In this article, Steve explains how to plan a reliable backup strategy for files that are infrequently changed using two clouds simultaneously.
One of the reasons we switched from our old Cloud Backup solution to CloudBerry MBS back in 2014 was the choice of independent backend cloud storage options that MBS offered. At the time only Amazon AWS had everything we needed (encryption, compression, fast retrieval, etc). And at $0.03/GB/month, we were able to offer better backup pricing to our clients and realize a larger profit.
But we became concerned with having all of the backups in the “AWS bucket”. What if AWS was not available, or more likely, we did something to delete the 90 days of backups through a faulty retention policy change? We had a single point of failure.
So when Google came out with their Nearline backend storage platform in mid-2015, we made a decision to setup a redundant backup model using both AWS and Google. That way if something happened to the backup data at one backend location, we would still have the other to rely on.
This has “saved our bacon” on more than one occasion when we went to restore a file and realized that one of the filters was set wrong in the Google plan, but were able to get the file from Amazon instead. And over time we have refined our strategy so that the primary backup goes to Google Nearline, and the secondary backup is sent directly to Amazon’s Infrequently Accessed Storage (Stadard IA) ($0.125/GB), then migrated to Glacier.
The bullets below highlight our backup policies:
- There are two Cloud file backup plans for each server – One for Google Nearline and one for Amazon SIA.
- We also do SQL backups to both cloud locations. And because you cannot do two Full/Incremental SQL backups of the same database, we set one to incremental and the other to “Fulls only”.
- We keep 90 days of backups/versions using the delayed purge feature to ensure versions are kept for a full 90 days.
- For clients using our Local Backup option, there is a third copy of the backup plan with a one-year retention (space permitting). We also do local Image/VHDx backups daily.
- We use Cloudberry Explorer to migrate data older than 122 days from SIA to Glacier, which is a real bargain at $0.004/Gb/Month. Because of the full/incremental purging nuances, migrating files to Glacier prior to 122 days can result in early deletion fees when using a 90 day retention period and monthly fulls.
- We send system level Hyper-V backups to Nearline once per month for Disaster Recovery, however we do not send the large data VHDx’s (D: drive) to the cloud as we already have the files stored in the cloud.
- For those clients who do not use Hyper-V, or whose Domain controller resides on the Hypervisor, we do Image-Based backups to the cloud once per month, but again only of the system drive (C:).
Another big advantage of having two cloud backups is that if one of them fails and the other runs successfully, it is no cause for concern. It is only when they both fail , usually due to permissions problems, that we need to investigate further. And while this approach does double or triple the number of plans that need to be managed, the MBS Monitoring console makes it very easy to identify what is happening.
We are currently considering using Google’s new Coldline platform ($0.007/GB/Mo) for large scale archiving of 3-4 TB for a potential new client, since it does not have the delayed retrieval and high retrieval cost of Glacier.
The great thing about using the combination of Cloudberrylab MBS, Amazon and Google is that we can provide our now 100+ clients with redundant, secure backups, at an extremely competitive price.
About Steve Putnam:
Steve retired in 2013 from a Fortune 500 firm as Director of Storage and Backup, after spending 40 years in IT. He now manages the Cloud Solutions Services for The PC Wizard, a small Managed Service Provider (MSP) owned by his son, who started the company in 1995 when he was only 15 years old. In this new role, Steve has more time to devote to exploring new technologies, as well as to engage in one of his passions – writing technical articles to assist others in maximizing their use of Cloud-based solutions.