Topic: [1500 TH] p2pool: Decentralized, DoS-resistant, Hop-Proof pool - page 704. (Read 2591919 times)
NOBODY SAY ANYTHING ABOUT THE L WORD
Also have to keep in mind, that blocks and shares don't propogate instantly, being a decentralized network.
When one of us finds a block, it is relayed to the network more like ripples on a pond. Your P2P tells at least 10 friends, and their P2P's tell at least 10 friends, and so on and so on. Depending on how many nodes there are, it could take a long time (relatively speaking) for every node to get the new block.
If another node also solves a block before the first block's ripple reaches it, it sends out its own ripples, and eventually those ripples meet and try and cancel each other out, like the earlier poster saw with it going back and forth, until the stronger ripple has dissapated the weaker ripple, resulting in the orphan.
-- Smoov
When one of us finds a block, it is relayed to the network more like ripples on a pond. Your P2P tells at least 10 friends, and their P2P's tell at least 10 friends, and so on and so on. Depending on how many nodes there are, it could take a long time (relatively speaking) for every node to get the new block.
If another node also solves a block before the first block's ripple reaches it, it sends out its own ripples, and eventually those ripples meet and try and cancel each other out, like the earlier poster saw with it going back and forth, until the stronger ripple has dissapated the weaker ripple, resulting in the orphan.
-- Smoov
Anyway, my point was more that all that passing ancient blocks to bitcoind is kinda buggy. Does bitcoind do anything drastic when a peer keeps sending the same block over and over?
It's wouldn't call it buggy since it's doing exactly what it was designed to do: when a peer announces a block, p2pool passes it on to the local bitcoin node. It's apparently just simplistic and doesn't attempt to keep track of blocks that it already told the local node about.
Also, the recent update just made p2pool more agressive about telling peers about blocks and so maybe you're seeing more messages about them. I don't think it hurts anything and I wouldn't change it since agressive distribution of blocks is a good thing in order to reduce orphaned blocks (which we have had more than our fair share of).
There seems to be something dodgy going on here:
Notice how p2pool keeps passing the same blocks to bitcoind over and over. Seems the orphan didn't want to go down without a fight...
Quote
2012-05-07 19:31:15.748325 GOT BLOCK FROM PEER! Passing to bitcoind! 52bb8595 bitcoin: http://blockexplorer.com/block/000000000000080aaeb78a2217dae7e1c66f431d48c6c4085026d50c52bb8595
2012-05-07 19:31:29.376311 GOT BLOCK FROM PEER! Passing to bitcoind! 58c76780 bitcoin: http://blockexplorer.com/block/000000000000015e628594f22a190a6ec8ba9c7af57a2c549bd34edd58c76780
2012-05-07 19:31:29.376311 GOT BLOCK FROM PEER! Passing to bitcoind! 58c76780 bitcoin: http://blockexplorer.com/block/000000000000015e628594f22a190a6ec8ba9c7af57a2c549bd34edd58c76780
Notice how p2pool keeps passing the same blocks to bitcoind over and over. Seems the orphan didn't want to go down without a fight...
I suppose your node is NTP-synced ? My node noticed the blocks like this :
[snip]
I notice 2 things :
- my node saw the blocks ~2 seconds later than yours
- there's a ~15 second delay between the two blocks
If you don't run ntpd 2 seconds isn't much, if you do it's a little worrisome : 2 seconds to broadcast such information is high. I'd expect something like 500ms at most.
I don't understand how a node could produce an orphaned block 15 seconds after another has been seen by p2pool if things are working correctly. If the node responsible for the orphan block could be found it may be useful to find out what happened from its point of view.
[corrected your quote]
My node is NTP-synced at midnight, and regularly off by 5s, so 2s by 19:30 isn't too unlikely. Note, however, that the timestamp of each block is ~1 minute after my node received them - and you complain about 2s?
Anyway, my point was more that all that passing ancient blocks to bitcoind is kinda buggy. Does bitcoind do anything drastic when a peer keeps sending the same block over and over?
Well - don't forget the obvious ...
Each 'pool' is whatever hardware, network and poorly tuned OS the person running p2pool is using.
Thus there can quite obviously be some people running p2pool with major latency issues in their setup.
In that case, consider that if they find a block, they are using a 'poorly tuned pool' to get the block out on the net to the block chain and share chain (rather than a low latency highly efficient pool running on high quality hardware with a well tuned OS with lots of network connections ...)
What's that comment about chains and weakest links ...
Luckily they wont find blocks very often ...
