N3200 Falls Unresponsive

[Draco]

We purchased the N3200 with three 750GB drives in RAID5 for primarily file backups, and have been running it for about 3-4 weeks. We started with the 1.00.01 firmware right off the bat when we received it.

I've been running into some odd issues. Just a heads up to the type of backups: our full Sunday/Monday backups for our servers comes in just over 50GB each. We can achieve read speeds in excess of 40-50 MB/s (FTP, possibly SMB as well) with jumbo frames enabled (4000). Writing is obviously slower, but I have not graphed that as of yet. Anyhow - After one of our large jobs involving Win2003 using SMB+mapped drive+ntbackup, it almost seems to weaken the device of sorts.

It seems to gradually get slower at which it then drops off the network completely OR slows to a CRAWL (eg. stalling in both the admin control panel and in network activity) which eventually leads to an unresponsive state (with error on the LCD front + tone).

I have my suspicions on jumbo frames possibly being the culprit, and I would love to ssh in, but there doesn't seem to be any sshd on this device that I can enable without throwing my warranty out the window. Correct me if I'm wrong of course. I would like to see what's going on inside with more detail. I'm sure a dmesg would tell me something more substantial than what's given at that silly/worthless logs panel. Is there some sort of beta program that willing participants can join? Is the firmware revision even to blame? Maybe I should roll back to the manufacturer-shipped 1.00.00...

Any help or insight would be appreciated. I wasn't pleased that it didn't email me at all over the weekend to report anything wrong despite the obvious degraded performance and subsequent halting of the device.

[tscholle]

My unit does not e-mail me either. I have talked to support about this. I let them dial in for them to look, they claimed it works if they point the smtp server at their server, but not at my server. That is great if I used their server, but I do not. Also my FTP connection seems to have issues.

Good luck, I will be watching your post to see if anything can help me!

[slypie]

I'm starting to get a bit worried as I've just forked out a lot for this device and 3x 750GB hard drives.

[tscholle]

Ya, I am wishing I had gotten something else, this unit can not seem to do anything I got it for and while support seems to try none of my issues have been resolved. I wish I could return the thing.

[slypie]

Ya, I am wishing I had gotten something else, this unit can not seem to do anything I got it for and while support seems to try none of my issues have been resolved. I wish I could return the thing.

What issues do you get with this device?

[Draco]

Just a small update:

Jumbo frames have been disabled and are probably going to stay that way.

I can't complain too much here. but it's been up and stable so far (6 days, 22 hrs, nothing note-worthy I might add), NFS speeds currently are between 21-28mb/s which is sufficient enough, but is hardly spectacular either considering I've milked 40+ out of this thing before.

The "small update" part I state is regarding a(nother) *possible* bug here with FTP. We're using on our backup server CentOS/ncftpget and it fails on one of our larger ntbackup bkf files which is currently 123GB in size. With about 50 seconds to go at the ~121GB mark, it craps out with "Remote host has closed the connection." - wouldn't it be great to have more daemon feedback on the Thecus? Maybe I might be able to resolve why it consistently dies at this point and only at a file this large... or maybe it's just an ncftp issue? I suppose I couldn't rule that out yet, but it's a pretty fat chance.

So now I fell back to just using NFS copy operations instead. (for some reason, FTP was.. on average, 5-10mb/s faster or so.)

Alright, I'll stop now... this wasn't quite as small of an update as I stated it to be.