Under Attack Again
-
[quote name=“groll” post=“22217” timestamp=“1374241623”]
This comes from Bitcoin as 6 block every 10 minutes so 1 hour of block. this require 51% to sustain as 50% of the block represent 50% of the window. this is different in FTC with 2.5 minutes2 things: make it the other sustains it
The first seems nearly ok you need >50% of the block to do so. a 40% with luck can do. but should be short live if not sustained.
to sustain you should need >50% of the block at current diff so for 2.5 it’s not 6 but 24 so median should be calculated on 49 blocks not 11. reducing the window to 6X2.5 can also be a solution, this will lead to some block rejected for time diff but 15 minutes seems ok. I don’t have time to check this exactly but i think GetAdjustedTime() is current time adjusted by the network time adjustement message. if it’s the case the adlustment should be lower then this to prevent DOS on time. (gettime() + adjust maxed at 35 *60)
note: dayligth saving times should not change anything as it’s UTC so this should have no impact. This was the reason in old time we puts 2h window in time validation of transaction in banking. even so 1-2 blocks would be rejected that is less problematic then what we currently have.
so my correction proposal is to replace the 2 * 60 * 60 by 15*60.
[/quote]That’s right, 2 hours future limit comes from Bitcoin and still there.
[code] // Check timestamp
if (block.GetBlockTime() > GetAdjustedTime() + 2 * 60 * 60)
return state.Invalid(error(“CheckBlock() : block timestamp too far in the future”));[/code]We’re 4x faster, so 4x reduction to 30 minutes seems logical. 15 minutes may be too optimistic, though acceptable usually.
nMedianTimeSpan=11 is also from Bitcoin. Could increase it 4x as well to 44.
GetAdjustedTime() is a bit weird in Bitcoin:
[code]
//
// “Never go to sea with two chronometers; take one or three.”
// Our three time sources are:
// - System clock
// - Median of other nodes clocks
// - The user (asking the user to fix the system clock if the first two disagree)
//
static int64 nMockTime = 0; // For unit testingint64 GetTime()
{
if (nMockTime) return nMockTime;return time(NULL);
}void SetMockTime(int64 nMockTimeIn)
{
nMockTime = nMockTimeIn;
}static int64 nTimeOffset = 0;
int64 GetTimeOffset()
{
return nTimeOffset;
}int64 GetAdjustedTime()
{
return GetTime() + GetTimeOffset();
}
[/code]In fact, it relays what standard UNIX/POSIX time() call says. Return value is seconds since the Epoch (01-Jan-1970 00:00:00 UTC).
There is another limit in AddTimeData():
[code=util.cpp / Bitcoin] // Only let other nodes change our time by so much
if (abs64(nMedian) < 70 * 60)
{
nTimeOffset = nMedian;
}
else
{
nTimeOffset = 0;[/code]Litecoin is more strict, we follow:
[code=util.cpp / Litecoin]// Only let other nodes change our time by so much
if (abs64(nMedian) < 35 * 60) // Litecoin: changed maximum adjust to 35 mins to avoid letting peers change our time too much in case of an attack.
{
nTimeOffset = nMedian;
}
else
{
nTimeOffset = 0;[/code]May limit even more.
-
[quote name=“wesphily” post=“22337” timestamp=“1374274466”]
[quote author=ghostlander link=topic=2847.msg22239#msg22239 date=1374247221]
I’m afraid we have to act quickly. Either with advanced checkpointing or something else. People start to panic.
[/quote]All this means is that people who are smart are going to make a killing on the flip side.
[/quote]The majority of people are not smart, they just follow the crowd. That’s how democracy works. A serious attack with or without double spend + panic dump may have FTC delisted from BTC-e, and that’s definitely not good.
-
[code]enum { nMedianTimeSpan=44 };[/code]
…and…
[code] // Check timestamp
if (block.GetBlockTime() > GetAdjustedTime() + 30 * 60)
return state.Invalid(error(“CheckBlock() : block timestamp too far in the future”));[/code]Seem like good suggestions that I will review.
-
[quote]Code: [Select]
enum { nMedianTimeSpan=44 };
…and…
Code: [Select]
// Check timestamp
if (block.GetBlockTime() > GetAdjustedTime() + 30 * 60)
return state.Invalid(error(“CheckBlock() : block timestamp too far in the future”));[/quote]
median is better with odd number not even else it need to be the average of the 2 median value in even number so 43, 45, 47 seems better choice. in fact 47 is the 2h value as it’s: 2h target number of block -1. (bitcoin is 12-1 we should have 48-1)for the second one I write it a bit fast when i say 6*2.5=15 it should have been windows +1*2.5=11+1*2.5=30 as bitcoin is 12*10=2*60, 2h. since we up the median range 47+1*2.5= 120 min. so any value below so 120 min. can still be fine. but I would not go bellow 2 time the max adjusted to avoid other attack as explicitly stated in code(can be used to make similar DOS if you can contain the network to your node for a target). if our time can be change we should be inside the allowed time.
the time change need to be looked at. not sure what it is in bitcoin. but FTC and LTC use 35 minutes, putting max future below it is opening door to DOS I would keep this value below 1/2 the max future allowed (35 mintestes for 2h is fine, for 30 i would lower it to 15)
[code]/ Only let other nodes change our time by so much
if (abs64(nMedian) < 35 * 60) // Litecoin: changed maximum adjust to 35 mins to avoid letting peers change our time too much in case of an attack.[/code]going with 47 and keeping 2h would be my suggestion, reducing the 2 hours would require additional special handling as we have block with that . growing the number to 47 has no impact mostly if no time attack is done at that time else it can cause a network split for non updated client. a simple after block X to let poeple update should do the trick for this one
note: 47 will increase the range of time back the 51% can do. for retarget. as he will be allowed to get time 24 block back not just 6
-
Very interesting, what does the code say if the Block is in the past? As per the current block time anomalies?
re: Why does Block 53872 has a time before block 53871?
-
Here is my analysis of the inserted blocks over night and which coins were effected. Mainly UNOCS.
I resolved the mincoin error, now I’ve got a trend. It was due to having to type the new block figures into the spreadsheet by hand, where cut and paste was disabled on the dustcoin.
>>>>>>> Zero days means the calculated block start date coincides with the real coin announce start date.
Coin 21st July Blocks 20th July Blocks Blocks 18.7.13 Reward Block Mins Block Time (Hrs) BlockTimeCalctRunningDays CalculatedEffectiveMiningStartDay Announce date Days Early 21.7.13 Days Early 20.7.13 Days Early 18.7.13
Alphacoin 104203 0.5 0.008 #VALUE!
Bottlecaps (CAP) 38183 0.000 0.000 21 July 2013 24 June 2013 -27.37 -27.36 -26.44
Mincoin 75158 135109 75153 2 10 0.167 521.931 15 February 2012 01 May 2013 440.56 441.00 441.45
Ixcoin 148590 148465 147530 96 10 0.167 1031.875 23 September 2010 10 August 2011 320.51 319.65 314.07
Freicoin 30926 30911 30831 224 10 0.167 214.764 18 December 2012 11 May 2013 143.40 143.30 143.66
PPCoin 59571 59393 58526 363.3 10 0.167 413.688 02 June 2012 10 August 2012 68.32 67.09 61.99
NovaCoin 32860 32725 31855 9.85 10 0.167 228.194 05 December 2012 11 February 2013 67.83 66.90 61.77
Bitcoin 247685 247529 246664 25 10 0.167 1720.035 04 November 2008 03 January 2009 59.67 58.59 53.50
BitBar 16469 16332 15568 0.2 10 0.167 114.368 29 March 2013 02 May 2013 34.00 33.06 28.67
Namecoin 122957 122797 121886 50 10 0.167 853.868 20 March 2011 18 April 2011 28.50 27.40 21.99
Litecoin 392523 392008 389006 50 2.5 0.042 681.464 08 September 2011 07 October 2011 28.10 27.21 22.92
ChinaCoin 59728 59728 59727 88 2.5 0.042 103.694 08 April 2013 01 May 2013 22.33 22.34 23.25
Megacoin (MEC) 38070 37550 34572 250 2.4 0.040 63.450 18 May 2013 07 June 2013 19.08 18.23 14.18
Bytecoin 18427 18410 18402 50 10 0.167 127.965 15 March 2013 01 April 2013 16.60 16.49 17.35
PhenixCoin 67708 66722 64402 50 1.5 0.025 70.529 11 May 2013 20 May 2013 8.16 7.15 5.64
Stablecoin (SBC) 88943 87146 76865 25 0.6666666666667 0.011 41.177 10 June 2013 15 June 2013 4.81 3.99 0.14
Fastcoin (FST) 380482 373867 338167 32 0.2 0.003 52.845 29 May 2013 29 May 2013 -0.52 -1.43 -5.47
Digitalcoin (DGC) 270719 266825 246892 20 0.3333333333333 0.006 62.666 19 May 2013 19 May 2013 -0.70 -1.59 -5.29
Worldcoin (WDC) 382275 376879 349791 32 0.25 0.004 66.367 16 May 2013 14 May 2013 -2.00 -2.93 -6.71
Feathercoin 53912 53503 50362 200 2.5 0.042 93.597 18 April 2013 16 April 2013 -2.77 -3.47 -8.01[attachment deleted by admin]
-
FTC appears to be under attack again, just a heads up.
-
I agree, here’s the latest analysis chart which shows which coins are under attack last 3 days. It is in terms of the number of false days worth of coins, inserted.
Some of the “invalid” mined blocks could be due to extra “valid blocks” generated by manipulating the difficulty with DDoS on Pools.
It’s obvious to me, from this chart, it is an outside element doing the attacks on all “viable” crypto coins, shouldn’t stick together to fight them off?
Coin Additional Coin Days Added above block rate specification 18-21-July 2013 21st July Blocks 20th July Blocks Blocks 18.7.13
Alphacoin 104203
Bottlecaps (CAP) 38183
Test 1yr 10min -0.98 52560.000 52560.000 52560.000
Hypercoin (HYC) -0.98 20464 20464 20464
ChinaCoin -0.98 59728 59728 59727
Mincoin -0.94 75158 75155 75153
Bytecoin -0.80 18427 18410 18402
Nibble (NBL) -0.78 25027 25003 24912
Freicoin -0.32 30926 30911 30831
Luckycoin (LKY) -0.26 57050 56962 56015
PhenixCoin 2.47 67708 66722 64402
Terracoin 3.38 162577 162173 159438
Devcoin 4.45 98324 98205 97543
Digitalcoin (DGC) 4.54 270719 266825 246892
BBQCoin 4.60 501132 500053 493093
Stablecoin (SBC) 4.61 88943 87146 76865
Worldcoin (WDC) 4.66 382275 376879 349791
Franko 4.84 166681 164534 149921
Megacoin (MEC) 4.85 38070 37550 34572
Fastcoin (FST) 4.90 380482 373867 338167
Litecoin 5.13 392523 392008 389006
Feathercoin 5.19 53912 53503 50362
BitBar 5.28 16469 16332 15568
NovaCoin 6.00 32860 32725 31855
Bitcoin 6.11 247685 247529 246664
PPCoin 6.28 59571 59393 58526
Ixcoin 6.38 148590 148465 147530
Namecoin 6.46 122957 122797 121886[attachment deleted by admin]
-
[quote name=“wrapper0feather” post=“22479” timestamp=“1374397712”]
I agree, here’s the latest analysis chart which shows which coins are under attack last 3 days. It is in terms of the number of false days worth of coins, inserted.Some of the “invalid” mined blocks could be due to extra “valid blocks” generated by manipulating the difficulty with DDoS on Pools.
It’s obvious to me, from this chart, it is an outside element doing the attacks on all “viable” crypto coins, shouldn’t stick together to fight them off?
Coin Additional Coin Days Added above block rate specification 18-21-July 2013 21st July Blocks 20th July Blocks Blocks 18.7.13
Alphacoin 104203
Bottlecaps (CAP) 38183
Test 1yr 10min -0.98 52560.000 52560.000 52560.000
Hypercoin (HYC) -0.98 20464 20464 20464
ChinaCoin -0.98 59728 59728 59727
Mincoin -0.94 75158 75155 75153
Bytecoin -0.80 18427 18410 18402
Nibble (NBL) -0.78 25027 25003 24912
Freicoin -0.32 30926 30911 30831
Luckycoin (LKY) -0.26 57050 56962 56015
PhenixCoin 2.47 67708 66722 64402
Terracoin 3.38 162577 162173 159438
Devcoin 4.45 98324 98205 97543
Digitalcoin (DGC) 4.54 270719 266825 246892
BBQCoin 4.60 501132 500053 493093
Stablecoin (SBC) 4.61 88943 87146 76865
Worldcoin (WDC) 4.66 382275 376879 349791
Franko 4.84 166681 164534 149921
Megacoin (MEC) 4.85 38070 37550 34572
Fastcoin (FST) 4.90 380482 373867 338167
Litecoin 5.13 392523 392008 389006
Feathercoin 5.19 53912 53503 50362
BitBar 5.28 16469 16332 15568
NovaCoin 6.00 32860 32725 31855
Bitcoin 6.11 247685 247529 246664
PPCoin 6.28 59571 59393 58526
Ixcoin 6.38 148590 148465 147530
Namecoin 6.46 122957 122797 121886
[/quote]All coins are being attacked? Hm…
-
Nevermind Will edit in the morning
-
[quote name=“wrapper0feather” post=“22468” timestamp=“1374389156”]
Very interesting, what does the code say if the Block is in the past? As per the current block time anomalies?re: Why does Block 53872 has a time before block 53871?
