Tony wrote:
The dump that you see on your screen is caused by a buffer overflow.
Thank you.
Tony wrote:
but rather the hardware changing it’s mind on how to apply the rules and not tell us.
I'll have to think about that.....
Tony wrote:
then there is only one common denominator left to consider and that is the board.
I beg to differ.
Tony wrote:
I understand that it does fine with Windows and Knoppix, but that is irrelevant
I disagree. The relevance is that the hardware is running satisfactorily with several rather large suites of software!
Tony wrote:
since your goal is to run NASLite. Drivers, kernel, and everything else NASlite is Linux 2.4.33. That is not the case with Windows or Knoppix.
Yes, of course.
Are you seriously telling me you exclude the possibility of a malfunction in one of NasLite's modules?
Extraordinary.
So Naslite is mathematically verified for correctness?
Tony wrote:
All that aside, try the other board and see if the crashing persists. Even if you can’t disable the Mbit NIC on it, it should give you an indication if the problem continues to occur.
But it probably won't, will it? If the result of the problem is buffer overflow, the time required to process the buffers, along with the quantity, will be (possibly considerably) more than the time available at gigabit speeds, so it would simply mask the problem, leading to a false sense of security.
Tony wrote:
Version 2.04, due out in a week or so will resolve your onboard NIC issue and you’ll be able to use your Gbit NIC.
We'll see.
Tony wrote:
It is virtually impossible for us to support all possible hardware scenarios equally well. Regardless, considering the compact footprint of NASlite, the range of hardware it supports is substantial by any measure.
Of course. But it's amazing that you don't even hint at the possibility of a software error!
Tony wrote:
I understand your frustration
Do you?
You & your colleague don't have to gather 3 long strings from disparate sources every time you change a small part of your NAS. We do, a lot.
You probably don't have to repair several hd's on a regular basis using alternative software to NL2.
It's like walking on eggs.
Tony wrote:
and will try and help with what I can. There are some hardware scenarios that are simply incompatible, so moving away from those also removes the cause and resolves the problem.

& adds to the expense & huge amounts of wasted time.
Tony wrote:
Lastly, if anyone else is experiencing this behavior, can you please speak up.
Now that is a good suggestion.
Please bear in mind that a fellow Customer has similar problems.
It would help if we knew if many others were actually using NFS, & *at gigabit speeds*. I suspect far more use SMB.
Perhaps a thread initiated by yourself would get more attention?
I'm currently doing more tests, but not altering the hardware, & so far there are no *apparent* errors.
Instead I've changed the file service......
However, I need to use NFS, or it's all pointless.
Even though some transfers have gone ahead, they are "sticking" part way through for several seconds.
I'll put in the sata again, & I've a 43 gigabyte file to copy as a test to run overnight.