Eliminate Transition Points for More Efficient Cloud Data Protection

Tintri Cloud Connector eliminates transition points that create bottlenecks in competing approaches to cloud data protection.

When you think about use cases for cloud storage, data protection and disaster recovery are likely among the first things that come to mind. But not all approaches to cloud data protection are the same. When Tintri introduced Cloud Connector, our goal was to create a simple-to-use solution that is extremely efficient in terms of capacity and network bandwidth consumption. A series of blogs in the coming weeks will look at various aspects of cloud data protection. This post weighs the pros and cons of the most common cloud data protection workflows.

Transition points and cloud data protection

Cloud data protection workflows involve replication of data (often in the form of a snapshot) from your on-premises storage to the cloud. The figure below shows the most common workflows.

The main thing to notice is that in every one of these approaches there is at least one transition point—in effect a middleman—before data reaches the S3 object store. While these transition points may confer some advantages (like enabling you to keep using your existing backup software), each one introduces another point of management (usually with a separate interface) and also adds cost. Perhaps more importantly, these transition points also have a negative effect on the speed and efficiency of data movement, the network bandwidth needed, and the amount of storage consumed. It is far more efficient to move data directly from primary storage to cloud storage.

Tintri cloud connector eliminates transition points

Tintri asynchronous replication operates in two modes:

• Node-to-node. VMs are replicated from one Tintri storage system to another.
• Node-to-S3. VMs are replicated from Tintri storage directly to an S3 object store.

Using Tintri Cloud Connector, you can create and replicate individual VM snapshots:

• Across a cluster of Tintri storage arrays
• To a third party on-premises object store like IBM COS
• To a third party public object store like AWS S3
• Simultaneously to both Tintri storage and S3 object stores

Tintri shrinks the workflow for replicating data to the cloud as shown below.

This means that there are no transition points to increase complexity, introduce bottlenecks, or reduce storage and network efficiency.

Tintri cloud connector benefits

Eliminating transition points from the data protection workflow offers a number of additional “end-to-end” advantages that competing approaches lack:

• A single point of management with a single management schedule for on-premises and cloud data copies
• Common metadata format
• End-to-end checksums
• Complete preservation of storage efficiency
• Faster, more efficient recovery

Future blogs in this series will discuss several of these topics in more detail.

Next time we’ll look more closely at management and the importance of granularity.