p2pool - Decentralized, Absolutely DoS-Proof, Pool Hopping-Proof Pool [archival]

[m3ta]

After 36 hours of port correctly forwarded, still only connected to 10 peers.
I think I reached the "I only believe it if I see it" phase regarding more than 10 peers being connected.

[1715394001]

1715394001

[1715394001]

1715394001

[1715394001]

1715394001

[fehknt]

I currently have 30 peers reported.  Just reset to update to the latest from git - hit 10 peers quickly, after couple minutes it's up to 13.  So it CAN happen...

[DeathAndTaxes]

Also I would use a port scanning tool to make sure the port is open and forwarded and not getting hung up somewhere between the wide open internet and the local daemon. 

I find http://www.canyouseeme.org/ useful for a sanity check.

[dirtycat]

After 36 hours of port correctly forwarded, still only connected to 10 peers.
I think I reached the "I only believe it if I see it" phase regarding more than 10 peers being connected.

im with you.. I have NEVER seen more than 10.

[rjk]

After 36 hours of port correctly forwarded, still only connected to 10 peers.
I think I reached the "I only believe it if I see it" phase regarding more than 10 peers being connected.

im with you.. I have NEVER seen more than 10.

Same... perhaps its it choking on NAT? While I have a NAT in place, it has multiple public IPs, one of which is assigned to the server I was using - basically each and every port is forwarded. This is instead of assigning the server a public IP, because it makes internal networking easier for me.

[broken]

After 36 hours of port correctly forwarded, still only connected to 10 peers.
I think I reached the "I only believe it if I see it" phase regarding more than 10 peers being connected.

im with you.. I have NEVER seen more than 10.

I have the same problem. I noticed that when i get a new ip address (24h disconnect) my bitcoin and namecoin connections drop to 8 and stay there untill i restart them. But that doesnt work for p2pool. Only one time I had 14 peers in p2pool, but never again.


edit: We need IPv6   (No more NAT)

[btc_artist]

It would be really cool if this were integrated directly into cgminer.  forrestv what do you think of that?  Have you had any dialog with ckolivas at all? 

[m3ta]

There just aren't many users currently (only me at the moment). Can you post any Python error messages you see?

Only you? wth, am I invisible?
No errors, just... 10 peers. That's it.

[m3ta]

It would be really cool if this were integrated directly into cgminer.  forrestv what do you think of that?  Have you had any dialog with ckolivas at all? 

+1. I'd be willing to sponsor this.

[m3ta]

Also I would use a port scanning tool to make sure the port is open and forwarded and not getting hung up somewhere between the wide open internet and the local daemon.  

I find http://www.canyouseeme.org/ useful for a sanity check.

Been there, done that.

Code:

Success: I can see your service on xx.247.9.10 on port (9333)
Your ISP is not blocking port 9333

Also, first thing I did was an nmap from an external server. Result: "open". I knew that, but I double-checked anyway.

So.. yeah. Maybe i'm being paranoid, but the whole ">10 peers" really seems utopic/flawed.

[c_k]

Peers really doesn't matter, if you have 10 you're doing just fine

Here's a snapshot of my stats after running for 12 hours with a fresh git checkout before I launched p2pool:

Code:

2012-01-18 07:56:29.497000 Pool: 84459MH/s in 17381 shares (12581/17385 verified) Recent: 3.66% >3089MH/s Shares: 122 (17 orphan, 1 dead) Peers: 12
2012-01-18 07:56:29.497000 Average time between blocks: 0.74 days
2012-01-18 07:56:29.497000 Pool stales: 9% Own: 16±6% Own efficiency: 93±7%
2012-01-18 07:56:36.796000 New work for worker! Share difficulty: 165.577356 Payout if block: 1.576102 BTC Total block value: 50.155000 BTC including 95 transactions

[m3ta]

Peers really doesn't matter, if you have 10 you're doing just fine

Here's a snapshot of my stats after running for 12 hours with a fresh git checkout before I launched p2pool:

Code:

2012-01-18 07:56:29.497000 Pool: 84459MH/s in 17381 shares (12581/17385 verified) Recent: 3.66% >3089MH/s Shares: 122 (17 orphan, 1 dead) Peers: 12
2012-01-18 07:56:29.497000 Average time between blocks: 0.74 days
2012-01-18 07:56:29.497000 Pool stales: 9% Own: 16±6% Own efficiency: 93±7%
2012-01-18 07:56:36.796000 New work for worker! Share difficulty: 165.577356 Payout if block: 1.576102 BTC Total block value: 50.155000 BTC including 95 transactions

I was gonna say "12 peeers? arrghhhhh Envy is a deadly sin!", but if you say 10 is ok, then.. cheers.

[fehknt]

Might be because I have a static IP - back up to 25 peers as of now.  Well, I'm not contributing (much) hashing power but at least I'm giving a little network stability?

[gnar1ta$]

Been running merged for 24+ hrs with cgminer on defaults and I 8 and was a bit shocked at the E, U, and Rejected, which would normally be 120%, 19.5/m and .6%. Can anyone recommend settings changes or do these numbers look good?

