now that is a nas test bed

[teknojnky]

http://www.anandtech.com/show/6241/building-the-2012-anandtech-smb-soho-nas-testbed/


schwiiing!

 

[stevech]

product comparisons seemed to be (a) unrealistically heavy multi-user loads and (b) apples-to-oranges in terms of products, e.g., comparing a mid-range to a high-end, in speed and cost.

 

[thiggins]

Thanks for sharing that.

Maybe I missed it, but I couldn't find what the performance goals were for the testbed. Can all 12 VMs simultaneously handle 125 MB/s read/write?

 

[teknojnky]

Thanks for sharing that.

Maybe I missed it, but I couldn't find what the performance goals were for the testbed. Can all 12 VMs simultaneously handle 125 MB/s read/write?

given an all SSD storage with dedicated drives and NIC's per VM, and 64 gig of ram and dedicated RAM shared disk, I would expect it to deliver 125mbs per vm, even though I do not see it explicitly stated either.

on first page

Our goal was to simulate a SMB (Small to Medium Business) / SOHO (Small Office / Home Office) type environment for the NAS under test. From the viewpoint of our testing, we consider a SMB as any setup with 5 - 25 distinct clients for the NAS. Under ideal circumstances, we could have had multiple PCs accessing the NAS at the same time. However, we wanted a testbed which didn’t require too much space or consume a lot of power. It was also necessary that the testbed be easily administered. These requirements ruled out the possibility of multiple distinct physical machines making up the testbed.

further in

The guest OS on each of the VMs is Windows 7 Ultimate x64. The intention of the build is to determine how the performance of the NAS under test degrades when multiple clients begin to access it. This degradation might be in terms of increased response time or a decrease in available bandwidth. Both of these can be measured using IOMeter. While IOMeter is the controlling program installed in the host, each of the VMs run the Dynamo workload generator component. Dynamo and IOMeter communicate through the internal network to ensure that there is no effect on the benchmark runs.

even if the system can not saturate 12x 125mbs, I would certainly expect it to take a pretty stout nas unit to even test that.

 

[teknojnky]

product comparisons seemed to be (a) unrealistically heavy multi-user loads

Well that is the intention, to see how a nas performs under heavy multi-user loads.

this is not much different than how a high end home user expects a nas device to perform when streaming multiple high bitrate HD streams, and/or perform other tasks without lagging (ie bittorrent/dvr storage/time machine backups/iscsi volumes/etc).

and (b) apples-to-oranges in terms of products, e.g., comparing a mid-range to a high-end, in speed and cost.

I don't think the goal was necessarily to compare only the 2 devices presented against each other, but to see how they perform individually and to provide baselines for future nas reviews.