Bitcoin Core, can't sync. Receiving wrong header/blocks?

[hmmkay]

Had a working version of the 0.9.3 Bitcoin Core.
Tried upgrading to 0.10 a few days ago and can't get a full node running.

Did a RAM test, and it didn't show any errors.
Tried going back to 0.9.3, getting errors. (with and without the bootstrap.dat - bootstrap itself doesn't show errors, it starts after getting blocks from the network)
Tried 4 times with 0.10, deleting and redownloading the blockchain, keep getting errors. Frustrating to say the least.

debug.log:

"2015-04-19 12:22:39 ERROR: ConnectBlock() : inputs missing/spent
2015-04-19 12:22:39 InvalidChainFound: invalid block=0000000000000004178e43c489f4372d899999d382ff4980a489ec5c1b9c4b50  height=268448  log2_work=73.645528  date=2013-11-07 21:21:08
2015-04-19 12:22:39 InvalidChainFound:  current best=0000000000000000cda3f044efb14dee09a13937aeca64f1b2b2a1905d094a1a  height=268447  log2_work=73.645313  date=2013-11-07 21:10:01
2015-04-19 12:22:39 ERROR: ConnectTip() : ConnectBlock 0000000000000004178e43c489f4372d899999d382ff4980a489ec5c1b9c4b50 failed
2015-04-19 12:22:39 InvalidChainFound: invalid block=0000000000000004178e43c489f4372d899999d382ff4980a489ec5c1b9c4b50  height=268448  log2_work=73.645528  date=2013-11-07 21:21:08
2015-04-19 12:22:39 InvalidChainFound:  current best=0000000000000000cda3f044efb14dee09a13937aeca64f1b2b2a1905d094a1a  height=268447  log2_work=73.645313  date=2013-11-07 21:10:01
2015-04-19 12:22:39 LoadExternalBlockFile: Processing out of order child 0000000000000000bb11ca9e12ea219f515f998b788b913367db21e06e8226e1 of 0000000000000004178e43c489f4372d899999d382ff4980a489ec5c1b9c4b50
2015-04-19 12:22:39 ERROR: AcceptBlockHeader : prev block invalid
2015-04-19 12:22:39 ERROR: ProcessNewBlock : AcceptBlock FAILED"

....

2015-04-19 12:42:40 ERROR: AcceptBlockHeader : block is marked invalid
2015-04-19 12:42:40 ERROR: invalid header received
2015-04-19 12:42:40 ProcessMessage(headers, 162003 bytes) FAILED peer=1
2015-04-19 12:42:42 ERROR: AcceptBlockHeader : block is marked invalid
2015-04-19 12:42:42 ERROR: invalid header received
2015-04-19 12:42:42 ProcessMessage(headers, 162003 bytes) FAILED peer=6
2015-04-19 12:42:43 ERROR: AcceptBlockHeader : block is marked invalid
2015-04-19 12:42:43 ERROR: invalid header received
2015-04-19 12:42:43 ProcessMessage(headers, 162003 bytes) FAILED peer=8
2015-04-19 12:42:43 ERROR: AcceptToMemoryPool : nonstandard transaction: non-final
2015-04-19 12:42:44 ERROR: AcceptBlockHeader : block is marked invalid
2015-04-19 12:42:44 ERROR: invalid header received
2015-04-19 12:42:44 ProcessMessage(headers, 162003 bytes) FAILED peer=4
2015-04-19 12:42:46 ERROR: AcceptBlockHeader : block is marked invalid
2015-04-19 12:42:46 ERROR: invalid header received
2015-04-19 12:42:46 ProcessMessage(headers, 162003 bytes) FAILED peer=2
2015-04-19 12:42:46 ERROR: AcceptToMemoryPool : nonstandard transaction: non-final
2015-04-19 12:42:53 ERROR: AcceptBlockHeader : block is marked invalid
2015-04-19 12:42:53 ERROR: invalid header received
2015-04-19 12:42:53 ProcessMessage(headers, 162003 bytes) FAILED peer=9
2015-04-19 12:42:57 ERROR: AcceptBlockHeader : block is marked invalid
2015-04-19 12:42:57 ERROR: invalid header received
2015-04-19 12:42:57 ProcessMessage(headers, 162003 bytes) FAILED peer=3
2015-04-19 12:43:08 ERROR: AcceptBlockHeader : block is marked invalid
2015-04-19 12:43:08 ERROR: invalid header received
2015-04-19 12:43:08 ProcessMessage(headers, 162003 bytes) FAILED peer=5

"

How is this possible? Why am I receiving wrong headers/blocks?
Who's sending these?

