Masternode Payment Variance
- Thread starter eduffield
- Start date
At this point my concern is that there may be a bug in voting. At least one MN has been paid FOURTEEN TIMES (XwSPJo3v4253WqS7kRFm7Huo3gDojRWuev). I mean there's "variance" and there's "winning the lotto."
P.S. El Presidente was kind enough to show that my MN WAS voted on for a payment, but that a bad actor did not pay it. Thanks! Now at least I know my node is working =)
P.S. El Presidente was kind enough to show that my MN WAS voted on for a payment, but that a bad actor did not pay it. Thanks! Now at least I know my node is working =)
Last edited by a moderator:
Nothing is broken, quite on the contrary.
You're hurting yourself by regenerating the masternode key. Essentially, if you also dont regenerate IP and new 1k deposit you can run into a little trouble, but in any case, new genkey = loosing all your votes and dropping down the bottom of the voting list.
On testnet I went 44+h until first payout, and that was with testnet with 30 something nodes up. Variance is only established if you let it roll until equal distribution. Until then, its sort of like "first-come, first served" in a random kind of way.
Yes, I think it was Stef who also pointed that out to me a couple days ago--but I had already done it =(
In any event, chaeplin showed that I should have gotten paid once already, but my MN won the voting on a block mined by a bad actor, so no payment. Plus, upon waking this morning, I found that I have received my first payment!
I myself have gotten 8 payments. Yesterday I got one and the day before I got 3. Today so far, none.
So yes, it is pretty random.
I have another node as well with 0 payments, so I guess I would be in the same position if I only had that node. Give it some time, it works, that is already proven.
You just need to roll the dice and win
So yes, it is pretty random.
I have another node as well with 0 payments, so I guess I would be in the same position if I only had that node. Give it some time, it works, that is already proven.
You just need to roll the dice and win
Hi all
Is there a way to check to see if our master-node has tried to been paid but hasn’t due to bad pool not paying.
Thanks
Is there a way to check to see if our master-node has tried to been paid but hasn’t due to bad pool not paying.
Thanks
Ummm Ya I have been paid once on Amazon ec2 but I changed to vultr so I could have my node in the uk. It's been on since Sunday night all green on elbzo masternode list. Might have tried to been paid but can't find any where which suggest it might be down to bad pool owner ????
To everyone showing doubt.
I got two masternodes, one of which have gotten 10 or so payments. The other had 0 until today when the first one popped in! So dont give up people!
I got two masternodes, one of which have gotten 10 or so payments. The other had 0 until today when the first one popped in! So dont give up people!
chaeplin updated his page at http://drk.poolhash.org/masternode.html - now you can filter your expected payouts
--> https://bitcointalk.org/index.php?topic=421615.msg7633822#msg7633822
My 4 nodes were elected 33 times, but only got paid 17 times - which almost exactly matches the current 'bad actor hashrate ratio' - so for me the voting system is working as expected, beside 'bad actors' and non-deterministic vote-lists througout the network. The rest boils down to variance and bad luck
--> https://bitcointalk.org/index.php?topic=421615.msg7633822#msg7633822
My 4 nodes were elected 33 times, but only got paid 17 times - which almost exactly matches the current 'bad actor hashrate ratio' - so for me the voting system is working as expected, beside 'bad actors' and non-deterministic vote-lists througout the network. The rest boils down to variance and bad luck
Last edited by a moderator:
I'm glad to see that El's masternode page now has the version number. It's good to see who is playing and who isn't.
That said, are there any changes planned for this weekend? I'm going out of town, so... no surprises, please !
That said, are there any changes planned for this weekend? I'm going out of town, so... no surprises, please !
You know it gonna happen the minute you go out the door, don't you?I'm glad to see that El's masternode page now has the version number. It's good to see who is playing and who isn't.
That said, are there any changes planned for this weekend? I'm going out of town, so... no surprises, please !
I missed the first days of MN payments because it caught me off guard travelling.
Chaeplin your the man... Thanks I can now see I have tried to been paid. I'm all working thank you
We need to get this to everyone worried about their Master Nodes. I have 2 because I got into DRK late and thats all I could afford. 1 has been online 6 days 7 hours with only 1 payment and the other 6 days 1 hour with 1 full payment and 2 half payments. I wasn't worried but it is encouraging to see that my 1st node had been voted 9 times in those 6 days but only one was paying and node 2 had been voted 7 times. Thats 16 votes for 2 master nodes in 6 days. Thats alot better than what I was seeing on the just payment end.chaeplin updated his page at http://drk.poolhash.org/masternode.html - now you can filter your expected payouts
--> https://bitcointalk.org/index.php?topic=421615.msg7633822#msg7633822
My 4 nodes were elected 33 times, but only got paid 17 times - which almost exactly matches the current 'bad actor hashrate ratio' - so for me the voting system is working as expected, beside 'bad actors' and non-deterministic vote-lists througout the network. The rest boils down to variance and bad luck
I'm seeing fewer and fewer ~0.5 DRK payouts. In fact, none since 07:38am EDT on June 3rd. So the "bad actor hunt" must be working.
