly writeable clone.
We’ve enabled Global Snapshot Clones by building a browsable catalog of snapshots that allows any participating VAST clusters to subscribe to and receive immediately consistent, full read-write access to a clone mounted to a local path. As soon as the clone is initialized it’s immediately available on the target – prioritized data blocks are intelligently streamed from the source, enabling writes to the cloned data while synchronization is ongoing.
What’s this look like in action? Here are some examples:
Multi-region service provider: An application or user can be moved to a different region and its data will be immediately available for full read / write access locally in the new region. This mobility allows service providers to optimize their resources, transparently moving apps and users along with their data to the ideal data center. A Global Snapshot Clone is created in the target region, and the application has a fully consistent and writeable view of its data while 100% of the data is copied to the destination.
Media & Entertainment: A video editor making a highlight film needs access to a large collection of videos from recent sporting events but will only use a subset of the assets. A Global Snapshot Clone presents a seamless view of 100% of the assets while only moving the required subset of the data and only when requested by the editor.
Ransomware recovery: Clones can also be made locally within a single VAST cluster. In the event of a ransomware attack, a Global Snapshot Clone is made to a local target, providing instant access to the data set on the same VAST cluster and enabling the business to resume.
All the examples provide a consistent view of a point-in-time copy of the required data and a seamless experience for the end user. Global Snapshot Clones complement our cluster-to-cluster replication capabilities with more options for how data movement is initiated and how much of data is moved to support a wider set of use cases.
How do Global Snapshot Clones work?
Symmetrical Access and Performance with Asymmetric Capacity Utilization
In addition to instant accessibility and usability of your globally shared datasets, one of the most exciting aspects of Global Snapshot Clones is the flexibility it offers in managing different types of data access needs while breaking the historic consistency and capacity tradeoffs required to give local users high-performance access to globally sourced data.
Global Snapshot Clones allow organizations to choose how and when to move the data to match the use case. When a snapshot clone is created, the Background Sync option controls how and when the data is copied to the target clone.
A snapshot clone of research data is created with Background Sync enabled
Background Sync On is ideal when moving or distributing data for well understood workloads – for example, on-site users relocating to a new campus. “Background sync = on” begins replicating 100% of the user’s data as soon as the clone is initiated. If the users begin to access the data at the new location before the sync is complete, the requested data blocks are prioritized to ensure the end user has seamless access just as if all data was local.
Background Sync Off provides on-demand data movement for less predictable or ephemeral workloads. With “background sync = off” data movement is initiated on demand. For example, an artist at an animation studio requires access to the full collection of textures for a project but will only need a subset for a given model. “Background sync off” provides immediate on-demand access while streaming only the required data blocks to the destination clone. Users and applications get consistency and immediate local access to data while bandwidth and capacity on the destination site are preserved.
By offering this level of flexibility, Global Snapshot Clones break the tradeoffs that traditional replication methods forced businesses to make. Businesses can now choose the replication approach that works best for their specific use case without sacrificing speed or completeness. This means businesses can more effectively manage their data and use it to drive their operations without worrying about the limitations of their replication system.
Global Snapshots Within a Single Cluster
Although named “Global,” these clones are not limited to multi-cluster workflows. The Global Snapshot Clones functionality can also be utilized locally to duplicate virtual machines, git repositories, and other assets commonly copied as “golden images” for multiple users and applications. As mentioned in the introduction, a local clone provides immediate recovery from ransomware attacks providing instant access to a known good data set so your business can get back to production without waiting for time consuming restores.
The future of data movement for an intelligent global namespace
Our Global Snapshot Clones capability is a preview of the innovation we will bring to a global namespace that can stretch from edge to cloud. Today VAST clusters are deployed not just in enterprise data centers but in remote locations including ships at sea and small form factor ruggedized systems designed for edge deployment. As data sets and workflows become ever more decentralized VAST will bring a new set of intelligent data management features that allow customers to seamlessly access their data at the right place and time. Global Snapshot Clones are just the first taste of immediately accessible and usable global data and we can’t wait to show you what is coming next.
