Home Server Board

Well... Basically, I had a 4x500GB RAID5 array in my server connected to the nVidia chipset's RAID controller. I then had 3x750 discs in a Thecus NAS, again RAID5d. Then (home server on a budget, remember) a robocopy script copying everything from the server to the NAS every Monday night. Well, one of the HDs in the NAS failed (Seagate firmware problem), and when I pulled it I also noticed another of the HDs was actually the wrong model, albeit a 750 too, so I ended up sending them both back to where I bought them, hence no more NAS for the time being. I figured after a year of 24/7 flawless operation, the server would stay up for a week or two without. A week later, the 680i board in my PVR box died with the 'double dash of death', so no server board backup (getting nervous now, but what are the chances of two boards failing before the NAS is up again, right?). Then, a week later, NAS still not back up cos Seagate are dragging their feet with the firmware fix for RMAd drives, and the server 680i fails with the same double dash error. So, in the space of two weeks, no NAS backup, no backup mobo to swap out if it all goes to hell, and then no server board :) Ta-daa!!!

I've read lots of reviews now and looked at loads of Supermicro and Tyan boards, and the user reviews on Newegg are pretty damning of the lot. Considering what you guys say too, I'm now thinking maybe a decent desktop board (Asus P5Q-E?) is reliable enough for me while being expandable. Software RAID will have to do for a month or two until I can scrape together the readies for an Adaptec 31205 or 3805. You then reckon leaving the NAS as JBOD, right?
 
How you do it is pretty much up to you. But make sure you have the complete dataset in two places.

from the detail above it seems you were just a victim of timings. The chances of the board failing in the window you didn't have backup facilities are one in tens of thousands.

The only way to guard against this is to ensure you replace failed drives promptly. Depending on the controller you can do this instantly by designating an additional drive in the array as a spare. This drive sits blank and is only used if another drive fails. This is fairly common on enterprise controllers but not sure about WHS etc.
 
You must log in or register to reply here.
Back
Top Bottom