• Forum has been upgraded, all links, images, etc are as they were. Please see Official Announcements for more information

PROVEN - client is corrupting wallets upon exit

Status
Not open for further replies.
Oh maybe make a backup first?
Make a backup of a file that was corrupted before I ever got a chance to make a backup? Do you even read?

CAN'T COPY WALLET.DAT WHILE DARKCOIN-QT IS MEMORY RESIDENT!
HAVE TO EXIT TO MAKE A BACKUP!
CORRUPTED UPON EXIT!
MAKE A BACKUP OF WHAT?

Can you read?

I'll say it again: I should have exported the key before exiting. That's the only backup I can be sure of.
 
No. And doing that won't prove anything. Pay attention.

This happened once before on a much, much older version.

Another Straw Man, because I've said this is the second time, and the first time we are all well aware of. I just let it slide because all the pumpers were throwing a fit about it. I even believed them when it was suggested the problem was elsewhere. Propulsion recovered it for me. I was told by two different people that it was caused by two different things that I did not do, and that it was my fault, even thought I did not do what they said I did to cause it, and only one of them could have been true... So, neither of them were right about the problem, or that I caused it.

Same problem, same symptoms.

Same price pumper trolls mad that I'm talking about it, skipping over useful info, pretending not to understand, etc...

Will try the dump, had no idea that even existed.
"What happen's in Vegas stay's in Vegas..."

What happen's in old client versions on old wallet.dat files really has no bearing on what's happening on a current wallet client on a brand newly created wallet.dat file. When you first described your issue you mentioned creating 8 brand new wallet.dat files with a brand new 10.15.16 client. You mentioned the 8 were created fine and it was then the 9th one that was corrupted. Since this was all being done on fresh wallet.dat files not ones that were being upgraded from previous version's of software then it should be entirely reproducible. We aren't talking about a 10.13.15 wallet.dat file that has had ds+ run on it stopped copied to a new 10.15.14 version then had more ds+ run on it then upgraded to 10.15.16 and opened and closed. We are just talking about a vanilla wallet being created. I don't know when that code was last touched but I bet it hasn't been touched in many many moons.
 
Make a backup of a file that was corrupted before I ever got a chance to make a backup? Do you even read?

CAN'T COPY WALLET.DAT WHILE DARKCOIN-QT IS MEMORY RESIDENT!
HAVE TO EXIT TO MAKE A BACKUP!
CORRUPTED UPON EXIT!
MAKE A BACKUP OF WHAT?

Can you read?

I'll say it again: I should have exported the key before exiting. That's the only backup I can be sure of.
I was talking about making a backup of hte corrupted wallet.dat file before you try dumping it to a human readable error. Just trying to save you from having your next comment be, "I corrupted it even worse now and forgot to make a backup before I tried the db dump"
 
I read your topic writes "PROVEN".. yet you haven't pasted any real log of anything.

You seem to be the only one having this very error. I tried replicating on different linux distributions and hardware, no dice, works every time perfectly.

Also, when loading a wallet with 1000 Dark.. might be somewhat a bit intelligent to keep it's private key somewhere, just in case you know...
 
What happen's in old client versions on old wallet.dat files really has no bearing on what's happening on a current wallet client on a brand newly created wallet.dat file.
You're suggesting that I ignore that which the evidence most clearly points out? Damn, your defense of that growing army of Straw Men knows no limits...

Thanks for the db dump info tho, will give it a shot.
 
I read your topic writes "PROVEN".. yet you haven't pasted any real log of anything.

You seem to be the only one having this very error. I tried replicating on different linux distributions and hardware, no dice, works every time perfectly.

Also, when loading a wallet with 1000 Dark.. might be somewhat a bit intelligent to keep it's private key somewhere, just in case you know...

I've come to accept that fact that even on this forum, pumper trolls have no shame, and pretty much everyone is a pumper troll... No better than BCT.
 
I'll say it again: I should have exported the key before exiting. That's the only backup I can be sure of.

1: Run wallet.
2: Quit wallet.
3: Make copy of wallet.dat
4: Start wallet again. wallet.dat corrup? Delete it and Goto 1:
5: Quit wallet
6: make a diff of your backup wallet.dat with the actual one
7: diff shows a difference (means one of your wallet.dat files is corrupt: Goto 1:
8: diff shows no differnce -> voila

But you're right, exporting the key is the more straightforward approach.
 
I tried replicating on different linux distributions and hardware, no dice, works every time perfectly.

Precisely why I got sloppy, then it bit me...

Seems to be fine for months now, wrote it off as a fluke with my system, not the client...

It happened again. Enjoy beating up the messenger, he doesn't back down. You'll only hurt your fists...
 
Bad ram chips in that machine? If it's unreproducible in software perhaps it's hardware related?
 
Yet you don't show any *proof*, I'd just remove the "PROVEN" in your title, it is misleading.
I showed you everything I've got. The log file ends abruptly. The file shrinks. There is nothing more. That's part of the problem. Hard to track down a problem with so little to go on.

