Recovery of a corrupted wallet

2

I have a very old wallet.dat that I've put into my .bitcoin folder. When trying to run bitcoind, the service stops giving the following error

2019-06-11T18:14:02Z Using wallet directory /home/nom/.bitcoin
2019-06-11T18:14:02Z init message: Verifying wallet(s)...
2019-06-11T18:14:02Z Using BerkeleyDB version Berkeley DB 4.8.30: (April  9, 2010)
2019-06-11T18:14:02Z Using wallet wallet.dat
2019-06-11T18:14:02Z BerkeleyEnvironment::Open: LogDir=/home/nom/.bitcoin/database ErrorFile=/home/nom/.bitcoin/db.log
2019-06-11T18:14:02Z Renamed wallet.dat to wallet.dat.1560276842.bak
2019-06-11T18:14:02Z BerkeleyEnvironment::Salvage: Database salvage found errors, all data may not be recoverable.
2019-06-11T18:14:02Z BerkeleyEnvironment::Salvage: WARNING: Unexpected end of file while reading salvage output.
2019-06-11T18:14:02Z Salvage(aggressive) found no records in wallet.dat.1560276842.bak.
2019-06-11T18:14:02Z Error: wallet.dat corrupt, salvage failed
Error: wallet.dat corrupt, salvage failed
2019-06-11T18:14:02Z Shutdown: In progress...

Edit. Here is the content of the DB log

file unknown has LSN 1/151495, past end of log at 1/28 Commonly caused by moving a database from one database environment to another without clearing the database LSNs, or by removing all of the log files from a database environment Page 0: metadata page corrupted Page 0: could not check metadata page wallet.dat: DB_VERIFY_BAD: Database verification failed file unknown has LSN 1/151495, past end of log at 1/1080 Commonly caused by moving a database from one database environment to another without clearing the database LSNs, or by removing all of the log files from a database environment wallet.dat.1560341204.bak: DB_VERIFY_BAD: Database verification failed file unknown has LSN 1/151495, past end of log at 1/2500 Commonly caused by moving a database from one database environment to another without clearing the database LSNs, or by removing all of the log files from a database environment wallet.dat.1560341352.bak: DB_VERIFY_BAD: Database verification failed

Edit. Using db_dump I am able to see the content of the wallet file ( addresses, keys etc)

Is there any way I could debug this, perhaps recovering just part of the keys or any useful data from the DB itself? Any useful hints how to approach such recovery?

Jernej

Posted 2019-06-11T18:21:11.010

Reputation: 51

Using db_dump -p <wallet file>, do you see anything lines that contain ckey or key? Does the output have HEADER=END near the start of the output and DATA=END at the end? – Andrew Chow – 2019-06-13T23:28:18.000

Yes, there are plenty of lines with \0key entries and indeed, both HEADER/DATA entries appear. Also if it helps db_verify says the DB is fine. The wallet was created with a fairly old client (2010?) – Jernej – 2019-06-14T06:17:11.527

Answers

1

It sounds like your wallet.dat file is still intact, but some of the environment data has gotten corrupted (possibly due to moving wallet files).

Try doing

db_load -r lsn wallet.dat

This will reset the LSNs in the wallet.dat file which should cause the errors in the db.log to go away. Once you do that, delete the database/ folder and start Bitcoin Core.

Andrew Chow

Posted 2019-06-11T18:21:11.010

Reputation: 40 910

This seems to make the wallet load just fine, though I wonder if all keys were erased? Is this error msg expected? 22019-06-14T16:40:30Z [default wallet] Keys: 0 plaintext, 0 encrypted, 0 w/ metadata, 0 total. Unknown wallet records: 0 ? – Jernej – 2019-06-14T16:41:53.017

That doesn't sound good. It seems like all of the keys were erased, and that's not supposed to happen. If you do db_dump on the file again, do you see what you saw earlier? Do you still see lines beginning with key? – Andrew Chow – 2019-06-14T18:43:00.680

Yes I can still see that. As a matter of fact diff seems to be saying that the two dumps are equivalent. – Jernej – 2019-06-14T18:59:55.667

Asked: 2019-06-11T18:21:11.010

Viewed: 264 times

Active: 2019-06-14T14:03:15.677