Each 'pool' is whatever hardware, network and poorly tuned OS the person running p2pool is using.
Thus there can quite obviously be some people running p2pool with major latency issues in their setup.
In that case, consider that if they find a block, they are using a 'poorly tuned pool' to get the block out on the net to the block chain and share chain (rather than a low latency highly efficient pool running on high quality hardware with a well tuned OS with lots of network connections ...)
What's that comment about chains and weakest links ...
Luckily they wont find blocks very often ...
There seems to be something dodgy going on here:
Quote
2012-05-07 19:31:15.748325 GOT BLOCK FROM PEER! Passing to bitcoind! 52bb8595 bitcoin: http://blockexplorer.com/block/000000000000080aaeb78a2217dae7e1c66f431d48c6c4085026d50c52bb8595
2012-05-07 19:31:29.376311 GOT BLOCK FROM PEER! Passing to bitcoind! 58c76780 bitcoin: http://blockexplorer.com/block/000000000000015e628594f22a190a6ec8ba9c7af57a2c549bd34edd58c76780
Notice how p2pool keeps passing the same blocks to bitcoind over and over. Seems the orphan didn't want to go down without a fight...
2012-05-07 19:31:29.376311 GOT BLOCK FROM PEER! Passing to bitcoind! 58c76780 bitcoin: http://blockexplorer.com/block/000000000000015e628594f22a190a6ec8ba9c7af57a2c549bd34edd58c76780
Notice how p2pool keeps passing the same blocks to bitcoind over and over. Seems the orphan didn't want to go down without a fight...
I suppose your node is NTP-synced ? My node noticed the blocks like this :
Code:
2012-05-07 19:31:17.798942 GOT BLOCK FROM PEER! Passing to bitcoind! 52bb8595 bitcoin: http://blockexplorer.com/block/000000000000080aaeb78a2217dae7e1c66f431d48c6c4085026d50c52bb8595
2012-05-07 19:31:17.799030
2012-05-07 19:31:19.374251 New work for worker! Difficulty: 0.999985 Share difficulty: 671.595424 Total block value: 50.106500 BTC including 171 transactions
2012-05-07 19:31:19.540366 P2Pool: 17346 shares in chain (17350 verified/17350 total) Peers: 50 (40 incoming)
2012-05-07 19:31:19.540528 Local: 1460MH/s in last 10.0 minutes Local dead on arrival: ~5.9% (3-10%) Expected time to share: 32.8 minutes
2012-05-07 19:31:19.540603 Shares: 136 (8 orphan, 8 dead) Stale rate: ~11.8% (7-19%) Efficiency: ~97.5% (90-103%) Current payout: 0.2078 BTC
2012-05-07 19:31:19.540763 Pool: 322GH/s Stale rate: 9.5% Expected time to block: 5.6 hours
2012-05-07 19:31:20.746810 New work for worker! Difficulty: 0.999985 Share difficulty: 675.393531 Total block value: 50.106500 BTC including 171 transactions
2012-05-07 19:31:22.689407 P2Pool: 17347 shares in chain (17351 verified/17351 total) Peers: 50 (40 incoming)
2012-05-07 19:31:22.689564 Local: 1467MH/s in last 10.0 minutes Local dead on arrival: ~6.3% (3-11%) Expected time to share: 32.9 minutes
2012-05-07 19:31:22.689667 Shares: 136 (8 orphan, 8 dead) Stale rate: ~11.8% (7-19%) Efficiency: ~97.8% (90-103%) Current payout: 0.2078 BTC
2012-05-07 19:31:22.689767 Pool: 325GH/s Stale rate: 9.8% Expected time to block: 5.5 hours
2012-05-07 19:31:25.756062 P2Pool: 17347 shares in chain (17351 verified/17351 total) Peers: 50 (40 incoming)
2012-05-07 19:31:25.756237 Local: 1460MH/s in last 10.0 minutes Local dead on arrival: ~6.4% (3-11%) Expected time to share: 33.1 minutes
2012-05-07 19:31:25.756316 Shares: 136 (8 orphan, 8 dead) Stale rate: ~11.8% (7-19%) Efficiency: ~97.8% (90-103%) Current payout: 0.2078 BTC
2012-05-07 19:31:25.756401 Pool: 325GH/s Stale rate: 9.8% Expected time to block: 5.5 hours
2012-05-07 19:31:29.498357 P2Pool: 17347 shares in chain (17351 verified/17351 total) Peers: 50 (40 incoming)
2012-05-07 19:31:29.498509 Local: 1445MH/s in last 10.0 minutes Local dead on arrival: ~6.4% (3-11%) Expected time to share: 33.4 minutes
2012-05-07 19:31:29.498593 Shares: 136 (8 orphan, 8 dead) Stale rate: ~11.8% (7-19%) Efficiency: ~97.8% (90-103%) Current payout: 0.2078 BTC
2012-05-07 19:31:29.498685 Pool: 325GH/s Stale rate: 9.8% Expected time to block: 5.5 hours
2012-05-07 19:31:32.871357
2012-05-07 19:31:32.871528 GOT BLOCK FROM PEER! Passing to bitcoind! 58c76780 bitcoin: http://blockexplorer.com/block/000000000000015e628594f22a190a6ec8ba9c7af57a2c549bd34edd58c76780
I notice 2 things :
- my node saw the blocks ~2 seconds later than yours
- there's a ~15 second delay between the two blocks
If you don't run ntpd 2 seconds isn't much, if you do it's a little worrisome : 2 seconds to broadcast such information is high. I'd expect something like 500ms at most.
