Pages:
Author

Topic: overflow bug SERIOUS (Read 63969 times)

founder
Activity: 364
Merit: 7065
August 16, 2010, 05:54:55 PM
#68
Un-upgraded nodes have the correct chain most of the time, but they are still trying to include the overflow transaction in every block, so they're continually trying to fork and generate invalid blocks.  If an old version node is restarted, its transaction pool is emptied, so it may generate valid blocks for a while until the transaction gets broadcast again.  0.3.9 and lower nodes still must upgrade.

The SVN now has the code we needed to automatically reorg the block chain without having to delete the blk*.dat files manually.  I knew I couldn't write that code fast and carefully enough yesterday, so I went with the quick manual option.
sr. member
Activity: 252
Merit: 268
August 16, 2010, 05:15:32 PM
#67
So at this point, would a client that has not upgraded also have the correct chain?
Yes. The only exception would be if an unpatched client made another fake transaction and managed to verify it. They would be able to spread that fake block to a few other nodes that haven't upgraded, but the upgraded clients seem to have more power than the unpatched clients, so the bad link would not last long and would not be spread by upgraded clients.
full member
Activity: 199
Merit: 2384
August 16, 2010, 05:13:00 PM
#66
So at this point, would a client that has not upgraded also have the correct chain?
full member
Activity: 158
Merit: 100
August 16, 2010, 09:35:10 AM
#65
Would it be useful to have RPC command to chop off the tail of the chain manually,
instead of removind and redownloading the whole chain?
founder
Activity: 364
Merit: 7065
August 16, 2010, 07:59:38 AM
#64
It looks like we overtook the bad chain somewhere around 74689.  0.3.9 and lower nodes have been responding with the current block number for some hours now.

That means it's no longer necessary to delete blk*.dat before upgrading.  You can just upgrade and it'll reorg away the bad block chain.

Thanks to everyone for the quick response!
member
Activity: 111
Merit: 10
August 16, 2010, 12:07:50 AM
#63
Question about fallout:  I had a transaction that I submitted after the bad block, using the bad block chain.

What is the status of that transaction?
From what I can tell, my (updated) sending client wallet shows the deducted amount.

Will it get reincorporated into the fixed chain, and will the recipient be able to spend it?
Right, it will get reincorporated into the fixed chain.  The transaction won't disappear, it'll still be visible on both sides, but the confirmation count will jump back to 0 and start counting up again.

It's only if you generated a block in the bad chain after block 74638 that the 50 BTC from that will disappear.  Any blocks in the bad chain wouldn't have matured yet.

Interesting.. fascinating to watch this work its way through the bowels of the system.
It was block 73746 that I generated.
    CTxOut(nValue=50.00000000, scriptPubKey=0x8DDD5C7DADB2D43AC5F186)
08/12/10 02:35 generated 50.00

And I sent 49.00 of it to 19Nzg21hQQDAY5GTdTTuUVPA4MaE7AusXz (using the broken chain)

Now I'm waiting for that node to figure out it was received, and it's using the new chain.
sr. member
Activity: 308
Merit: 258
August 15, 2010, 10:57:39 PM
#62
The bad chain is also slowed down as more nodes upgrade.

We've already generated 14 blocks since 74638.  The builds of 0.3.10 were uploaded about 2 and 3 hours ago.  Of the nodes I'm connected to, more than half are already 0.3.10.  I would say we probably already have more power than the bad chain.

I think it'd probably be a good idea still to come out with another version that rejects connections from older versions - otherwise the network might remain rather fragmented for a while. :/
We want to keep the older clients connected so that when the correct chain overtakes the incorrect chain, they will switch back to the correct chain. Although I don't know the specifics of how far back in the chain the old chain will accept a branched chain.
The old clients should accept it all the way back to the last snapshot of release, so this being found so quickly and so long after the last release, it should work in theory. Here's a good test of the theory  Grin
sr. member
Activity: 252
Merit: 268
August 15, 2010, 10:09:52 PM
#61
The bad chain is also slowed down as more nodes upgrade.

We've already generated 14 blocks since 74638.  The builds of 0.3.10 were uploaded about 2 and 3 hours ago.  Of the nodes I'm connected to, more than half are already 0.3.10.  I would say we probably already have more power than the bad chain.

I think it'd probably be a good idea still to come out with another version that rejects connections from older versions - otherwise the network might remain rather fragmented for a while. :/
We want to keep the older clients connected so that when the correct chain overtakes the incorrect chain, they will switch back to the correct chain. Although I don't know the specifics of how far back in the chain the old chain will accept a branched chain.
hero member
Activity: 574
Merit: 513
August 15, 2010, 09:50:00 PM
#60
4b.  Currently the network has a combination of good and bad nodes (those running versions older than 0.3.10).  If you only connect to nodes running old versions, you will be stuck at 74638 blocks.  You can fix this by running the client with the options "-addnode=75.158.131.108 -addnode=99.27.237.13 -addnode=76.235.234.64 -addnode=74.137.15.169 -addnode=68.68.99.14"

http://www.bitcoin.org/wiki/doku.php?do=show&id=74638_nodes
full member
Activity: 221
Merit: 102
August 15, 2010, 09:39:55 PM
#59
The bad chain is also slowed down as more nodes upgrade.

