I had 5 drives in my array (1 parity and 4 data). My parity drive was 750 gig drive and I bought a Seagate 1TB drive to replace that with.

 

I added the drive to my system and the first thing I noticed was the hard drive activity light... it's always on for this drive (from the moment the MOBO gets power the activity light is always lit).

 

So I replace the parity drive and the system rebuilds parity. Next I try moving some movies to the server (20 gig movies). It took me 5 attempts before I was able to successfully write a movie to the server (writing to a user share). 4 of the 5 times the transfer gave me errors. I don't remember exactly what the error was... I think the last one had a winrar error of some sort (I'm not sure why this error came up... winrar is the default file viewer for the ISO I was transferring though). Before I added this 1TB drive as parity I had no issues writing to the unRAID server.

 

Also, the transfer speeds were much slower then before. With my 750 gig drive I could transfer the movie in about 20 minutes. With the 1TB drive it took about an hour.

 

I'm guessing there's something wrong with this drive? Can these type of errors happen with a bad parity drive? There's no errors posting on the http main page. Just to rule everything out I put back the 750 gig drive as parity and am rebuilding parity again so I can test transferring files.

 

The drive is a Seagate Barracuda ES.2 ST31000340NS. I thought I'd buy an ES drive for the parity slot thinking it would be better... guess not?

The only way to give you advice would be to have you attach a copy of your syslog, before you reboot again.  That way, it might have clues about why your new drive is slower.

 

It could be almost anything.  From the drive itself being slower, to it having problems interacting with your hardware, to something completely unrelated (activity on your LAN, or bad memory)

 

If 4 out of 5 transfers of files result in errors, I strongly suspect a hardware issue... the memory and the power supply being the most likely suspects.

 

Joe L.

I already rebooted last night when I swapped back my 750 gig as parity.

 

I copied some movies with the 750 gig drive back in place and there are no more issues and transfer speed is back to about 13.3 MB/sec (so says Windows Vista) and transfer time is about 25 minutes (instead of a little over an hour). With the 1TB drive Vista said the speed as about 9.6 MB/sec (but sometimes my PC would stop sending the file to the server for 10 or so seconds... like the server couldn't keep up with my PC).

 

I'm pretty sure it's not my hardware that's causing the problem... it could be the hard drive and how it interacts with my hardware though. Memory is fine... passes memtest and is writing with no errors now. Powersupply is a PC Power 750 watt with single 12v rail at 60A (so my system isn't maxing this out with just 6 hard drives).

 

Something else I noticed the other day with the 1TB drive as well... when windows said it had finished writing the file refreshing the HTTP status page showed the file was still being written to the hard drives. It took about 30 seconds after windows said it had finished to stop writing. I found this odd since with my 750 gig drive when windows says it's done so does the unRAID server.

 

I also heard (only once while I was in the room) a loud click sound coming from the server during once my my tests. Almost like a drive had been spund down quickly (but louder then usual).

 

I'm still pretty certain it has to be hard drive related. If it weren't, why would I have no issues with my 750 gig as parity with alll else being constant? The only possible answer (I can think of) is that the 1 TB drive draws more power then the 750 gig and that's showing a weakness in my Power Supply. But I put in another hard drive and have no issues with the 750 gig as parity. 750 gig parity plus another hard drive should be drawing more power then the 1TB as parity.

 

All signs point to a bad hard drive. The problem didn't happen until I added the drive to my array. Maybe the SMART sensor is bad (hence the HD activity light always on) and it is creating write errors but unRAID isn't showing those because the SMART sensor isn't working? Maybe there are no write errors but the drive itself has a defect (like it spins slower then it should or something is bent inside). I have no idea. I'm used to hard drives either working or not working. But I do know that hard drives can still work but cause some issues...which is what I think I have going on here.

I would suggest is that you run "smartctl" on the new 1T drive.

A description of how to use it, and a link to download it is here: (part way down the page, where it describes hard disk failures)

http://lime-technology.com/wiki/index.php?title=Troubleshooting

 

Also, while the 1T drive is in place and acting poorly capture a syslog for us to analyze.  Who knows, it might have the clues needed to figure out what is happening.

A description of how to capture the syslog is on the same page listed above.

 

Joe L.

I shipped the drive back to Newegg this morning so I can't run any of those checks. I RMA'd it for a replacement so I will be getting the same make/model drive back.

 

I'll tun all those tests once the drive has been added to my array. It could be this make/model drive isn't compatible with something in my system (like Chenbro backplanes or DFI MOBO. Hopefully it works though so I won't have to troubleshoot.

 

Thanks for the quick replies Joe. I'll report back here once I get my new drive installed and tested.

Just for some further education. See the associated link

http://smartmontools.sourceforge.net/BadBlockHowTo.txt

 

We should probably have smartmon tools installed as part of the regular distribution.

This way we could get warnings about potential problems early on.

Thanks for the extra information... more stuff to read

 

I have to agree with you... I think there should be a few tools installed as part of the regular distribution.