Any help is greatly appreciated.

[achow101_alt]

Does anything happen if you just run bitcoin core 0.10.0 for a while? If it doesn't give you an error message, then it should be syncing up fine. You might have just connected to a bad node, and core should resolve the issue itself.

[hmmkay]

Usually it slows down after a while. Going from 3MB/s to a few kb/s. Then I check the log and it shows it has a bad block somewhere and can't connect.

Since there is no rollback feature I'm aware off, the only thing to do is delete everything and start downloading again.

I've restarted the whole blockchain download...

Got a fatal error:

2015-04-19 16:22:48 ERROR: CheckProofOfWork() : nBits below minimum work
2015-04-19 16:22:48 ERROR: ReadBlockFromDisk : Errors in block header
2015-04-19 16:22:48 *** Failed to read block
2015-04-19 16:23:09 ERROR: ProcessNewBlock : ActivateBestChain failed
2015-04-19 16:23:09 ERROR: CheckProofOfWork() : nBits below minimum work
2015-04-19 16:23:09 ERROR: ReadBlockFromDisk : Errors in block header
2015-04-19 16:23:09 *** Failed to read block
2015-04-19 16:23:09 ERROR: ProcessNewBlock : ActivateBestChain failed
2015-04-19 16:23:09 ERROR: CheckProofOfWork() : nBits below minimum work
2015-04-19 16:23:09 ERROR: ReadBlockFromDisk : Errors in block header
2015-04-19 16:23:09 *** Failed to read block
2015-04-19 16:23:09 opencon thread interrupt
2015-04-19 16:23:09 addcon thread interrupt
2015-04-19 16:23:09 dumpaddr thread stop
2015-04-19 16:23:09 net thread interrupt
2015-04-19 16:23:10 ERROR: ProcessNewBlock : ActivateBestChain failed
2015-04-19 16:23:10 msghand thread interrupt
2015-04-19 16:23:10 Shutdown: In progress...
2015-04-19 16:23:10 StopNode()
2015-04-19 16:23:10 UPNP_DeletePortMapping() returned : 714
2015-04-19 16:23:10 upnp thread interrupt

[Amph]

try a different wallet.dat, otherwise there is something wrong with your os

there aren't many alternatives

this user had the same issue https://bitcointalk.org/index.php?topic=957509.msg10498465#msg10498465, try to contact him, perhaps he solved it

[hmmkay]

How come it worked before? Didn't change a thing. Blaming the OS seems a bit random.

I've always removed the entire folder, the standard wallet.dat is also removed.

[hmmkay]

Tried again, different error:

2015-04-19 17:14:23 opencon thread interrupt
2015-04-19 17:14:23 addcon thread interrupt
2015-04-19 17:14:23 dumpaddr thread stop
2015-04-19 17:14:23 msghand thread interrupt
2015-04-19 17:14:23 net thread interrupt
2015-04-19 17:14:23 Shutdown: In progress...
2015-04-19 17:14:23 StopNode()

Going to give 0.9.3 a try.

[Amph]

How come it worked before? Didn't change a thing. Blaming the OS seems a bit random.

I've always removed the entire folder, the standard wallet.dat is also removed.

maybe you have installed something new, some windows update or other stuff

Tried again, different error:

2015-04-19 17:14:23 opencon thread interrupt
2015-04-19 17:14:23 addcon thread interrupt
2015-04-19 17:14:23 dumpaddr thread stop
2015-04-19 17:14:23 msghand thread interrupt
2015-04-19 17:14:23 net thread interrupt
2015-04-19 17:14:23 Shutdown: In progress...
2015-04-19 17:14:23 StopNode()

Going to give 0.9.3 a try.

then the walle.dat had some blame...

[hmmkay]

Got many errors trying again. Also tried different disks to write to.
Right now writing to a non-raid disk. Seems to be going better than normal, fingers crossed.

About the wallet.dat, seems weird because a new one is generated at every install.

Is there an option to connect to 0.10 peers only?

