Why MSPs Love Block Level Differencing (FastBIT)
by Sean Curiel, on Apr 13, 2021 7:54:24 AM
Modern data protection solutions offer us a wide range of methods for backing up. To start with, we have the option of backing up at the file-level or taking an entire system image. These backup formats each have their own benefits and unique purpose for data security.
Next, we are presented with options such as Differential Backup (capturing all file-level differences since the last Full backup) and Incremental Backup (capturing all file-level differences since the last backup of any type). These features give most of us the tools we need to reduce the size of our backup data to something more manageable. Combined with compression and the ability to remove unnecessary data (retention functionality), data-storage costs become less of a concern for most of us.
What About Extremely Large Data Sets or Limited Bandwidth?
There are organizations such as MSPs, who handle the backup of dozens, if not hundreds of clients. For MSPs this often means dealing with extremely large data-sets in a variety of network environments. In such a situation, you may have clients backing up to the cloud with a limited backup window and varying upload bandwidth speeds. This is where we see technologies like File-Level Block Differencing (like NovaBACKUP’s FastBIT technology) making a huge impact on backup for Managed Service Providers.
What is File-Level Block Differencing?
Let’s assume the initial backup of files has been made. The next time a backup is scheduled to take place, a block-level comparison between the previously backed up files and files being backed up is made. Now ONLY the differences, or changes in the files at the block level, will be sent to the cloud. This results in a massive reduction in the amount of data transferred, creating a cloud backup scenario that is as fast and efficient as possible.
Changes to existing files in the average workplace are often small, usually representing somewhere between 1-15% of the original size of the file. Backup integrity is ensured through extensive error checking. If for some reason the block comparison cannot be made, files are simply compressed and backed up using the standard incremental method.
What Does It Look Like in Practice?
The results that Managed Service Providers can experience are dramatic. Let’s imagine that we have a client with a single 20GB database file and a 100 Mbit connection to the internet. At 6 MB/s we can estimate roughly 1 hour for the initial backup. After one week, perhaps the client has added 1,000 new records to the database. This would likely add just ~2MB of new data, which could be transferred to the cloud storage in less than 1 second. While this scenario is oversimplified, we can visualize the implications on a broader scale.
Direct benefits that MSPs experience:
- Fastest Possible Backup Times
- Flexibility in Client Backup Scheduling Window
- Reduced Storage Requirements
- Reduced Client Bandwidth Requirements
What About the First Initial (Large) Backup?
To take advantage of File-Level Block Differencing, an original full backup of all files must be created. As this initial backup can often be large and thus time consuming, NovaBACKUP offers a free backup seeding service to our MSP partners. This service allows Managed Service Providers to ship a copy of their (encrypted) customer data directly to a NovaBACKUP data center to eliminate the need for a full online-backup, saving time and resources. Following this, end users can immediately take advantage of block-level differencing for the fastest backup times possible.
We invite you to speak with our MSP team to learn more about our managed backup solution, NovaBACKUP Cloud for MSPs.