I don't understand how a node could produce an orphaned block 15 seconds after another has been seen by p2pool if things are working correctly. If the node responsible for the orphan block could be found it may be useful to find out what happened from its point of view.
There seems to be something dodgy going on here:
Notice how p2pool keeps passing the same blocks to bitcoind over and over. Seems the orphan didn't want to go down without a fight...
Quote
2012-05-07 19:31:15.748325 GOT BLOCK FROM PEER! Passing to bitcoind! 52bb8595 bitcoin: http://blockexplorer.com/block/000000000000080aaeb78a2217dae7e1c66f431d48c6c4085026d50c52bb8595
2012-05-07 19:31:29.376311 GOT BLOCK FROM PEER! Passing to bitcoind! 58c76780 bitcoin: http://blockexplorer.com/block/000000000000015e628594f22a190a6ec8ba9c7af57a2c549bd34edd58c76780
2012-05-07 19:32:22.195304 GOT BLOCK FROM PEER! Passing to bitcoind! 5e0bae54 bitcoin: http://blockexplorer.com/block/00000000000009f544164219e9d5580020c802b27b4e30d61924497b5e0bae54
2012-05-07 19:34:44.081425 GOT BLOCK FROM PEER! Passing to bitcoind! 8f095de4 bitcoin: http://blockexplorer.com/block/00000000000004b35e124ad23992137c87e1c2d37e8a15189879db408f095de4
2012-05-07 19:36:49.426890 GOT BLOCK FROM PEER! Passing to bitcoind! 58c76780 bitcoin: http://blockexplorer.com/block/000000000000015e628594f22a190a6ec8ba9c7af57a2c549bd34edd58c76780
2012-05-07 19:37:08.294949 GOT BLOCK FROM PEER! Passing to bitcoind! 52bb8595 bitcoin: http://blockexplorer.com/block/000000000000080aaeb78a2217dae7e1c66f431d48c6c4085026d50c52bb8595
2012-05-07 19:42:05.599642 GOT BLOCK FROM PEER! Passing to bitcoind! 58c76780 bitcoin: http://blockexplorer.com/block/000000000000015e628594f22a190a6ec8ba9c7af57a2c549bd34edd58c76780
2012-05-07 19:44:03.164002 GOT BLOCK FROM PEER! Passing to bitcoind! 52bb8595 bitcoin: http://blockexplorer.com/block/000000000000080aaeb78a2217dae7e1c66f431d48c6c4085026d50c52bb8595
2012-05-07 19:31:29.376311 GOT BLOCK FROM PEER! Passing to bitcoind! 58c76780 bitcoin: http://blockexplorer.com/block/000000000000015e628594f22a190a6ec8ba9c7af57a2c549bd34edd58c76780
2012-05-07 19:32:22.195304 GOT BLOCK FROM PEER! Passing to bitcoind! 5e0bae54 bitcoin: http://blockexplorer.com/block/00000000000009f544164219e9d5580020c802b27b4e30d61924497b5e0bae54
2012-05-07 19:34:44.081425 GOT BLOCK FROM PEER! Passing to bitcoind! 8f095de4 bitcoin: http://blockexplorer.com/block/00000000000004b35e124ad23992137c87e1c2d37e8a15189879db408f095de4
2012-05-07 19:36:49.426890 GOT BLOCK FROM PEER! Passing to bitcoind! 58c76780 bitcoin: http://blockexplorer.com/block/000000000000015e628594f22a190a6ec8ba9c7af57a2c549bd34edd58c76780
2012-05-07 19:37:08.294949 GOT BLOCK FROM PEER! Passing to bitcoind! 52bb8595 bitcoin: http://blockexplorer.com/block/000000000000080aaeb78a2217dae7e1c66f431d48c6c4085026d50c52bb8595
2012-05-07 19:42:05.599642 GOT BLOCK FROM PEER! Passing to bitcoind! 58c76780 bitcoin: http://blockexplorer.com/block/000000000000015e628594f22a190a6ec8ba9c7af57a2c549bd34edd58c76780
2012-05-07 19:44:03.164002 GOT BLOCK FROM PEER! Passing to bitcoind! 52bb8595 bitcoin: http://blockexplorer.com/block/000000000000080aaeb78a2217dae7e1c66f431d48c6c4085026d50c52bb8595
Notice how p2pool keeps passing the same blocks to bitcoind over and over. Seems the orphan didn't want to go down without a fight...
