Back to "serious business" after my belated April Fool’s Day flight of fancy.
“Depends” is a word with multiple relevant meanings in IT. Thanks for opening that can of worms, Juan.
As technology architects designing systems, we all live and die based on the ability to correctly identify, qualify and quantify need, and then build solutions to problems dependent on that analysis, accounting for product capability, best practice, cost and other things. I think Juan wrote a great bit about this in his last post on hardware versus software deduplication.
We also rely on correct understanding of system-to-system and user-system dependency (interoperability and user acceptance). Whenever some technology gets somewhat close to plug-and-play (think Fibre Channel in 1998 versus today), the IT hydra sprouts new heads, and something new with a stack of dependencies emerges – think 10GbE today. Tied to that is the ability to support / keep a solution running, which Jack correctly pointed out depends as much on the skills/culture/comfort of the teams managing it as anything inherent to the technology itself.
And finally “Depends” can also refer to the popular brand of adult diaper we've all worn at one point or another during particularly stressful projects. IT is a dangerous job; it takes its toll.
We engineers have a love/hate relationship with the fact that so much depends. Our jobs would be less stressful, but too boring (or maybe completely non-existent), if everything just didn’t depend so darn much…
Fortunately, We Can Show You How Much It All Depends.
I'm all about the merciless and single-minded promotion of Datalink Labs, so I’ll note that the purpose of the lab is dealing with these depends-related problems:
1. Showing customers how these solutions actually look, feel and work so they’re well-educated (and therefore, hopefully more comfortable) during the decision-making process.
2. Increasing our technical teams’ understanding of how things work (independently, and in complex interdependency) to the highest level possible. Our solutions rarely touch just one vendor these days (see Kent's notes on unified support), so this is critical.
The labs are one tool in our arsenal to ensure that we can figure out why, when, and how much things depend, so we can translate that into solutions. Further piggybacking onto Juan's last post with a dedupe theme, here's a list of deduplication-related technologies available in the lab:
Symantec – NetBackup PureDisk, Deduplication Appliance, Client Deduplication for efficient data protection
Quantum – DXi backup appliance for efficient data protection
NetApp – Deduplication for FAS, plus other efficiencies offered by FlexClone, SnapMirror, SnapVault, etc.
Hitachi Data Systems – Archive single-instancing, and dedupe-enabled virtual tape for storage and data protection efficiency
EMC – Data Domain, VNX, Avamar – efficiency for primary/archive storage and data protection
IBM – ProtecTIER virtual tape for efficient data protection
VMware – Linked Clones for efficiency; backup appliance
Riverbed – WAN acceleration (features deduplication)
SEPATON – Virtual tape offering deduplication for efficient data protection
CommVault – Source and target deduplication for data protection
... yes, long list. If I missed some, I'm sure someone will correct me.
If you've made it to this point, I sense you are now experiencing a powerful urge drawing you to Datalink Labs, so you can watch these solutions compete and understand their differences, both large and subtle. Request info today!
Let’s be careful out there. Regards, Rick