The title of this episode really should be written in large, friendly letters. When you have a storage infrastructure with hundreds, thousands or more disk drives, hard disk failure is simply a fact of life. There’s no way to avoid it. But there are ways to deal with it. Welcome to the fifth video in our short series explaining our scalable storage solution: SymplyPERIFERY. In this video we’re going to go over what makes our in-facility appliance truly scalable and robust – and we’ll cover a bit about compliance while we’re here.
SymplyPERIFERY takes disk failures in its stride and treats them as routine events to be managed and not an emergency situation requiring immediate human attention and expensive intervention. It’s made possible by a unique system of continual, automated oversight.
In SymplyPERIFERY a health processor continually checks for failed hardware, bit rot – that can cause silent data corruption, replica and erasure coding anomalies, and even network problems. All the nodes in your cluster participate in these functions, which makes failure detection near instant. In the event of a failed drive or node the data whose redundancy is affected (lost replicas) will immediately be re-protected in a massive, parallel, cluster-wide recovery process as long as there’s spare capacity available. All nodes holding an “orphaned” replica will trigger the creation of an additional replica on another node. Since replicas are spread across the cluster, all nodes contribute by doing a small portion of the work. In this way, a disk drive that might take days or even weeks to regenerate in a typical RAID configuration can be re-protected in less than 15 minutes. Which is pretty epic!
In the case of erasure coded data, a very similar active regeneration is triggered and the full power of the cluster is again applied to the recovery process. This is the reason why SymplyPERIFERY is capable of delivering significantly more durability than an identical erasure coding scheme in an otherwise comparable hardware environment. That’s a fancy way of saying we’re better than other storage vendors. In either case, the bigger the cluster, the faster this recovery process runs.
And because this process is automated, SymplyPERIFERY takes care of the hard part so you don’t have to worry. With software-based data redundancy being managed by the system all you need to do is replace hardware – so nodes or disk drives – at a time that’s convenient for you.
This is what we mean when we talk about a self healing system. The software that manages SymplyPERIFERY hardware continually looks for problems and creates copies of data as needed – so you can be assured that nothing will be lost in any event.
This functionality also helps with legal content protection. SymplyPERIFERY provides additional capabilities that enable you to meet regulatory mandates that require content to be stored on non-erasable, non-rewritable media.
Perfiery lets you use Legal Hold to create a point-in-time snapshot of a specified set of files at a specified time. The files are then immutably stored regardless of what happens to the original file or the cluster as a whole. This patented technology means you can prove in a court of law that content has not been tampered with or altered. Integrity seals are based on content only and can be upgraded as newer hashing algorithms replace outdated ones. These features along with Object Immutability (WORM) encryption in flight and at rest, combined with detailed audit trails make SymplyPERIFERY compliant with regulation from the U.S. Securities and Exchange Commission, the Financial Industry Regulatory Authority, the Commodity Futures Trading Commission, and Directives of the European Parliament. So you can implement a 3-2-1-1-0 data protection strategy, ensuring data is protected against ransomware attacks, security breaches, and accidental deletion.
OK, I think we can leave it there. Next time we’ll cover something that we’ve only really touched on: metadata and the uniquely powerful way it works in SymplyPERIFERY.
