The philosophy underlying SDS is that data services and management control are both abstracted from the storage appliance, allowing for agility and scale in operations, while preventing the historical lock-in to specific vendors that has pushed storage costs up significantly.
Clearly, this looks like an attractive business proposition to the IT community, which has led to older products being rebranded as “SDS” where clearly it’s business as usual. At this time, a good rule of thumb is that true SDS packages come from startup companies. This carries a bit of risk for the user, which can be mitigated by careful selection of the software vendor.
Making the Right Choice
Key decision points in today’s market revolve around ease of use, and particularly how complex a new installation is to set up. In the ideal world, policy files would be common across all of the better vendors; basic, and even advanced operations, should be easy and intuitive to set up; recovery should allow files, images or volumes to be identified and restored with low pain and high accuracy and saved data should be in a format that other backup tools can access rather than some proprietary format that is aimed at creating lock-in.
Good tools are hardware agnostic. They handle SANs, NAS and DAS with all the standard protocols. Just as importantly, good tools are cloud agnostic. They can talk to ALL the major cloud services, providing backup for cloud storage, and tiering between different archiving services.
Cost tracking is a needed feature. The art of storage management in the cloudy future will revolve around total cost management with the implication that data will move around a lot in a highly competitive and rapidly evolving market. We are still in the early phase of this and a “Good” rating implies reporting is available, though in two tears year time having “what if” capability will be the norm for that rating.
The last major decision point is of course price. With a spectrum covering historical vendors who still charge many thousands for a solution, and startups at a level that falls in line with cloud prices measured in cents, this is perhaps the first measure to be applied in looking for a short list.
Why Traditional Vendors Struggle
The historical model for pricing is a lead weight around the neck of the traditional vendor. They look for client engagements yielding thousands of dollars, even down into the SMB space. They are loaded up with the need for training, expensive support and add-on features.
Technically, they suffer, too. Most traditional software comes from a tape-based era, with evolutions for disk-based backup in recent years. The cloud is requiring yet another makeover. A good analogy is if Chrysler were building today’s green vehicles on a Dodge Dart chassis. There’s just a lot of baggage!
The SDS Contrast
The SDS startups are coming from current master design specifications that conform to how data flows during its lifecycle today. User interfaces fall in line with current expectations, rather than overstretched CLI paradigms. Flow is optimized for efficiency and parallel scalability allows transfer speeds to be sped up or throttled as needed.
The Economic Argument
Cloud-based backup and archiving is very cheap today, with numbers in the cents per month range. Clearly, the economic argument for using cloud services is compelling and the result is a wholesale movement of archiving and backup to the cloud.
The transition to new SDs applications for this space isn’t moving as fast, even given the mismatch of traditional vendor pricing and the cloud. In part, this is laziness…there are petabytes of existing archives to reprocess out to the cloud, while the usual fear of change adds to the lethargy…where users feel that the status quo using existing traditional software makes for a smooth transition. But often that smooth transition is to a locked in proprietary structure in the cloud.
Inevitably, staying with traditional vendors will lead to much higher archiving costs, but the fears of transition have to addressed by both the new SDS software and by cloud vendors. The tools should ease the transition by automating the move from existing solutions in the salt mine. Cloud vendors should make available upload from tape and other media and minimize the cost of doing so.
Software-Defined Storage is about to remake the storage market. Startup backup software vendors are aligned already to the scale-out philosophy enshrined in the SDS approach and aim to deliver hardware agnostic solutions. These offer great agility and resulting cost savings over traditional approaches and best service the cloud backup/archiving approach that is fast becoming the industry standard.
The software vendor base is fairly broad so some categorization is useful. There are three identifiable groups. These are the SDS backup startups, the non-SDS startups and the traditional vendors. Our focus is on the SDS backup startups, but it’s essential to understand where the other players sit, especially as claims of “SDS-like properties” abound, causing some confusion.
Click here to view comparison sheet in PDF.