<html><head><style type='text/css'>p { margin: 0; }</style></head><body><div style='font-family: Verdana; font-size: 8pt; color: #000000'>Hi,<br><br>Drobo's not my first choice either, but I had to be up quick and I didn't have another system to build a RAID array with. I've hooked up the Drobo on a Mac with HFS plus (gives me a potential 16 TB array), and it is slow.<br><br>Once I have my data on it, and off of my failing raid, I'll most like go back and rebuild the raid and use that. I'm much more comfortable with it, and the speed is great. Then I'll relegate the drobo to backups.<br><br>My software 4TB RAID-5 had been running fine for 2 years, and my older 1TB RAID-5 is still running after 5 years. What happened was: I threw two extra GB of RAM into the box, and I didn't notice that I started having serious speed/cpu issues. Any HDD access would use 100% CPU, in kernel. About 4 hours after I put the RAM in, I had a HDD failure. I bought a new HDD and put it in the array, which is when I noticed the CPU usage. I didn't think it was a RAM issue. Rebuilds were VERY slow (estimate 90 days for 4TB). I eventually figured it was the RAM, took the RAM out, and my problems went away. A rebuild was now going to take 5 hours. At 97.5% rebuild, an existing drive in the array got a permanent read error, causing the rebuild to stop. There's no data on that sector, but I guess rebuild doesn't care. So, out of a 5 disk array, 1 is a new disk, and 1 has a read error during rebuild. I'm SOL. The array comes up degraded just fine, so I had to get my 3TB of data off the array as soon as I could. I didn't realize the Drobo was so slow: scp file transers over GigE are at 24MB/s.<br><br>I'll continue with the scp copy, OSX sharing the Drobo over SMB, and my backends using it for storage of Live TV and movies, etc.<br><br>We'll see how it goes.<br><br>Gerald<br><br><br>----- "John Drescher" <drescherjm@gmail.com> wrote:
<br>> > Personally, I don't understand the fascination with Drobo. It's a<br>> > proprietary system, so you're stuck with what they want you to have.<br>> <br>> I do not either. This will be no more reliable than the raid 5 you are<br>> coming from. To me its very unusual that 2 drives had a hardware<br>> failure at exactly the same time. If both drives did not have an<br>> hardware failure that prevents reading the drives this is most likely<br>> a recoverable situation if you have software raid. With hardware raid<br>> (hardware raid controller) you may not be able to force the drives to<br>> mount so this is a less reliable option in my opinion.<br>> <br>> >For just Myth, storage groups with a couple drives are likely a better option.<br>> <br>> Totally agreed.<br>> <br>> John<br>> _______________________________________________<br>> mythtv-users mailing list<br>> mythtv-users@mythtv.org<br>> http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users<br>> </div></body></html>