We've already generated 14 blocks since 74638.  The builds of 0.3.10 were uploaded about 2 and 3 hours ago.  Of the nodes I'm connected to, more than half are already 0.3.10.  I would say we probably already have more power than the bad chain.

I think it'd probably be a good idea still to come out with another version that rejects connections from older versions - otherwise the network might remain rather fragmented for a while. :/
founder
Activity: 364
Merit: 7065
August 15, 2010, 09:38:21 PM
#58
On Windows, findstr /c:"version message" debug.log

It looks like the bad chain was on block 74678 recently.  Can't wait to overtake it.

On the stats at http://nullvoid.org/bitcoin/statistix.php  there's been 5 blocks per hour in the last 3 hours.  We had a difficulty adjustment about a day ago that should have put it back to 6 blocks per hour.
full member
Activity: 210
Merit: 104
August 15, 2010, 09:28:57 PM
#57
... Of the nodes I'm connected to, more than half are already 0.3.10. ...
That's great news! How did you figure that out, though? I was considering writing a client specially tailored to put itself out there and try to get as many people to connect as possible, then survey their versions and write out a log (in addition to all the normal bitcoin stuff).
sr. member
Activity: 308
Merit: 258
August 15, 2010, 09:22:05 PM
#56
I'm chucking as much CPU at this as I can.  Yeah, bit of an unfair advantage I guess until everyone upgrades.  Lips sealed
My wife's PC already generated 2 of the new blocks (LOL), luck is on her side I guess.
founder
Activity: 364
Merit: 7065
August 15, 2010, 09:16:10 PM
#55
The bad chain is also slowed down as more nodes upgrade.

We've already generated 14 blocks since 74638.  The builds of 0.3.10 were uploaded about 2 and 3 hours ago.  Of the nodes I'm connected to, more than half are already 0.3.10.  I would say we probably already have more power than the bad chain.
member
Activity: 111
Merit: 10
August 15, 2010, 09:02:07 PM
#54
Ya know, in a way, this actually temporarily decreases the difficulty to generate blocks until everyone upgrades. Yeah, it'll be slower, but that just gives each client more time to try to generate the difficult hash.

Of course, this happened right after the difficulty kicked up to 511.7, so all these new "good" blocks will have to contend with that.  And a good chunk of the network grind power is off on the wrong chain, so new blocks may take a lot longer than 10min to find.  At what point will the balancing figure this out and possibly kick the difficulty back down again?
founder
Activity: 364
Merit: 7065
August 15, 2010, 08:12:05 PM
#53
Most people running clients are not reading this message thread.  So...  Silly questions:

1) How will this continue to affect version 3.8.1 (pre-catastrophe) clients with bad block chain?
2) How will this affect clients that upgrade to 3.8.10 but don't remove their block chain files?
1) Once more than 50% of the node power is upgraded and the good chain overtakes the bad, the 0.3.10 nodes will make it hard for any bad transactions to get any confirmations.
2) If you didn't remove your blk*.dat files, you're not helping to contribute to that 50%, and you'll still show bad transactions until the good chain overtakes the bad chain.
newbie
Activity: 15
Merit: 0
August 15, 2010, 08:06:13 PM
#52
I added the addnode options and info about the 'stuck at 74638 blocks' issue at https://bitcointalksearch.org/topic/m.9595
newbie
Activity: 32
Merit: 0
August 15, 2010, 08:02:35 PM
#51
Most people running clients are not reading this message thread.  So...  Silly questions:

1) How will this continue to affect version 3.8.1 (pre-catastrophe) clients with bad block chain?
2) How will this affect clients that upgrade to 3.8.10 but don't remove their block chain files?
founder
Activity: 364
Merit: 7065
August 15, 2010, 08:02:24 PM
#50
I did all steps, now my client is 0.3.10 and it stopped at block 74638. Is all fine?
If you still show 74638 blocks then you aren't connected to any 0.3.10 nodes.  

For today, try adding these parameters: 
-addnode=75.158.131.108 -addnode=99.27.237.13 -addnode=68.68.99.14

See
https://bitcointalksearch.org/topic/sfg-828
founder
Activity: 364
Merit: 7065
August 15, 2010, 08:00:45 PM
#49
Question about fallout:  I had a transaction that I submitted after the bad block, using the bad block chain.

What is the status of that transaction?
From what I can tell, my (updated) sending client wallet shows the deducted amount.

Will it get reincorporated into the fixed chain, and will the recipient be able to spend it?
Right, it will get reincorporated into the fixed chain.  The transaction won't disappear, it'll still be visible on both sides, but the confirmation count will jump back to 0 and start counting up again.

It's only if you generated a block in the bad chain after block 74638 that the 50 BTC from that will disappear.  Any blocks in the bad chain wouldn't have matured yet.
Pages:
Jump to: