Topic: [ANN][RIC] Riecoin: constellations POW *CPU* HARD FORK successful, world record - page 47. (Read 685265 times)

I think we're both right.

The fork caused the diff drop, by making it seem as if the hashrate was actually only about 60% of what it was.

The diff dropped too low for the real hashrate, so after the fork healed, the hashrate was huge for the diff.  That caused the diff spike.

And with the diff spike, some huge miner(s?) dropped out.

And you're right - this is the kind of diff swing that newer diff salgorithms like DGW were designed to prevent.  Let's hope it settles down for now.

That's probably from the fork - if the network was cut in half, then half the mining output during the fork got thrown away (in practice, it won't be that much, because there's no reason for exactly half to be on each part).  So the diff drop will overcompensate this time, and then it'll bounce pretty high next time, and then start settling back down to normal for the amount of mining power that's out there.

Not very often recently, but I gather there were a lot of stability problems about a year and a half ago.  I was completely away from RIC for a while (sabbaticals are the best thing ever for academics... :-), so I wasn't really watching.  If you look back in the forum, though, you can probably find some discussions about crazy-high orphan rates when the network first started getting DDoS'd. 

Here's a fork reported in October 2017:  https://bitcointalksearch.org/topic/m.23395787

Here's a discussion from 2015 about the DDoS attacks - https://bitcointalk.org/index.php?topic=446703.4080

That accusation (Xpoolx accusing Simba (RiePool.ovh) of the DDoS) is one of the ones I was referring to.

And here's a plot of the historical orphan rates (orphans are what happens when there's a race to complete the chain -- a fork is just when the race goes on for a longer time):

https://chainz.cryptoid.info/ric/orphans.dws

You can see that things got really bad in late 2016 until early 2017.  Some time around then is when several people started adding extra DDoS-protected nodes to the network to try to reduce the chaos.  The orphan rates calmed down, but there's been a new spike the last few days.  (Amusingly, the DDoS is why I started paying attention again, because whomever was doing it launched an attack strong enough it took all of Carnegie Mellon University offline -- I had left a Riecoin node running in my office for the last few years, completely forgotten about it, and then... un-forgot. :-)

How often do these forks happen?

Unsure.  But we seem to have had another fork.  Not sure if it's related to that or not -- high tx volume affecting block propagation times?  Seems unlikely at 200k unless some of the nodes are already slow because of DDoS.

This one wasn't as bad as the previous, but it'll still be 72 blocks before the wallets come out of warning mode.  About 3 hours.

Fantastic - thank you!

Just as a heads-up, my Frantech node just got hit by something that took it offline for a minute or two (or had a hardware glitch).  Investigating, but I suspect this means the DDoS remains alive.

Expect attacks such as very large numbers of connections from many nodes to try to exhaust your numconnections= limit, bandwidth exhaustion attacks, TCP synfloods, and attempts to get your node to initiate outbound traffic.

Also, someone seems to be trying to attack the network right now with massive numbers of transactions.  They're not doing anything, but they'll use a bit more CPU and bandwidth than you may have thought you'd normally need.  This started at about block 809154 (about two hours ago).

I have set up two nodes hosted on OVH

176.31.136.231
176.31.210.206

Network appears stable.  Block explorer shows happy, and my wallet has stopped warning about the fork.  Don't know how long Poloniex waits.  But all appears happy.

I'll try to get another well-connected public node up in the next few days.  sigh.

There was a 54 block fork.  It's resolved, but the wallets "pause" for 72 blocks after the fork to make sure things are stable:


Polo and other places should be operational again by block 809120, which is 29 blocks from now.  Approximately 29 * 2.5 = 72 minutes from now.

I don't believe I've insulted you, so perhaps you would return the courtesy.  Nor did I accuse you, though I did mention that others have -- you'll note that I specifically said I don't know whether any of it is true.  And -- seriously, are you making threats?  Please act like an adult.  This is a geeky, prime number coin.  With the blocks per day you're mining on Riepool, it can't be worth any real money to you.  Unless, of course, you *are* in collusion with the botnet(s) that are mining most of the coins, in which case, I can understand why you'd be so upset about all of this.  But otherwise, your anger seems strangely out of proportion to the stakes involved.

In regards to this fork/orphan thing -- the last time I think it was because one of my public nodes got messed up, but it's been up and stable this time.  I'm happy to set up another DDoS-shielded node to try to help stabilize things.

Has anyone tested whether OVH's nodes are safe for Riecoin nodes?

  -Dave

ublock, like all those who do NOT mining under "All Others" will have major problems of orphanage, if that SOLO miner does not decide to do things properly.

Needless to lash out at RiePool without ever having any kind of evidence of what was claimed.  

The problems here are very different!

for ziiip, dga and other "idiots": I advise you to stop with your ridicolous and unfounded accusations. We have already begun to take action, if you continue we will go ahead even outside the forum... Please STOP!

People often use connect= so they don't get connected to whatever nodes trigger the huge DDoS floods.

As an unrelated aside:  XpoolX, your clock seems to be ahead.  The block explorer shows things like 5 min, 12 min, 6 min, 18 min... where the 6 minute old block (in block sequence, but out of time order) is an XpoolX block.  I don't think this matters in any serious way, but you might want to fix it before it gets so out of sync your blocks start getting rejected.

i think using addnode will be better option here, correct me if im wrong, your commands will limit wallet to connect those nods only :-?

it seems that the accumulation of ric is insane on poloniex and waiting for its halving.

did you add some nodes to the conf file like the following below??

i cant sync my wallet, no block connections .... any idea?

uBlock.it is running much better now. Attack is now being successfully mitigated. 

Apologies for the issues over the last week.
Happy mining.

I traced some of the major solo addresses that showed up around the DDoS start to their Poloniex deposit addresses and reported them to a contact in law enforcement to see if there were any leads they could follow up on.  Haven't heard back, but I wouldn't expect to.  I posted about the analysis about six months ago. 

So... have you actually tried to follow the money? You know, for science.

I suspect that there's a large botnet that monetizes by mining RIC and uses the DDoS to eliminate the competition.  But it could also be a pool or something - others have speculated, for example, that Simba is actually responsible for it, to try to maintain a monopoly on the pool fees.  I have no information one way or another about that.  But it's usually the case that following the money is the path.