Did we just orphan our own block? We had 105, then 105 orphaned, then 106.
Yes!
Did we just orphan our own block? We had 105, then 105 orphaned, then 106.
OK, who tested his experimental quantum computer running in his basement on p2pool between 5:30PM and 5:35PM GMT today ? 4 blocks in a matter of 3 minutes including one orphan ? Insane luck !
This is at time like this that you can't stop wondering if luck really has no memory...
This is at time like this that you can't stop wondering if luck really has no memory...
Maybe consider lower diff a bit and less longpool messages? I haven't share over last 24hrs...
I have only ~170MH/s and most of my worker get works are break by longpool reset.
In description it is about 10s between longpolls but in most cases i get longpool restart every 5s...
Also after upgrading to 11.1 i`m getting ALOT of dead on arrival (about 30% vs under 5% on 10.3), same worker and config.
Since you have a small rig, it might be worth mining at p2pmining.com. He runs a public, merged mining p2pool node with diff 1 shares designed for smaller miners.I have only ~170MH/s and most of my worker get works are break by longpool reset.
In description it is about 10s between longpolls but in most cases i get longpool restart every 5s...
Also after upgrading to 11.1 i`m getting ALOT of dead on arrival (about 30% vs under 5% on 10.3), same worker and config.
To get difficulty 1 share at p2pmining.com add a +1 to the end of your username/payout address. Larger miners may find it work better to submit higher difficulty shares.
They're coming so fast now we managed to orphan one of our own at 179105!
http://blockchain.info/orphaned-blocks
http://blockchain.info/orphaned-blocks
If we don't hit another block for a week I'm going to kill you. Never mention good luck. Ever. It's the first rule. LOL.
They're coming so fast now we managed to orphan one of our own at 179105!
http://blockchain.info/orphaned-blocks
http://blockchain.info/orphaned-blocks
Woohoo! I finally got my first p2pool block. I had been starting to get worried because my rough estimates of my personal expected time to block had already come and gone. Even when you know it is an estimate, and an estimate of the 50% probability time, no less, I've gotta say, it can still be hard not to get worried.
Just noticed the warning : block might have been orphaned from the chain... So I assume it's OK.
p2pool bug ?
If I'm not mistaken, p2pool found a block yesterday at 21:04 GMT but my node doesn't report it in the last blocks see original page here if you go to it quickly enough :
http://linode.bouton.name:9332/static/
Screencap coming later.
If I'm not mistaken, p2pool found a block yesterday at 21:04 GMT but my node doesn't report it in the last blocks see original page here if you go to it quickly enough :
http://linode.bouton.name:9332/static/
Screencap coming later.
Also, I have a quick question, if you solve the bitcoin block do you also solve a namecoin block at the same time? I'm pretty sure that is what merged mining means.
Any Bitcoin block will definitely (at least for now) be a Namecoin block, because Bitcoin's difficulty is higher than Namecoin's.
I didn't find any for about a month or so, then I got 2 towards the end of april, and two more with each block I've found recently. I think I may have found one or two more way back when.
But I just sell it off right away. Still, it is a nice little bonus.
But I just sell it off right away. Still, it is a nice little bonus.
lucky, i barely find any. so you got like 200 nmc?
Yeah, I knew that alt mining wasn't pooled, but so far every time I have found a bitcoin block, I also found a namecoin block.
Of course, I've only found like, 3 or 4, so the sample size is pretty small, but still.
Of course, I've only found like, 3 or 4, so the sample size is pretty small, but still.
Jump to: