DSM-G600, DNS-3xx and NSA-220 Hack Forum

Some questions for you

1 - Why are you at the web menu if you can access the data without problem?
2 - Why are you using beta firmware and storing live data?
3 - Why are you storing live data without a backup?

We all make choices in life - you've made so poor ones, fortunately, it looks like you have a chance to recover without losing your data.

One thing I want to leave with you - if you value the data - back it up, or at some point you will lose it, just so you understand, it doesn't matter where you store it, on the PC, on the DNS-323, or on a $1M storage device - you need to back it up.

I have made an upgrade to 1.04b58 und then i format the disk. Then i put my files on the device and all is ok for a week or so. Since some days, there appears the problem, that i can not go to the web-interface for example to create new users.
@fordem i have the DNS-323 for backup, but i have no other disk with 500GB to clean the DNS-323 and reformat.

Every Firmware you can download by DLINK has a "b" in it and the 1.3 is bugy too. But this helps not for the problem. Nobody here who can say what is to do? Access via Telnet, edit a file ?

fordem wrote:

The point is - it's YOUR data - it's YOUR responsibility to ensure it's secure - and one of the things you don't do with valuable data is store it on a device that your testing "beta" firmware on.  Beta firmware is firmware that has been released for test purposes, and has not been approved for public use - it is expected to have flaws.

The "b" behind a D-Link version no. stays for "built" so for a build no. And NOT for Beta!!!

fordem wrote:

What I am saying is that as of today's date - November 27th 2007 - the latest officially released code is 1.03, release date 4/5/2007 - and that anthing with a higher number or later date is beta.

Here's my source.

hi fordem

We agree that files like music, movies, docs etc. are the same in USA and in Europe.

Yesterday the version 1.04b61 was like your´e words than a Beta version around the world. Today D-Link Germany put this version on their public server. Up this moment it´s in Germany a officially released. But not for D-Link UK and not fir D-Link US and at this moment for no other D-Link country else.

So, than this version might be for US DNS-323 customers a beta version but for German DNS-323 customers it´s not.

Sorry, but your ´e wrong.

Last edited by Tilly (2007-11-27 21:59:58)

fordem wrote:

What I am saying is that as of today's date - November 27th 2007 - the latest officially released code is 1.03, release date 4/5/2007 - and that anything with a higher number or later date is beta, and therefore experimental.

Well, I agree that people should keep multiple copies of their important data, but just FYI, the 1.03 has a bug which will cause it to format the wrong HDD too:
http://dns323.kood.org/forum/t885-OMFG- … rible.html
I really don't understand why we still haven't see a new official firmware on dlink.com after more than 7 months of waiting. DLink knows all the bugs in 1.03, but they just kept release beta firmwares instead of stable ones. As you said, not like using an official firmware, if you use a beta firmware and something bad happens, you can't blame DLink because u are already told that it is a "beta" firmware.

However I am wondering if the can-not-blame-DLink-for-beta-firmware thing is the real reason that DLink hasn't released any official firmware in the past 7 months, they just don't want to take the responsibility. I mean, how difficult it can be to JUST make sure the unit would not format the wrong HDD?

Last edited by dickeywang (2007-11-28 13:25:44)

fordem wrote:

First - if you check the same post you link to - you'll notice that in my response I mention that I have swapped many disks in & out of my DNS-323 without having the problem that you did.

I'm not denying that a problem exists, but simply pointing out that not all users have experienced the problem - which may mean it is related to the environment or what the user is doing - and makes it difficult to recreate.

For D-Link to fix the bug, it will need to be reproducible in their labs - perhaps the elusive nature of this bug has been the delay.

Hi forderm

Yes that´s  absolutly correct!

Because i don´t understand the most problems DNS-323 users have, i guess most if their problems are self-made problems.

Out of curiosity - what do you for a living?  I'm guessing it's not firmware/software development.

Changes made to firmware to fix one bug have a way of creating others, and as such the tendancy is to avoid making "unnecessary" changes.

It's easy to say the "DNS is very likely to format a wrong HDD if the new HDD was formatted as NTFS/FAT32 before it is installed into the DNS", but you have no statistical evidence to support this - and there is at least one thread where the user claims it happened with a new drive - posts in a forum such as this are called "anecdotal" evidence, and will not usually be used UNLESS the complaint can be reproduced in a controlled environment.

You don't think it has anything to do with what a user does - unless one uses a hacked firmware (u need to turn off the DNS, install the HDD, then turn on the DNS again - so in principle only the firmware matters in this case and blaming users for this bug is simply irresponsible. )

Well - I don't agree - again anecdotal, but why is it that the users who have more experience, like myself, do not appear to have experienced the problem?

If you check the forum - you will find posts reporting several different 1.04 beta releases, showing that development has been ongoing - can you say definitively that the problem was not fixed in an early beta release, and that other issues may have delayed the final release.

