In the description of Amazon’s EBS (Elastic Block Store), they say that
The latency and throughput of Amazon EBS volumes is designed to be significantly better than the Amazon EC2 instance stores in nearly all cases. You can also attach multiple volumes to an instance and stripe across the volumes. This is one way to improve I/O rates, especially if your application performs a lot of random access across your data set.
I wonder how they do that. I presume that EBS volumes are generally stored on separate machines, accessed through the network, and that instance stores are generally local disk. Could be that the local disks are really slow and the SAN disks are really fast and it’s just luck that it came out that way… But I wonder, if you had the same speed of disks in the two cases, is there a way to design things to come out as described above. Have to think about that one…