[/quote]It’s allowed if within limits. There may be some clock drift, though servers are supposed to synchronise their clocks through NTP at least daily.
-
[quote name=“111111111111111111111” post=“22476” timestamp=“1374395173”]
FTC appears to be under attack again, just a heads up.
[/quote]That’s not an attack. The difficulty has decreased recently from 104.66 to 74.01, so profitability vs. BTC is 250%.
-
Here’s the latest chart’s showing the additional “coins mining days per day”, above what should have been mined. The red bars are from the change in block production, during attack. The blue shows where it has fallen back to currently.
I think there are still too many injected blocks in the “current condition”. I’ll retest later and see if indicates a more subtle attack, or if it is just due to lag in the difficulty change compensating for hash rate variability.
[attachment deleted by admin]
-
Big Vern dropped us again about 40 mins ago.
His tweet “FTC trade paid is suspended again due to an attack that has broken the chain again.” -
[quote name=“Pyxis” post=“22502” timestamp=“1374413062”]
Big Vern dropped us again about 40 mins ago.
His tweet “FTC trade paid is suspended again due to an attack that has broken the chain again.”
[/quote]Was only for a brief period while I checked out some stats. It was put back online quickly.
BigVern
-
Help, I think we are under attack again. Some addresses are getting impossible block rates, and my mining has gone to zero.
eg.? 6oXrGm4vwKUhqQR1hJaCMZoi2Eh7Weu2RW
1660a2cc29… 54421 2013-07-21 13:38:39 200 68400.66264736 FTC
6342aceef3… 54424 2013-07-21 13:47:18 200 68600.66264736 FTC
c9bb2291d0… 54425 2013-07-21 13:47:19 200 68800.66264736 FTC
7ea1e75d60… 54426 2013-07-21 13:47:20 200 69000.66264736 FTC
6f9ca0366c… 54428 2013-07-21 13:47:21 200 69200.66264736 FTC
40366f6003… 54431 2013-07-21 13:47:22 200 69400.66264736 FTC
2a2f7ae9df… 54430 2013-07-21 13:47:23 200 69600.66264736 FTC
0169b5ac0b… 54427 2013-07-21 13:47:25 200.000195 69800.66284236 FTC
9d46bf6d9b… 54432 2013-07-21 13:47:27 200.005 70000.66784236 FTC -
These block rates are not impossible. It is allowed to find blocks shorter than the allocated time window if the blockchain is shorter than it is expected to be at this point in time.
It’s my opinion that somebody has realized that many blockchains are underfilled, and are using their massive hashpower to get these blocks quickly until the chain reaches it’s maximum possible height.
I’m still finding blocks on fcpool, but not as often as expected - this is due to the large hashrate currently on the network. I’d expect difficulty to rise soon.
BigVern
-
*******
[attachment deleted by admin]
-
With the diff jumping again and our hash rate about to drop we better be ready for another attack. Also if you have not seen the post about the hash rate on the stats page its been doing some weird stuff.
-
Find the latest trend in Extra Mined Coin Days, in terms of days mined before the announce date. For the last 5 days.
I have split the chart up to show the detail
1.All coins being analysed
2.Just coins in excesses of specification, no IxCoin.
3.Just coins moving closer to excessive coins to specification.
4.Just coins which have not produced the correct level of coins.There could be some modifications to the figures once someone checks the spreadsheet, but the trend is correct. I am sampling at about 10 to 11 GMT, so there will be a slight variability in 24 hr sample.
[attachment deleted by admin]
