XigmaNAS

Is a comparasion on FreeNAS, not in Nas4Free, but result can be interesting for people reading this article.

http://forums.freenas.org/showthread.ph ... ce-Testing

Interesting, but it just confirms my long held opinion that if you build a good working box to begin with there is little to be gained from tweaking the basics. If you want to improve performance you need to make a real effort and invest in better hardware/software. A lot of effort was spent for little return.

Thanks Raul.

very nice and interesting post, worth reading it.

For me it confirms also that you need to know what you want prior building a server.
For a office environment I would mainly go for performance, speed and reliability,
for a home NAS I would go for low power consumption and at a performance level I am happy with.

In my case I can only choose to use the onbord (Realtek) NIC 'or' to get a new mainboard and processor,
would it be worth it? I doubt...

Well worth the read and I agree with al562 You really need to know what you want before you build a system.

I wonder why he only used a single nic.. kind of a waste of speed with the drives pushing out 300MB/s leaving the nic as the bottleneck. Though he would have to upgrade his switch to something that supports LACP and add more gigabit ports to the nas and his system. Then again.. it was for just a single nic test

We did a lot of looking at what happens with the NICs at the m0n0wall lists... The main difference is additional support (v-lan tagging, and things like that) and the CPU load. If you have CPU to burn, that is not noticeable, but on a smaller system, it can make a big difference. The other, and less common difference was stability. Occasionally the RT chips would just be somewhat unstable for no good reason...

Lee Sharp wrote:Occasionally the RT chips would just be somewhat unstable for no good reason...

except for being a Realtek card