Gold collapsing. Bitcoin UP.

[adamstgbit]

@Christoph Bergmann
how did you mange to do that!?

 

[Justus Ranvier]

validating signature is one of the most CPU expensive operation

Do you plan to experiment with GPU acceleration for that part at all?

 

[Christoph Bergmann]

don't know. It was stucked at block 374 963, I restarted, got the message, go back to satoshis announcement. Fuck it.

maybe my hdd is too old, too cheap, too often overridden ...

friday afternoon - cheers'

 

[Richy_T]

I believe thats already in effect sort of, bitcoin uses a gossip network to relay TX across the network efficiently. altho gavin speculated there were some inefficiencies in its impl.

If the ratio of block traffic to non-block traffic is 12:88, I'd say that's pretty inefficient.

 

[jl777]

Do you plan to experiment with GPU acceleration for that part at all?

Not sure it is worth the effort for me to do as it will just crunch away in the background.

Hopefully it wont take 20 hours to verify all the sigs! If it does, then I will have to look into something

 

[Justus Ranvier]

Hopefully it wont take 20 hours to verify all the sigs! If it does, then I will have to look into something

The blockchain isn't getting any smaller...

Who knows, maybe some day Bitcoin will be doing so well that you need GPU acceleration just to keep up with the transactions on the wire?

 

[adamstgbit]

don't know. It was stucked at block 374 963, I restarted, got the message, go back to satoshis announcement. Fuck it.

maybe my hdd is too old, too cheap, too often overridden ...

friday afternoon - cheers'

your hdd is an unlikely culprit....
when you started syncing you started from 0? or some old blockchain from 2013??
did you stop and start the app while it was syncing? ( maybe the code does not gracefully shutdown? )
did the app crash at any point during the syncing process?

 

[Richy_T]

don't know. It was stucked at block 374 963, I restarted, got the message, go back to satoshis announcement. Fuck it.

That number rings a bell. I think when I has a Raspberry Pi node, it got stuck on that block. Unfortunately, the filesystem took a dump and it wouldn't boot before I could look much into it.

 

[adamstgbit]

that block is ~100 days ago
maybe it was that block with the spam that took a miner 5mins to validate?
on a home computer it could maybe take hours to validate? ( might not even be doable without enough RAM? )
if you thought the program was frozen and then force quit ( maybe windows thought it was frozen and forced quit the app itself ...)
if so its likely that it forced-quit during a write_to_hdd operation
and thats how you corrupt the DB.

 

[jl777]

The blockchain isn't getting any smaller...

Who knows, maybe some day Bitcoin will be doing so well that you need GPU acceleration just to keep up with the transactions on the wire?

The past blockchain size issue is solved via iguana parallel sync and with a 1MB block, there is no issue with keeping up. So until 2019 there is no problem

but even with 10x tx capacity increase via interleaved block, I still see no problems to complete the validation in time. Though I wrote about using GPU to parallel process the time consuming RPC calls, like listunspent in the iguana childboard

 

[Erdogan]

hmm ... now I'm at block 374 963 and it does nothing since 30 min. 0% cpu. Anybody know what to do?

@jl777 Good idea. So the user can use his client as a wallet and a node and can validate old transactions when he has time to do.

Take a look in the debug log, and see if there is a clue.

 

[adamstgbit]

The past blockchain size issue is solved via iguana parallel sync and with a 1MB block, there is no issue with keeping up. So until 2019 there is no problem

but even with 10x tx capacity increase via interleaved block, I still see no problems to complete the validation in time. Though I wrote about using GPU to parallel process the time consuming RPC calls, like listunspent in the iguana childboard

can we get a short and sweet " what is iguana "

am i right in assuming its a test program to see how fast you can validate the blockchain using GPUs?

google tells me " what is iguana" is a iguana


oh wait i found it
https://bitco.in/forum/threads/iguana-parallel-sync-full-btc-blockchain-in-30-minutes-uses-half-the-space-but-starts-up-instant.911/
[doublepost=1458352468,1458351564][/doublepost]"sync full BTC blockchain in 30 minutes,"
are you kidding me?!?!

i still dont fully get what you have done here, i dont have 8cores, if i used this method of syncing and "validating 2000Blocks at a time?" would i see a huge improvement in the time that it takes to sync?

why are you saying until 2019 ?