Demanding more when there is no more does not justify the agenda of pumper troll.

Something is causing it to exit abnormally. Very abnormally. So abnormally that it just plain dies with no further output. It's just fucking gone and it fucks the wallet.dat as it goes. I suspect some bit of code that flushes upon exit. Opens wallet to [do something for exiting] write part, total fail out of RAM entirely. Since the file is a convenient size difference, 512 to 488, exactly 24k... Maybe it's some kind of header-footer formatting, and the bulk is flushing fine? Data in chunks/categories, the bug hunk (probably keys and addresses) seems to write, but then something else doesn't make it and the exception is not trapped?

Ignoring a problem just because it is difficult and rare is not a plan for success. I let it slide once. Not this time. Especially not now that it has repeated itself identically.
 
1: Run wallet.
2: Quit wallet.
3: Make copy of wallet.dat
4: Start wallet again. wallet.dat corrup? Delete it and Goto 1:
5: Quit wallet
6: make a diff of your backup wallet.dat with the actual one
7: diff shows a difference (means one of your wallet.dat files is corrupt: Goto 1:
8: diff shows no differnce -> voila

But you're right, exporting the key is the more straightforward approach.
That's how I do all my new wallets, it's repetitive but it means I can keep an eye on my shit.

I had one QT corrupt on me once, lost 30 DRK, I renamed a wallet.dat whilst I still had the QT running, it corrupted after that. Kind of my own fault...

That's the only time I've had an issue with it corrupting...
 
Never said to ignore the issue. You simply state that its PROVEN, while the only claim is your own experience in your own words.

Were you able to replicate it on the *same* machine? This is likely an operation relying mostly on ram, as coingun said it could be a bad stick.
I have doubts about this, but you could be out of memory, ending up with the system killing the wallet at some point during it's shutdown?
 
I showed you everything I've got. The log file ends abruptly. The file shrinks. There is nothing more. That's part of the problem. Hard to track down a problem with so little to go on.

I understand your anger, but a bit more information would help tracking this one down tremendously:

- which operating system/version
- fresh install or did you already install other software packets. If so,which ones
- did the computer and other software ran stable before or were there already other crashes

With this information it would be at least possible to find someone with a similar configuration and try to reproduce it.

Edit:eth1 made a good point, USB sticks ARE very insecure, I had already a couple of them doing nasty things...
 
Never said to ignore the issue. You simply state that its PROVEN, while the only claim is your own experience in your own words.

Were you able to replicate it on the *same* machine? This is likely an operation relying mostly on ram, as coingun said it could be a bad stick.
I have doubts about this, but you could be out of memory, ending up with the system killing the wallet at some point during it's shutdown?

That's a very convenient piece of busted RAM... It manages to cause a problem with wallet.dat of darkcoin-qt 2 times in exactly the same way, and the system runs perfectly stable in every other way...

Oh, and no, it's a different machine entirely. So, theory moot.
 
I understand your anger, but a bit more information would help tracking this one down tremendously:

- which operating system/version
- fresh install or did you already install other software packets. If so,which ones
- did the computer and other software ran stable before or were there already other crashes

With this information it would be at least possible to find someone with a similar configuration and try to reproduce it.

Well, it's not common to include things that don't happen... It didn't rain today. Weird, right?

If I came out saying "this system has months of uptime and is rock solid" you'd just accuse me of trying to cover up in advance... Why would someone mention that?

Already stated, Xubuntu 14.04/0.10.15.16 darkcoin-qt - client version likely irrelevant because this happened once before in exactly the same way several months ago.

Can we please focus instead of search for other things to blame that aren't real/involved?
 
Just trying some ideas. As far as I know, you don't know what caused it either right ?

Also, please stop being an ass with the people in here, we are simply trying to help.
 
I understand your anger, but a bit more information would help tracking this one down tremendously:
eth1 made a good point, USB sticks ARE very insecure, I had already a couple of them doing nasty things...
All media currently involved is magnetic. 2x USB hard drives, internal hard drive...

Hardware is not the problem... I'd have other issues. I don't. Which is why I didn't mention having any. I don't have any to mention...
 
camosoul you have to understand that you are not giving us anything to work with. And what you did provide was tried (by me even) and unable to reproduce.

You come here, scream proven, yet there is no reliable way to reproduce it. This all does sound like something that is shitting itself on your end. So until someone
can recreate it or you come up with a reliable way of doing so, arguing here is moot.

This thread got way of topic already, and you should probably rename it to "possible that client is corrupting wallet - need more help testing" and stop being an ass.


Just re-read your fist post, seems you are running the darkcoin-qt from command line, how did you close the client? Is it safe for us to assume you have not used ctlr+c? Cuz that would be just silly
 
Last edited by a moderator:
Status
Not open for further replies.
Back
Top