I'm going to leave you with one last thought - we, as users, are all guessing at what has been done, when it has been done and why it has been done.  Some of us have more or less experience and as such have different and in some cases unrealistic expectations.  If you think D-Link doesn't deliver what you feel reasonable, you have the option of buying another brand, good luck with whatever you choose.

dan.crouthamel wrote:

It's a never a bug ... it's always user error ..

I won't deny that bugs exist, but, I've learned the hard way, that user error occurs a lot more often than most users will admit - and remember I'm a user too.

If I might be allowed to go "off topic" briefly - as a very young field engineer (I was working with NCR at the time) I spent almost two weeks trying to track down an intermittent problem that affected both machines at one customer site, I never personally saw the problem and all the system diagnostics ran without error.

One day I went to site because the customer had called yet one more time and on entering the data center realized that frequently when I went to the site, that a particular operator was always present when I got there - I went to her supervisor and suggested that not only was this operator always present, but the problem always affected the machine she was at - the supervisor said she thought she had noticed it but that she was not certain.

Armed with that knowledge, I went back into the data center and stood behind the operator and watched her work - in less than 15 minutes, I had not only personally witnessed the problem, but could also tell how the operator was causing it.  I then asked her supervisor to join me so that she too could see for herself.

Back on topic - in some cases, bugs that show up in the real world, often don't show up in a test lab, because the people in the test lab are professionals who have been trained in a particular way and wouldn't do things the same way that an untrained end user would.

End users may also have a wider range of equipment, especially older equipment, than you would find in a typical lab - and finally - end users often have unrealistic expectations, for example, you can connect a DNS-323 to an old 10 mbps hub and get it to work, but the throughput will be way below the advertised estimates on the package.

Let's take this - format the wrong disk if the "new" disk is NTFS/FAT32 formatted scenario - as an example.  I have learned (again the hard way) not to use disks with data on them as replacements in a RAID array, I have had the wrong disk formatted (this was on a Novell NetWare system in the early days of NetWare).

Could the experience gained working with RAID over almost 18 years be the reason why I have not had my DNS-323 format the wrong disk?

If this is in fact the cause of the wrong disk format problem - and I'm not certain that it is - should it be considered a bug that needs fixing, or simply document it and post a warning to the users, not to create the scenario under which it's known to occur?

As an example - if you smoke whilst refueling your car you run the risk of an explosion, but has that been considered a bug that needs to be fixed (perhaps by reformulating the fuel - diesel has much less risk of explosion - or by redesigning the fuel dispensing mechanism)?  No - what has been done is to simply post "NO SMOKING" signs.

So - if you run the risk of formatting the wrong disk by inserting a disk with NTFS/FAT32, should it be fixed in the firmware or simply by posting a warning?

dickeywang - can you describe a set of circumstances which you can guarantee will, 100% of the time, cause the unit to format the wrong disk?  This is actually the first step into fixing the problem - identifying the actual conditions under which the problem will occur.

And - no - I was drinking a cup of coffee and suddenly the coffee starts to explode (I assume that's what you mean) is not to my mind a more accurate way to describe it.  The format of the disk does not spontaneously occur, it's triggered by a particular action, that of adding a second disk, along with a second, as yet to be identified condition -  just as the explosion is triggered by adding gasoline, along with a second condition which is the presence of a naked flame.

Last edited by fordem (2007-12-04 01:58:33)

I'm not asking you to do D-Link's job - I'm asking you to understand that to do the job takes time - and that the first step is to define the problem.

Intermittent problems take longer to locate - and I will define this problem as intermittent for no other reason than it is not 100% reproducible - ie, you cannot reproduce it at will.

You claim that it can happen if you insert a disk that has an existing NTFS or FAT32 partition - I will tell you right now, that I have done that without it formatting the wrong disk, so I know the DNS-323 will not format the wrong disk every time a disk with an existing NTFS or FAT32 partition is installed.

How many times has this happened to you?  Once?  Twice?  I'm guessing that you cannot even begin to describe the exact circumstances under which it happened - was it an NTFS partition or a FAT32 partition or did it have both?  These details may be critical to defining the problem and then locating the cause.

Another thing that you need to be aware of is how bugs are treated in a situation like this, especially the elusive ones - attention will be paid to the statistics - the number of complaints as a percentage of the units sold, if the numbers are low enough and the problem elusive enough, it will simply be ignored - the greater the number of complaints, and the easier it is to reproduce the more likely it is to be fixed.

Instead of complaining - put your shoulder to the wheel and push - if you help define the problem it's more likely to get fixed.

You think the gasolene and the naked flame thing is common sense?  Not every one agrees with you - the number of NO SMOKING signs in filling stations is testimony to that - and for what it's worth - I think wiping hard drives before reusing them is common sense also - I do it as a matter of habit.

By the way - I'm done with this discussion - yes - I'm disappointed by the delay in the release of updates, but I work in the industry and I understand the constraints.