Data Protection: A Topic Most Storage Vendors Prefer to Avoid
Having built a good portion of my career in the storage industry, it’s clear to me that storage vendors generally just don’t like talking about some topics. Companies aren’t shy to brag about storage throughput or IOPS performance, their attractive price per terabyte, or how their product is so easy to use. Panasas is no different here – we are also quick to tell you about the benefits in our ActiveStor platform.
But when it comes to discussing what can actually go wrong with their product, most storage vendors understandably find it an uncomfortable subject and go silent, despite the critical importance that data protection represents to their customers. Instead of embracing a need to discuss this topic, most vendors prefer to simply take the position that their products are so reliable and the odds of anything serious going wrong are so low that that they’re not worth discussing.
This seriously needs to change. Sales reps and marketing literature rarely discuss the probabilities of data loss, data unavailability, risks inherent with long RAID rebuild times, data recovery capabilities when one too many drives fail, how reliability changes with the scale of the deployment, and so on. But they should. The reality is that no storage product is perfect and failures in storage systems are inevitable – especially when deployed at scale. Most seasoned storage managers have experienced storage failures that were very unpleasant, if not catastrophic, at some time in their past so this is truly a subject that matters.
It IS worth discussing because it is YOUR data at stake.
Panasas is taking the lead in discussing data protection risk mitigation and disaster recovery. We’ll tell you, in detail, what we do to maximize data protection at scale. And we go a step further by discussing how we preserve availability and speed time-to-restore when systems experience catastrophic failures, making them much less catastrophic. Panasas ActiveStor hybrid scale-out NAS solutions deliver significant innovations in both reliability and availability:
- Per-file distributed RAID increases data reliability with scale
- RAID 6+ triple parity data protection implemented in software with erasure codes, 150x more reliable than dual parity approaches
- Rebuild performance scales linearly with the size of the system
- Extended File System Availability (EFSA) avoids complete file system outages and keeps unaffected files available in failure modes that would take down the entire file system of most other storage systems
- Quadruple-mirrored directory data protection, making it easy to identify and quickly restore affected files in the unlikely case of three simultaneous drive failures
We are not shy when it comes to discussing the “unpleasant” matters of data storage. For more information, please visit the Why Panasas and PanFS RAID 6+ product sections of our website to learn more about how we deliver reliability at scale.