Performance tests for Black/Green 5440/7200?

I have been searching for a long time today on the internet and on these forums and it is incredible how few actual real world performance tests there have been on the Drobos. I would love it if someone knows of where I can find (or has the hardware to do) actual tests on the difference between 5400/7200 or the difference between Caviar Black/Green drives. Ideally this would be done in DroboPro or DroboElite and with using various interfaces (iSCSI, Firewire 800, USB 2).

according to posts DRI have made (mainly on the old forums)

on drobo v1 and v2 - no difference to end users

drobopro - extremely MARGINAL difference in benchmarks - no real difference to real life usage

droboelite - definitely faster with 7200 rpm drives

since they havent really posted anything more, i cant comment on FS/drobo S

Thank you for the reply. That is basically what I figured, but I would love to see the actual stats, too so I don’t end up wasting any money either direction :frowning:

How full the Drobo is and how many discs are inserted all have dramatic influences on performance. My DroboPros all have 8 2Tb WD Green drives. The fastest is the DroboPro via iSCSI; even though it is full I still see it write at over 40MB/s. The slowest by far is my full DroboProFS where 16 or 17MB/s writes or reads are a good outcome. My empty DroboProFS gets26MB/s writes. Jumbo frames on the FS units. Dual disc redundancy also takes its toll.

As a test I bought a QNAP 219P and put 2 OCZ 120GB Vertex 2E drives in and set up as RAID 0. I connected the NAS to the same network as the Drobos. It would read way in excess of 70MB/s but writing was less than 30MB/s. These were brand new SSD units and will probably represent the fastest performance of media in a NAS. I doubt that there are many situations where a 7200 drive will be faster than a 5400 drive. There is too much processing going on.


wow - a 5 bay synology (DS1010+) with 5400 RPM drives in RAID 5 can read and write at over 100MB/sec over ethernet

I am excluding any reads and writes that occur into memory, this is what I get copying very large files. I have seen huge copying speed until the memory is used up. Of course it also depends on what read/write technique one is using; mine are just Win 7 files being dragged to a directory on the target.