Help would be highly appreciated. I've done the entire google return for "Darkcoin Pool" (etc) and only managed to sniff out a few. I've run out of pools to sniff out, need more. I'm monitoring loads, but the more the merrier, most of these have extremely low hashrate! Most updated right on user email request, so it does work!.
Please, drop in here and give me a hand!
https://darkcointalk.org/threads/ju...-collectively-help-drk-before-next-fork.1662/
Evan, will rc4's top-10 masternode system make payouts more even over the short term?
For instance, one of my nodes hasn't gotten paid in 4 days and another got
three payments in under an hour.
I'd like to see more even distribution. Would look better to investors!
Would selecting randomly from nodes that haven't been paid the longest be an option?
For instance, one of my nodes hasn't gotten paid in 4 days and another got
three payments in under an hour.
I'd like to see more even distribution. Would look better to investors!
Would selecting randomly from nodes that haven't been paid the longest be an option?
Zoom out and you will see a more even distribution, it's all matter of the scale.
Random is the best even distribution you can get btw. Look at mining. Finding hashes is most random. Some days you find 2 blocks, after that you spend weeks without anything.
I personally don't think it needs changing, I also believe that over a bigger timescales it will even itself out.
Also, rather than putting your eggs in one basket, how about you invest in a partial node? I've done that and set up a full masternode so I almost get daily payouts in some form or another
Also, rather than putting your eggs in one basket, how about you invest in a partial node? I've done that and set up a full masternode so I almost get daily payouts in some form or another
Zoom out and you will see a more even distribution, it's all matter of the scale.
Random is the best even distribution you can get btw. Look at mining. Finding hashes is most random. Some days you find 2 blocks, after that you spend weeks without anything.
Why is everybody so opposed to making sure each node gets its turn algorithmically?
Random selection from top of lists sorted by time-since-paid and time-visible-on-network would work fine.
Everybody would get a payment per time window of (number of masternodes mod 576) days.
I'm not against it, I just don't think its a high priority compared to the other features. It's working has expected and implemented so its not a bug.
This!
The randomization for MN payments is delivering the same results as mining - why are MN owners so impatient?
Why even think of changing the algorithm by introducing more complexity?
The more I think about this, the more I believe the regularity of payments is not only a non-issue, but "fixing it" could actually be undesirable for a number of reasons. First, development of your proposed feature would take valuable time and effort away from the development team working on more constructive features or functionality aimed at consumers. Even if they did add it, it would not increase each masternode's expected payout, so it wouldn't even create value for MN operators, especially over the long-term over which payments should even out. Second, it would make the selection of the next MN less random and more predictable... which could have negative consequences (e.g., perhaps it could enable an attacker to predict high-probability MNs for DDoS attacks to disrupt the network). Random selection makes guessing the likelihood of the next MN as difficult as possible, thus providing greater security to the network. Third, the system is working fine now... it was designed to attract hundreds of well-managed masternode operators to provide a stable, always-on network infrastructure for coin mixing and anonymity services; and that it what it has done. We now have 600+ stable MN's up and operating as desired, and failing to fix a perceived "unfairness" of how frequently individual MNs are paid seems an unlikely threat to this network's size and stability at this stage... in short, there's nothing broken to fix.
If you feel you are being unfairly compensated, you are free to exit your MN investment at any time. If you and others decide to take this option, it will increase the likelihood (and frequency) of the remaining MNs receiving a payment. For my part, I'm buying DRK as we speak to add another MN, because I think the payouts I'm seeing are suffient. I imagine there are others like me. Thus, there should be a "market equilibrium" of sorts that we reach. I just don't see a burning issue, and I see plenty of downsides to implementing a less random, less secure solution.
The last payment of my masternode was more than 90 hours ago. I am running 0.10.15.21 compiled from the source code from the git repository.
I understand that some variance is expected. But apart from this, how can I find out whether everything is OK with the voting process/network/other issues that effect the payment process?
I checked my debug.log on the masternode and detected a lot of "we are misbehaving" entries.
$ grep "we are misbehaving" debug.log | wc -l
Output for yesterday's debug.log: 408
Output for today's debug.log: 160
It is not a problem, is it? Sorry, it was not explained above clear enough.
I understand that some variance is expected. But apart from this, how can I find out whether everything is OK with the voting process/network/other issues that effect the payment process?
I checked my debug.log on the masternode and detected a lot of "we are misbehaving" entries.
$ grep "we are misbehaving" debug.log | wc -l
Output for yesterday's debug.log: 408
Output for today's debug.log: 160
It is not a problem, is it? Sorry, it was not explained above clear enough.
I think there is no random payment after the enforcement. You may want to run the daemon with -rindex option and delete peers.dat to see if you will get something after roughly 1200 blocks. The MN payments are distributed in a round robin fashion now and you should get a payment after ~1200 blocks. At least, my MN's are getting payments every ~1200 blocks when there are roughly 1200 MN's on the network.