 

[jl777]

2019 is when 2MB hardfork will be mainnet

on the parallel sync, it is mostly bandwidth limited:

03:08:34 52572.27 174.25
03:08:44 55078.78 237.92
03:08:54 40423.29 160.13
03:09:04 55779.03 216.22
03:09:14 61821.66 191.29
03:09:24 62304.47 223.34
03:09:34 36312.58 155.32
03:09:44 52449.76 238.48
03:09:54 45422.25 259.60
Time eth0
HH:MM:SS KB/s in KB/s out

03:10:04 47707.93 285.00
03:10:14 25276.91 155.40
03:10:24 39874.62 287.68
03:10:34 61400.63 342.74
03:10:44 42533.16 256.06
03:10:54 20870.65 136.98
03:11:04 59232.55 424.25
03:11:14 59525.28 399.82
03:11:24 54570.41 353.84
Time eth0
HH:MM:SS KB/s in KB/s out
03:11:34 24594.03 167.24
03:11:44 52450.59 330.78
03:11:54 60753.00 359.81
03:12:04 64870.80 394.11
03:12:14 66039.66 374.83
03:12:24 50707.53 276.17
03:12:34 62222.98 338.61
03:12:44 65961.79 337.86
03:12:54 45313.98 232.43

above is 10 second samples of download/upload bandwidth of the blockchain data.
I am still optimizing things as there are some annoying dropouts and idle time, well idle relative to 800% CPU usage and 50MB/sec

Now for a normal home connection of 20mbps, it certainly cant download the 65GB in 30 minutes. But I plan to make the readonly dataset that is compressed to less than 20GB available via bittorrent. so that wont even use any bandwidth of relay nodes and at 20 mbps would take several hours

sig validation happens at the very end and would add another 35GB that is needed to download so I think most users will just choose to rely on checkpoints

the latest versions seems to have slowed things down some:
v.72/72 (2000 1st.82) to 201 N[202] Q.129 h.402000 r.158000 c.0.000kb s.343051 d.137 E.137:434485 M.403290 L.403291 est.128 0.000kb 0:26:13 176.375 peers.127/256 Q.(0 0)

only at block 342051 after 26 minutes, but I am getting things in place so I can combine the current realtime blocks with the readonly data. when that works, I can speed it up again

 

[jl777]

improved it a bit:

v.158/169 (1995 1st.182) to 189 N[202] Q.23 h.402000 r.362000 c.0.000kb s.376752 d.182 E.182:432530 M.403302 L.403303 est.8 0.000kb 0:30:00 590.699 peers.125/256 Q.(0 0)

well actually a lot as 182 bundles are almost as far as block 376752. I think I might know why the dropout. Have to resync from scratch again. back in 30 minutes

 

[adamstgbit]

Wowzers!
I have no doubt this tech will get adopted by the reference implementation ( i just dont like saying core lol) sooner than later.

 

[jl777]

I have been repeatedly told it is impossible to do an independent core, plus everytime I mention something iguana does, I am chastised for saying it, because the reference implementation already does everything that is good.

took a few tries, but latest result is:

06:04:23 18608.43 430.93
06:04:33 28280.53 474.10
06:04:43 40975.05 283.98

Time eth0
HH:MM:SS KB/s in KB/s out
06:04:53 31281.95 177.31
06:05:03 10635.75 173.26
06:05:13 14985.52 160.32
06:05:23 4990.17 190.72
06:05:33 25855.66 380.26
06:05:43 35324.08 165.18
06:05:53 20279.60 189.87

Time eth0

HH:MM:SS KB/s in KB/s out
06:06:03 31138.11 194.71
06:06:13 20247.69 221.65
06:06:23 7881.91 147.81
06:06:33 22620.91 175.68
06:06:43 11464.06 148.74
06:06:53 40422.45 220.33
06:07:03 46758.93 306.15

