Recently we discussed how Vicinity creates a data fabric to enable data to move between storage systems, locations, and clouds.
In this blog, we’ll offer more specifics on how Vicinity controls the placement of data and uses multiple sites and multiple clouds – not only for protection and availability, but also to contain costs.
Multi-site replica placement
Vicinity allows data to be spread over multiple sites with fine-grain replica placement control in node groups called “zones.” Each site must host at least one data node, while there is no limit in the number of sites.
Why would you want to place data in different sites or zones? There are a number of reasons:
A zone can be defined by geographical parameters or based on other requirements. For example, a zone can be spread across a wide geographic area or it can be confined to specific IP addresses. It is up to the administrator to set these parameters.
In the most typical use case, an application can read/write data at one site while a replica is synchronously written to another site. This scenario has the benefit of fast access to the data, since the application is reading/writing close to where data is accessed.
In multi-site volumes, replicas are assigned to different zones, so each site has at least one copy of the data. There is no service disruption even in the case of a data center failure, since a full, up-to-date replica set exists elsewhere.
The potential benefits of cloud storage are widely known – disaster recovery, security, and, perhaps most significantly, cost efficiency: long-term archive can be far cheaper because you’re not buying and maintaining hardware.
Storage is expensive, and so are its environmental costs – power, cooling, floorspace – so we always recommend older, cold data be stored in the cloud.
Vicinity can use cloud storage in a number of other ways. When a virtual cloud drive is attached to a local volume, Vicinity can ensure one copy of the data is kept locally and one copy in the cloud so it’s always available in case of a local infrastructure failure.
Because a cloud drive has theoretically infinite storage capacity, Vicinity can also use it as “cloud overflow” if the data load grows larger than the on-premise capacity. If you later add local capacity, data can be shifted back on-premises.
When one cloud service is best for capacity, and one is best for performance, Vicinity can use both.
When one cloud service is best for this application, and one is best for that application, Vicinity can again use both.
When you need a different level of access or security for different cloud drives in different places, Vicinity can do that too.
When you need multiple cloud drives in different regions – well, you know, Vicinity does that. Use the same account or use a different account for each provider such as Amazon S3 and Microsoft Azure.
Cloud doesn’t have to mean giving up control
While Vicinity allows administrators a wide range of options for data placement, and unlimited sites and/or clouds, many of these processes occur automatically, requiring no special settings or changes to the devices and applications.
And Vicinity uses artificial intelligence to learn and adapt to the organization’s needs, meaning it evolves as you do.
You can learn more about it here.