Bitcoin Core fatal error

[TheNerd]

I'm running bitcoin-qt manually because I used to have a ton of trouble with bitcoind ran by Armory.
I was rebuilding databases when bitcoin-qt exited with fatal error and pointing to the debug.log file
and here is what it says:

2017-05-04 21:07:12 LoadBlockIndexDB: last block file info: CBlockFileInfo(blocks=129, size=131897238, heights=464381...464557, time=2017-05-01...2017-05-02)
2017-05-04 21:07:12 Checking all blk files are present...
2017-05-04 21:07:12 LoadBlockIndexDB: transaction index disabled
2017-05-04 21:07:13 LoadBlockIndexDB: hashBestChain=000000000000000005c7e0869610192c9ac4135649444f9f54e35940e268b011 height=409805 date=2016-05-02 01:37:47 progress=0.814528
2017-05-04 21:07:13 init message: Rewinding blocks...
2017-05-04 21:07:15 init message: Verifying blocks...
2017-05-04 21:07:15 Verifying last 6 blocks at level 3
2017-05-04 21:07:15 [0%]...[16%]...[33%]...[50%]...[66%]...[83%]...[99%]...[DONE].
2017-05-04 21:07:16 No coin database inconsistencies in last 7 blocks (8250 transactions)
2017-05-04 21:07:16  block index           11072ms
2017-05-04 21:07:16 init message: Loading wallet...
2017-05-04 21:07:16 nFileVersion = 130200
2017-05-04 21:07:16 Keys: 102 plaintext, 0 encrypted, 102 w/ metadata, 102 total
2017-05-04 21:07:16  wallet                   18ms
2017-05-04 21:07:16 mapBlockIndex.size() = 464558
2017-05-04 21:07:16 nBestHeight = 409805
2017-05-04 21:07:16 setKeyPool.size() = 100
2017-05-04 21:07:16 mapWallet.size() = 0
2017-05-04 21:07:16 mapAddressBook.size() = 1
2017-05-04 21:07:16 init message: Loading addresses...
2017-05-04 21:07:16 torcontrol thread start
2017-05-04 21:07:16 ERROR: ReadBlockFromDisk: Deserialize or I/O error - ReadCompactSize(): size too large at CBlockDiskPos(nFile=506, nPos=90495971)
2017-05-04 21:07:16 *** Failed to read block
2017-05-04 21:07:16 Loaded 63303 addresses from peers.dat  371ms
2017-05-04 21:07:16 init message: Loading banlist...
2017-05-04 21:07:16 init message: Starting network threads...
2017-05-04 21:07:24 Failed to connect best blockdnsseed thread start
2017-05-04 21:07:24 init message: Done loading
2017-05-04 21:07:24 msghand thread start
2017-05-04 21:07:24 addcon thread start
2017-05-04 21:07:24 opencon thread start
2017-05-04 21:07:24 net thread start
2017-05-04 21:07:24 tor: Thread interrupt
2017-05-04 21:07:24 scheduler thread interrupt
2017-05-04 21:07:24 torcontrol thread exit
2017-05-04 21:07:24 addcon thread interrupt
2017-05-04 21:07:24 msghand thread interrupt
2017-05-04 21:07:24 dnsseed thread interrupt
2017-05-04 21:07:24 opencon thread interrupt
2017-05-04 21:07:25 net thread interrupt
2017-05-04 21:07:25 Shutdown: In progress...
2017-05-04 21:07:25 StopNode()
2017-05-04 21:07:25 Shutdown: done

Im with Ubuntu 17.04, Armory 0.95 and bitcoin-core 13.2

[goatpig]

Looks like you're out of disk space

[TheNerd]

Yes for some reason my main disk switched to read-only mode while it was syncing. As a result DB is corrupted again and I have to start from scratch?
Thing is it takes few days every time and after few days crashes again.

[TheNerd]

Since the corruption is in the latest block, isn't there a way to rebuild with just them instead of rebuilding the whole thing for few days?

[goatpig]

You can get Core to check its own chain over. As for Armory, you have to rebuild the DB from scratch.

[TheNerd]

Yes that's what I'm doing wiht core. But if I delete even one block file it starts rebuilding the whole thing and that takes like 3 days on my i7 laptop.
Is there a way to not start from the beginning every time?


(I'm deleting the latest few block files when the db corrupts as advised in other thread here)

[goatpig]

Is there a way to not start from the beginning every time?

I don't know. You'd want to ask Core maintainers directly for this one.