Time eth0
HH:MM:SS KB/s in KB/s out
06:07:13 32445.86 141.49
06:07:23 22955.99 142.55
06:07:33 39299.35 215.33
06:07:43 61961.28 225.64
06:07:53 54240.60 232.75
06:08:03 53174.01 201.58
06:08:13 24793.50 142.19

Time eth0
HH:MM:SS KB/s in KB/s out
06:08:23 9828.02 101.05
06:08:33 6347.83 92.07
06:08:43 5186.33 86.09
06:08:53 4242.28 68.86
06:09:03 16211.78 174.63
06:09:13 45714.96 386.25
06:09:23 45487.96 324.41

Time eth0
HH:MM:SS KB/s in KB/s out
06:09:33 52642.57 365.60
06:09:43 54077.28 393.56
06:09:53 55753.80 410.28
06:10:03 51455.15 381.77
06:10:13 56764.98 438.63
06:10:23 52408.95 414.59
06:10:33 51425.45 379.06

Time eth0
HH:MM:SS KB/s in KB/s out
06:10:43 46166.61 341.18
06:10:53 49369.84 339.05
06:11:03 55672.13 332.73
06:11:13 59797.08 354.31
06:11:23 51755.78 369.98
06:11:33 50497.74 393.98
06:11:43 52754.56 367.91

Time eth0
HH:MM:SS KB/s in KB/s out
06:11:53 59322.11 433.70
06:12:03 47937.46 357.03
06:12:13 56950.92 422.38
06:12:23 53209.36 417.04
06:12:33 54700.66 448.26
06:12:43 57002.73 432.88
06:12:53 50785.61 340.52

Time eth0
HH:MM:SS KB/s in KB/s out
06:13:03 45457.70 315.13
06:13:13 57787.79 379.35
06:13:23 53414.41 401.90
06:13:33 58053.28 437.63
06:13:43 45885.87 356.71
06:13:53 58876.77 387.10
06:14:03 57000.18 385.52

Time eth0
HH:MM:SS KB/s in KB/s out
06:14:13 46199.99 352.39
06:14:23 55424.30 429.83
06:14:33 55528.16 431.24
06:14:43 57677.24 476.14
06:14:53 50447.69 348.14
06:15:03 55249.19 355.66
06:15:13 57209.82 393.19

Time eth0
HH:MM:SS KB/s in KB/s out
06:15:23 61214.88 394.63
06:15:33 55836.75 348.11
06:15:43 50873.37 282.32
06:15:53 48212.24 298.00
06:16:03 33110.28 215.61
06:16:13 56192.92 387.81
06:16:23 58893.04 349.64

Time eth0
HH:MM:SS KB/s in KB/s out
06:16:33 58322.29 360.59
06:16:43 60243.76 417.89
06:16:53 56117.59 330.99
06:17:03 59758.54 363.06
06:17:13 60542.36 305.17
06:17:23 62802.79 348.33
06:17:33 58397.53 322.29

Time eth0
HH:MM:SS KB/s in KB/s out
06:17:43 52921.23 307.84
06:17:53 59356.96 335.58
06:18:03 57459.44 281.15
06:18:13 55923.34 319.51
06:18:23 60766.14 358.44
06:18:33 57951.68 327.02
06:18:43 53224.88 273.15

Time eth0
HH:MM:SS KB/s in KB/s out
06:18:53 56952.82 271.96
06:19:03 55608.83 279.30
06:19:13 57317.33 295.16
06:19:23 50699.88 282.80
06:19:33 38171.22 279.14
06:19:43 66214.86 442.78
06:19:53 59430.25 477.68

Time eth0
HH:MM:SS KB/s in KB/s out
06:20:03 38240.57 289.85
06:20:13 33687.98 293.99
06:20:23 31512.57 193.54
06:20:33 48491.11 306.89
06:20:43 59070.11 434.52
06:20:54 57904.61 426.79
06:21:04 46081.06 337.54

