Let's see if the latest version in 0.93-bugfix fixes any of this behavior. As usual, wipe the existing DB.
The outcome is no different to yesterdays unfortunately (the missing addresses are identical to yesterdays 0.93bugfix branch). Just finished re-checking that I was using the correct build. Edit: just tried running latest Armory 0.93 bugfix with Bitcoin Core 0.9.3, and all problems are resolved (all wallets have the balances I would expect) You mean you were trying with Core 0.10 and a dedicated chain folder? I'm confused as to why it would fail under these circumstances but I guess I have no choice but to see for myself. Thanks for your effort though, it makes my life easier at least =P
|
|
|
Let's see if the latest version in 0.93-bugfix fixes any of this behavior. As usual, wipe the existing DB.
|
|
|
During the upgrade it didn't show anything so I deleted everything and rebuilt. Still blank. 0 transactions 0 balances. I'm on Windows. I'm using custom directories. Maybe there's a bug there?
If you're pointing at the wrong block data folder. What top block is Armory displaying?
|
|
|
The results with current 0.93bugfix are not necessarily an improvement. No mined coins shown at all, for instance.
pre or post 100 conf? Do you get the ledger in the lobby with the other transactions? Can you see those mined coins looking at address ledgers from the wallet property dialog? all are post 100 conf, can't see any record of them at all. I've reinstated the 0.92.99.2 Db, it works fine still (Db is the only change) Hmm could I see a WO? Otherwise gimme your machine's spec and OS
|
|
|
I've just scanned a wallet with coinbase rewards and I see them fine.
|
|
|
The results with current 0.93bugfix are not necessarily an improvement. No mined coins shown at all, for instance.
pre or post 100 conf? Do you get the ledger in the lobby with the other transactions? Can you see those mined coins looking at address ledgers from the wallet property dialog?
|
|
|
Ok, I'll try it out
Thanks, much appreciated It does fail at block 161352 every time.
Thanks for the block upload. Unfortunately I am unable to reproduce your issue. With the 24 block files you provided, I scanned up to block 197431. Can you try 0.93-bugfix and let me know if the issue reappears? Also please make sure to delete existing DB files first.
|
|
|
My bad T_T
Yeah 0.93-bugfix, obviously
|
|
|
Yes, I tried manually deleting & then re-building the Db using 0.92.99.3. Select addresses from a given wallet appear, but others do not. I'm not sure how the culling is determined, there's no obvious pattern at a glance.
No such problems with 0.92.99.2
Mind pulling dev and trying with that? Again, delete your former DB.
|
|
|
Did you delete the DB from the previous version? They share the same file names so LMDB would append the new DB to the old one, and that would conflict.
|
|
|
- Dozens of tx's fail to appear once the Db/scan is complete
.3 very fast but transactions no longer show
play with the wallet filter, should display the transactions.
|
|
|
Right I forgot about these guys. That's what we used last time someone uploaded us some block files
|
|
|
Could you send me your first 14-15 block files? (I expect your first 165xxx blocks to be covered by that few files)
What's the best way to transfer files of that size? easiest to set up would be FTP or torrent I guess. XDCC/fserve over IRC isnt too hard either. Maybe Mega allows for several 128mb files?
|
|
|
This is very odd, I'd like to observe that issue first hand. Could you send me your first 14-15 block files? (I expect your first 165xxx blocks to be covered by that few files)
|
|
|
Yeah it should crash instead. That throw is handled too gracefully.
|
|
|
I notice a few things wrong there: first, that something goes wrong at block 161252. Second, that when the scanning process halts unexpectedly, the rest of the program acts like it completed successfully.
It is complaining that block 161352 is missing, and halted scanning from that point onwards. Can you wipe the entire dbdir again and confirm that it chokes at the same block height?
|
|
|
Hi everyone, just a quick update: goatpig has created and HDD_optimization branch in the repo, and my first test confirms that it is dramatically better. Once Core was synchronized, it took only about 45 minutes to both build and scan the databases. This is now dramatically faster than the original LevelDB-based implementation of previous versions.
If you are running from the repo, I encourage you to check out the branch and try it. Otherwise, wait a day or two and we'll wrap up this optimization with a bunch of the other bug fixes into a .3 testing release.
I've tried two rounds now of "rebuild and rescan databases," and both times the process appears to complete successfully but when it's done the transaction history either shows a subset of what it should show, or nothing at all. Trying again with a manually wiped dbdir. Hmm yeah, forgot to mention that you have to wipe your dbdir, either manually or through Help -> Rebuild and Rescan
|
|
|
If you're in linux, it's very quick: https://bitcoinarmory.com/building-from-source/If you are on Windows or Mac... not so easy, and I would recommend you just wait. Especially Windows -- it's a total mess to setup the dev/build environment in windows! You're saying it's easier to build on Mac than on Windows... my feelings are hurt!
|
|
|
lmdb lock file, not sure what they shove in there, haven't looked at that part of its source.
|
|
|
no we dont have any public chat rooms
|
|
|
|