Attached Storage Testing

[thiggins]

This doesn't really belong in NAS discussions. But I figured this section has the audience who would be most interested.

I'm thinking about adding coverage for attached storage products to SmallNetBuilder. Since these are simpler products than NASes, reviews would be short (think a slightly expanded version of a "New To The Charts" article), but test data would be entered into a new Chart section.

Focus would likely be on multi-bay products with eSATA (and USB 2.0) interfaces.

My default plan would be to perform the same iozone and Vista SP1 filecopy testing that I use for NASes. Do you think this is enough? If not, what would you like to see?

Thanks in advance for your help.

 

[valentin]

CPU usage % would be good too.
________
Easy Vape Vaporizer

 

[thiggins]

CPU usage % would be good too.

Can you explain why, please?

 

[corndog]

Tim,

This would be a great idea. And in keeping with this (and with the bright orange "Net" in SmallNetBuilder), it would be great to review consumer-level NAS-heads too.

Three examples come to mind:

Addonics NAS Adapter
The Linksys NSLU2 "Slug"
DroboShare

There may be more, but I'm not aware of them. Sadly these devices are quite underpowered compared with current leading NASes. What I would like to see is a powerfully designed Pro-Sumer NAS-Head - a device that packs the power of the QNAP 509 or the ReadyNAS Pro but has no disks in it at all other than the flash for firmware. Instead it has 4 eSATA ports, 2 FW800 ports, and 2 USB ports on it. It should also have 2 to 4 Gigs of RAM. Then I can provide my own storage to attach to it, and it provides VERY fast access to that storage on my network.

Getting back to DAS reviews, I'd like to see reviews on products from Addonics, Vantech, Sans Digital, Lacie, Wiebetech, Sonnet, Glyphtech, and of course all of the DAS options from Thecus, Synology, Buffalo, and the others who provide NASes that are already reviewed here.

Being at the "pro-sumer" end of the Small Net Builder arena, and also being responsible for a large networking and storage infrastructure at work, I'm interested in the whole range from the cheapest consumer stuff right into the Pro-sumer space - including stuff you'd find at ridiculous media creation shops like Vistek and Henry's camara.

Bring it on man!

 

[valentin]

Can you explain why, please?

1 I think that if the CPU is loaded 100% then it is limiting the transfer speed

2 If I use a low-powered laptop as a torrent client via attached USB drive I would like to know if the laptop would be usable for anything else.
________
MFLB

 

[valentin]

OK, so I see it would very much depend on the CPU itself

Still it is important information. May be to test with somewhat powerful, any dual-core, and with an Atom or Nano.

Otherwise we would not know if the low-power processors can handle that load, I think they would not.
________
Vapir No2

 

[scotty]

I think this would be interesting to see...

Don't suppose you could test devices that are 'attached' via. Fibre Channel? j/k Honestly though there's a huge lack in higher end storage benchmarks online in general. Probably because most sites can't afford to be throwing SANs around...

Sounds interesting though, I'll be looking forward...

 

[thiggins]

1 I think that if the CPU is loaded 100% then it is limiting the transfer speed

2 If I use a low-powered laptop as a torrent client via attached USB drive I would like to know if the laptop would be usable for anything else.

Ok. I understand. Host CPU utilization.

Plan would be to test with the same new NAS testbed machine so as to not have the test machine be the limit.

 

[thiggins]

Don't suppose you could test devices that are 'attached' via. Fibre Channel? j/k Honestly though there's a huge lack in higher end storage benchmarks online in general. Probably because most sites can't afford to be throwing SANs around...

Uh, no. Sorry. I would not be aiming at that level of storage anyway.