[hmmkay]

"2015-04-19 22:37:32 ERROR: ConnectBlock() : inputs missing/spent
2015-04-19 22:37:32 Misbehaving: 73.39.224.201:8333 (0 -> 100) BAN THRESHOLD EXCEEDED
2015-04-19 22:37:32 InvalidChainFound: invalid block=0000000000000000114635fb382a473713ccc8e23d10006b699546309252666c  height=316053  log2_work=80.228102  date=2014-08-17 09:07:49
2015-04-19 22:37:32 InvalidChainFound:  current best=000000000000000028d040303a07a81dbd05e7d3a6fcc90ca4ee40b03d0b24a6  height=316052  log2_work=80.228016  date=2014-08-17 08:55:14
2015-04-19 22:37:32 ERROR: ConnectTip() : ConnectBlock 0000000000000000114635fb382a473713ccc8e23d10006b699546309252666c failed
2015-04-19 22:37:32 InvalidChainFound: invalid block=0000000000000000114635fb382a473713ccc8e23d10006b699546309252666c  height=316053  log2_work=80.228102  date=2014-08-17 09:07:49
2015-04-19 22:37:32 InvalidChainFound:  current best=000000000000000028d040303a07a81dbd05e7d3a6fcc90ca4ee40b03d0b24a6  height=316052  log2_work=80.228016  date=2014-08-17 08:55:14
2015-04-19 22:37:32 ERROR: AcceptToMemoryPool : nonstandard transaction: non-final
2015-04-19 22:37:32 Pre-allocating up to position 0x2000000 in blk00168.dat
2015-04-19 22:37:33 ERROR: AcceptToMemoryPool : nonstandard transaction: non-final
2015-04-19 22:37:33 ERROR: AcceptToMemoryPool : nonstandard transaction: non-final
2015-04-19 22:37:41 ERROR: AcceptToMemoryPool : nonstandard transaction: non-final
2015-04-19 22:37:46 ERROR: AcceptToMemoryPool : nonstandard transaction: non-final
2015-04-19 22:37:55 ERROR: AcceptToMemoryPool : nonstandard transaction: non-final
2015-04-19 22:38:23 receive version message: /Satoshi:0.9.2.1/: version 70002, blocks=352835, us=62.163.255.170:52001, peer=42"

So we've got a misbehaving node at 73.39.224.201, anyone care to find out who/where this is?
After downloading 21gb, for the 5th time....nice

[hmmkay]

Trying again with a bitcoin.conf file. Added 0.10 nodes only.

Bitcoin, so easy....

[hmmkay]

Nothing was wrong on my end, there seems to be malicious nodes out there.
A bit disappointing that I had to figure this out myself and nobody gave any good advice.

Anyway, to fix this, add a bitcoin.conf to your bitcoin directory.
https://en.bitcoin.it/wiki/Running_Bitcoin#Bitcoin.conf_Configuration_File

With 'addnode' or 'connect' you can specify which nodes bitcoin core should connect to.
Find good nodes here:
https://getaddr.bitnodes.io/

[goregrind]

Had the exact same problems. After redownloading the blockchain about 3 times I had enough with bitcoin core and used electrum

[gmaxwell]

Nothing was wrong on my end, there seems to be malicious nodes out there.
A bit disappointing that I had to figure this out myself and nobody gave any good advice.

You're asking the wrong question. There is nothing a malicious node can do to harm you, you should not have to manually control who you're connecting to, and doing so with connect is likely to leave you constantly dysfunction unless you're connecting to nodes you control on your own.

Your blockchain data was becoming corrupted; this doesn't have anything to do with your peers; its either due to your hardware or a software bug. You do not generally need to redownload to recover from corruption, reindex does that. Unfortunately there is a bug in 0.10.0 where reindex can cause corruption; it's fixed in 0.10.1rc available at https://bitcoin.org/bin/bitcoin-core-0.10.1/test/  and which will be made a full release in a few days (sooner if there are success reports from people using it).

[hmmkay]

Small update:

Although I got it working in the end with the bitcoin.conf and addnodes, another factor could have been my ISP.

Online games showed heavy packetloss every now and then, ran 'Pingplotter' and packetloss showed up to a few seconds of 100% loss every now and then. So I switched ISP last month and everything has been working fine, without addnodes/bitcoin.conf.

I'm not sure if it's related to the problem I had before, it would seem bitcoin would handle less than ideal connections as well.
But for future refence in case people have similar problems, you might want to check your connection in general.

[ProfitCenter]

I have been indexing again, and again.
It took me already seven nights to reindex the database....
One hour ago with 1 year indexing remaining the logs shows this
"LoadExternalBlockFile: Processing out of order child" on several lines.

I had to start the indexing again...

This is really frustrating, especially I bought some bitcoins one week ago, and they still don't show up