|
September 02, 2017, 04:57:21 AM |
|
Hello, I'm not one to join forums or ask people for help, but here I am hat in hand. I'll get right to it.
I've been using Armory for about 2 years now. About 2 weeks ago I woke up, turned on my PC, fired up armory and within a couple minutes received an "armory needs to shut down" message underneath a message that says:
"BDM thread failed: The scanning process interrupted unexpectedly, Armory will now shutdown. If the error persists, you will have to rebuild and rescan your database. If rebuilding and rescaning did not fix the issue, contact support. Refer to your log file for more details on the error."
I've had DB issues before with core that I've been able to resolve, but this was new. Note that I was running an outdated version of core and Armory 93.3 up until now. Core is running fine, and there are no missing blocks that I can discern. Core's error logs are clean. However, here is the pertinent part of the armorycpplog:
-WARN - 1504323063: (..\BlockWriteBatcher.cpp:1133) Finished applying blocks up to 477500 -WARN - 1504323265: (..\BlockWriteBatcher.cpp:1133) Finished applying blocks up to 480000 -ERROR - 1504323416: (..\BlockWriteBatcher.cpp:980) No block in DB at height 481824 -INFO - 1504323418: (..\BlockUtils.cpp:1518) Scanned Block range in 29035s -INFO - 1504323418: (..\BlockUtils.cpp:1521) Finished loading at file 985, offset 6002112 -INFO - 1504323418: (..\BlockDataViewer.cpp:157) Enabling zero-conf tracking -ERROR - 1504323421: (..\BDM_mainthread.cpp:429) BDM thread failed: The scanning process interrupted unexpectedly, Armory will now shutdown. If the error persists, you will have to rebuild and rescan your database. If rebuilding and rescaning did not fix the issue, contact support. Refer to your log file for more details on the error.
Clearly Armory does not like block 481824 for some reason. Below I will go over what I've done thus far. That block number has changed during some of these attempts to fix, however it's been that number the most, and it's currently that number again after a complete reset.
Here's what's transpired since:
1) I simply restarted Armory. Same issue, not a fluke. 2) I click the 'rebuild and rescan DB on next start up'. Waited for Armory to do it's thing. Same issue. 3) I deleted the 'database' folder from the Armory directory. Same issue. 3a) I used a backup of bitcoin core from July 19th to reload recent blocks as block 481824 was on Aug 24th. Same issue. 4) I -reindex'd bitcoin core. (3 days later) Same issue. 5) I -rescan'd bitcoin core. Same issue. 6) I deleted and reinstalled all my wallets multiple times. No help. 7) I downloaded and installed Armory 96.1. No change, except the error logs provide a LOT less information :/... Namely, it only provides the following:
Log file opened at 15:22:54.000: A:\Armory\Data\armorycpplog.txt -ERROR - 15:59:41.203: (..\SocketObject.cpp:447) POLLERR error in readAndWrite -ERROR - 15:59:41.469: (..\StringSockets.cpp:359) POLLERR error in readAndWrite
8. I deleted bitcoin core and downloaded the newest version. (3 days later - today) Same issue. Same block number - 481824. This block number exists in core with no issue.
2017-09-01 10:10:42 UpdateTip: new best=000000000000000000cbeff0b533f8e1189cf09dfbebf57a8ebe349362811b80 height=481823 version=0x20000002 log2_work=86.98539 tx=249036369 date='2017-08-24 01:52:51' progress=0.990820 cache=199.4MiB(173876tx) 2017-09-01 10:10:43 UpdateTip: new best=0000000000000000001c8018d9cb3b742ef25114f27563e3fc4a1902167f9893 height=481824 version=0x20000002 log2_work=86.985426 tx=249038235 date='2017-08-24 01:57:37' progress=0.990823 cache=203.4MiB(177815tx) 2017-09-01 10:10:44 Pre-allocating up to position 0x1100000 in rev00975.dat 2017-09-01 10:10:44 UpdateTip: new best=000000000000000000daf7a26d903543377d5cdddb962077e58fd11212479eea height=481825 version=0x20000000 log2_work=86.985462 tx=249040201 date='2017-08-24 02:26:36' progress=0.990845 cache=204.3MiB(180318tx) 2017-09-01 10:10:46 Pre-allocating up to position 0x700000 in rev00976.dat
Nothing logically makes any sense. The error will occur in 1 of 2 places: 1) at about 1 or 2% of importing one of my wallets (with 93.3, with 96.1 it goes all the way to 99% before erroring out), or 2) if the wallet is already in the system, it will get to the end of the 'scanning transactions' phase of boot up and then throw the exception.
Could a wallet file be corrupted? I have four wallets on the system. I tried to send funds from one wallet to the other wallet and added a small fee of .0005. I didn't mind waiting a few days for it to go through. However, when this error first started happening I could see behind the error pop-ups the list of my transactions. The one at the top (the last one) was this transaction between the two wallets still with 0 confirmations, with no other transactions after it (and I KNOW there were transactions after it). However, even when I try to import one of the other wallets that wasn't involved with this transaction to a fresh install of Armory I still get the same error.
I haven't tried 96.1 with the new version of core yet. I'll download 96.2 and do that now, but I absolutely do not expect any different results. I'm honestly not even sure what anyone can advise me to do at this point if a clean install of both programs doesn't resolve the issue.
If it matters, these are watching-only wallets. I use full cold storage implementation. That said, when I go to my offline wallets to get the private keys, to put into Electrum as a last ditch effort to recover my funds, it doesn't have any. How screwed am I? Thanks for anyone's time who contributes, sorry for the long read.
|