- Also, our customers look for fast connectivity and cost efficiency.
- It's TCP/IP vs. fiber channel, which tends to be more costly.
They need to improve the go-to-market for all-flash and converged infrastructure. What is your goal-to-market vision, and when to get there? They’re too slow compared to others and what they’ve done in the past. They were the leader in dedupe, but now, it’s not such an innovative edge.
It lacks flexibility in failover and failback, so we cannot granularly failover pieces. It's not easy to move one piece over to the other side.
Also, from the overall workload standpoint, all protocols are handled in just one physical architecture. So if I'm running dedupe, fiber channel, and other protocols on the same CPU core, I can’t load-balance. I’ve seen issues specifically with EMP, one core is maxed out, and I can’t use the other cores to handle it.
Fairly solid 5-9 array. FAS is a solid architecture in 90% of the environments.
Scalability especially in SMB range has been well-received. So long as the environment is sized correctly, it’s been good.
I have had both good and bad experiences, depending on what tier I get to initially. Now it’s tiered, whereas it used to be one senior guy.
If historically you’re a NetApp customer, it’s not as complex as cluster mode. It requires a lot more complexity – command line is not so friendly for storage admins. I’d recommend also sticking with what you know.