Question about sharding

Hello, everyone. I am taking M103 course, lesson 3, Sharding.

I’m confused when they say a rule of the thumb is having 2TB to 5TB per server. Here, data is being seen like it is stored per server (local storage) but in real production environments data is stored in NAS or dedicated storage solutions that can handle TB or PB.

I hope your comments.



Personally, I would not specify a NAS in a production environment database server. The NAS will add substantial latency for any operations that use disk operations.

What storage would you use? What is the recommended?

Any directly attached devices or SAN most likely RAID SSD or spinning.

Yes. Now I read SAN, yes, you’re right. I was wrong when I typed NAS instead of SAN.

Everything OK.