Time eth0
HH:MM:SS KB/s in KB/s out
06:21:14 76136.72 515.48
06:21:24 59867.00 394.22
06:21:34 59904.98 450.80
06:21:44 72706.01 418.26
06:21:54 62654.94 396.33
06:22:04 58394.74 376.46
06:22:14 37411.82 243.87

Time eth0
HH:MM:SS KB/s in KB/s out
06:22:24 60867.27 334.46
06:22:34 61979.32 383.86
06:22:44 59928.65 323.09
06:22:54 62966.96 346.77
06:23:04 44306.72 272.19
06:23:14 64106.22 366.36
06:23:24 56942.44 297.12

Time eth0
HH:MM:SS KB/s in KB/s out
06:23:34 59004.24 302.72
06:23:44 49701.54 221.25
06:23:54 56917.70 313.69
06:24:04 56399.29 295.54
06:24:14 63113.02 393.82
06:24:24 61270.35 378.45
06:24:34 64672.28 389.21

Time eth0
HH:MM:SS KB/s in KB/s out
06:24:44 62078.80 392.82
06:24:54 59409.10 353.30
06:25:04 62827.80 389.35
06:25:14 63927.91 344.69
06:25:24 61675.18 381.50
06:25:34 62690.08 375.58
06:25:44 62251.23 305.40

Time eth0
HH:MM:SS KB/s in KB/s out
06:25:54 50143.17 348.33
06:26:04 65032.11 387.71
06:26:14 65265.06 447.77
06:26:24 28856.31 344.91

almost no dropouts at all and after 20 minutes 400K blocks parallel downloaded, saved, and good progress doing the second and third pass on the bundles. but it was a bit faster with the old settings. at least I got the bandwidth going at full speed and this uses the optimized loading for the realtime bundle, which will speed up resuming

 

[grewalsatinder]

can we get a short and sweet " what is iguana "
am i right in assuming its a test program to see how fast you can validate the blockchain using GPUs?
google tells me " what is iguana" is a iguana
oh wait i found it
https://bitco.in/forum/threads/iguana-parallel-sync-full-btc-blockchain-in-30-minutes-uses-half-the-space-but-starts-up-instant.911/

I wrote this on a reddit post, so just putting it here as is. I think it covers 'What is Iguana?' question:


As jl777 says it's about solving problems of Bitcoin and Crypto currency all together: "privacy, speed, scalability, adoption".

• Iguana is UNDER DEVELOPMENT platform already including many solutions and features in it.
• Iguana is full MULTI Blockchain Wallet.

Right now Iguana supports:

• Full Bitcoin Wallet
• Full BitcoinDark Wallet
• More Blockchain Wallets can also be added with some JSON configuration.
• It is BLAZINGLY fast!
• Provides Bitcoin RPC API
• Uses a lot less disk space
• One Click installer for ALL platforms through Google Chrome as a Chrome App
• We are also working to support One Click App installer for other Web browsers like Mozilla Firefox, as many people don't want to use Google Chrome for different reasons.
• Also can run like a native daemon or service on any platform. Tested platforms are Linux, Windows, Android, Mac OS X & iOS (these days I'm troubleshooting iOS build, found glitches in build process).
• On mobile devices it would not be sending queries to external nodes, as mobile device would itself can be a FULL Bitcoin NODE!
• There are 100+ features and improvements done in iguana core over Bitcoin protocol ever since its been in development.
• It's GUI is in development by a team of half a dozen developers.
• But iguana core has got a lot covered in it's core API already if you want to make an app on top of it right NOW! Check http://docs.supernet.org
• I can not cover it's all features in this small space. You should considering spare your 5 minutes to read on that article. I bet it would be worth taking a look at it. Article LINK
• As it's GUI gets ready for beta tests iguana/SuperNET team will update it's availability on all media places. Specially keep an eye on iguana child-board on bitco.in forum

 

[Norway]

F2Pool just mined a Classic block. Cool!

http://nodecounter.com/#block_explorer

 

[freetrader]

Someone please tell me this is buttcoin leaking?

http://united-bitcoin.com/

 

[albin]

@freetrader

I'm having a very hard time figuring out what exactly they're trying to parody, the site is very unfocused.