Code:

 cgminer version 2.1.2 - Started: [2012-01-16 15:03:27]
--------------------------------------------------------------------------------
 (5s):1356.2 (avg):1353.6 Mh/s | Q:89492  A:3620  R:210  HW:0  E:4%  U:2.18/m
 TQ: 8  ST: 9  SS: 0  DW: 80039  NB: 185  LW: 86049  GF: 1  RF: 0
 Connected to http://10.0.0.2:9332 with LP as user miner2
 Block: 000005c8ec2e9da784e646e5651e59c4...  Started: [18:44:10]

Code:

Pool: http://10.0.0.2:9332
Has long-poll support
 Queued work requests: 89237
 Share submissions: 3822
 Accepted shares: 3612
 Rejected shares: 210
 Reject ratio: 5.5%
 Efficiency (accepted / queued): 4%
 Discarded work due to new blocks: 79812
 Stale submissions discarded due to new blocks: 0
 Unable to get work from server occasions: 1
 Submitting work remotely delay occasions: 0

Here is p2pool - 2 rigs connect here.

Code:

2012-01-17 18:43:14.505698 New work for worker! Share difficulty: 172.908439 Payout if block: 2.024835 BTC Total block value: 50.037500 BTC including 50 transactions
2012-01-17 18:43:16.670359 Pool: 83856MH/s in 17357 shares (17361/17361 verified) Recent: 4.68% >3925MH/s Shares: 461 (28 orphan, 28 dead) Peers: 10
2012-01-17 18:43:16.670527 Average time between blocks: 0.74 days
2012-01-17 18:43:16.670613 Pool stales: 11% Own: 12±3% Own efficiency: 99±3%

[forrestv]

Been there, done that.

Code:

Success: I can see your service on xx.247.9.10 on port (9333)
Your ISP is not blocking port 9333

Also, first thing I did was an nmap from an external server. Result: "open". I knew that, but I double-checked anyway.

So.. yeah. Maybe i'm being paranoid, but the whole ">10 peers" really seems utopic/flawed.

I made my node connect to your address (using the -n option), so you should have at least 11 peers now. Something might be wrong with the share ranking algorithm, though if I look at the address database, it says that your node has only been up for about 22 hours (since unix timestamp 1326773440). Does that sound right..?

It would be really cool if this were integrated directly into cgminer.  forrestv what do you think of that?  Have you had any dialog with ckolivas at all?  

I worked a bit with him to get cgminer and P2Pool working well together. I don't know how useful that would be... A lot of people have multiple machines running cgminer and one P2Pool instance. Why do you want it integrated? I'd think some kind of monitoring tool that started P2Pool and cgminer might be better.

Been running merged for 24+ hrs with cgminer on defaults and I 8 and was a bit shocked at the E, U, and Rejected, which would normally be 120%, 19.5/m and .6%. Can anyone recommend settings changes or do these numbers look good?

P2Pool has characteristics very different from most mining pools. Work is given a difficulty of 8 instead of 1, which explains the decreased U. Work expiring more quickly (every 10 seconds instead of 10 minutes) decreases cgminer's displayed efficiency and increases rejects.
P2Pool's displayed efficiency is very close to 100%, so everything seems okay. Efficiency is the ultimate indicator of how well everything is working.

[m3ta]

I made my node connect to your address (using the -n option), so you should have at least 11 peers now. Something might be wrong with the share ranking algorithm, though if I look at the address database, it says that your node has only been up for about 22 hours (since unix timestamp 1326773440). Does that sound right..?

I got your latest changes from Git, had 2 mns downtime, so yeah, 22 hours at the time of your post.. that is correct.
Interestingly enough, I did notice the "Peers:11 (1 incoming)" this morning, and now, it even got to "Peers: 12 (2 incoming)".

Sooooo, it seems it's .. erm.. "hard" to go over 10 peers, but it can be done.
And if, like c_k said, "10 is enough", then everything is working fine.

[forrestv]

Win32 development snapshot: http://u.forre.st/u/lirlewiy/p2pool_win32_633dd68.zip I've changed the packaging a bit - everything is combined into one executable file now. Tell me if you have any problems with it!

[JWU42]

I might play with this later tonight.  Like what I am reading as it seems to be the perfect mix of solo mining and pooled mining.

[dirtycat]

wow looks like we are up to 100 ghs! and I got 11 peers! woohoo!!

[btc_artist]

It would be really cool if this were integrated directly into cgminer.  forrestv what do you think of that?  Have you had any dialog with ckolivas at all?  

I worked a bit with him to get cgminer and P2Pool working well together. I don't know how useful that would be... A lot of people have multiple machines running cgminer and one P2Pool instance. Why do you want it integrated? I'd think some kind of monitoring tool that started P2Pool and cgminer might be better.

I was thinking it would be nice to just be able to download one program, set your payout address and run it to start mining on P2Pool.

But now that you mention it and I think about it, I'm ok with installing 3 programs (bitcoind, cgminer, p2pool). Like you mention, a monitoring tool that would start and monitor all 3 programs would be nice.