[1500 TH] p2pool: Decentralized, DoS-resistant, Hop-Proof pool

[bitpop]

I have 15 incoming
http://bitpoppool.geekgalaxy.com:9332/static/graphs.html

[manifold]

Ahh.. ok no i have no static ip address and I don't even run 24/7... ;-)

[Panda Mouse]

I have 15 incoming
http://bitpoppool.geekgalaxy.com:9332/static/graphs.html

How you put yours data (graphs) to outside server?
I have it only locally

Panda Mouse

[bitpop]

I forward the port and use ddns

[rudrigorc2]

great, using it from now on!

[DeathAndTaxes]

Looks like the people who where not upgraded got kicked from the network.

Well technically they are on an incompatible and minority fork.  That fork can continue to exist forever but hopefully they notice the 70%+ (from their point of view) drop in hashing power and take the time to actually upgrade their p2pool once a month or so.

[DeathAndTaxes]

Ahh.. ok no i have no static ip address and I don't even run 24/7... ;-)

static IP isn't necessary as long as your IP address is relatively stable.  Looking at the incoming peers chart it looks like >8 hours is where # of incoming peers rapidly increases.  For most users DHCP leases tend to be 14 days or longer (I have the same dynamic IP address as I did last year) so there is no need for a static IP address.

Keeping p2pool running is important.  nodes rank other nodes based on how long they have seen them.  If you are always seen as a very short lived and unstable node it is unlikely any node will ever voluntarily connect to you (as there are other better options available).  Nodes are oppertunistic (as they should be).   They seek out the best possible nodes.

Currently looking at the code it looks like longevity is the only factor is how "good" a node is.  That probably could be expanded to include things like latency, share notification (what % of shares did that node report), bad shares (nodes should verify shares before forwarding), version reported, etc.

One thing to keep in mind if that even if you don't mine 24/7 you can keep p2pool & bitcoind running 24/7 (both use minimal resources & bandwidth).

Some people suggested that a node has to be "reliable" for people to connect to it. The node is running 24/7, but the internet connection reconnects every 24 hours. Maybe this is the reason.

If the IP address changes then other nodes will see you as a new 0 minute node.  Nodes identify other nodes based on IP address.  Theoretically that could be changed but I don't think it is a pressing concern.  One way I could see that changing is by nodes identifying themselves by a public key.  This would allow nodes to retain information on other nodes even when IP address changes.

A simpler patch would be to have the node list include dns entries.  Then using dyndns one could always be the same url despite IP address changing.  Personally I like the public key method better (prevents the need for frequent dns lookups) but using dns is much simpler.

[DeathAndTaxes]

Some people suggested that a node has to be "reliable" for people to connect to it. The node is running 24/7, but the internet connection reconnects every 24 hours. Maybe this is the reason.

If the IP address changes then other nodes will see you as a new 0 minute node.  Nodes identify other nodes based on IP address.  Theoretically that could be changed but I don't think it is a pressing concern.

[kano]

Some people suggested that a node has to be "reliable" for people to connect to it. The node is running 24/7, but the internet connection reconnects every 24 hours. Maybe this is the reason.

If the IP address changes then other nodes will see you as a new 0 minute node.  Nodes identify other nodes based on IP address.  Theoretically that could be changed but I don't think it is a pressing concern.

Hmm interesting - my IP address changes every day also.

So does this mean that new connection nodes are effectively discouraged?
If your description above is correct then they are discouraged.
I guess there's some technical details that would explain that?

[Panda Mouse]

I forward the port and use ddns

Many thanks, but can you explain how to do this in windows 7?

Panda Mouse.

[forrestv]

I forward the port and use ddns

Many thanks, but can you explain how to do this in windows 7?

Panda Mouse.

It should be as simple as forwarding port 9332 through your router to your computer and then going to http://YOUR.IP.ADDR.ESS:9332/static from outside.

[bitpop]

Or get a ddns from no-ip.com

[Panda Mouse]

I forward the port and use ddns

Many thanks, but can you explain how to do this in windows 7?

Panda Mouse.

It should be as simple as forwarding port 9332 through your router to your computer and then going to http://YOUR.IP.ADDR.ESS:9332/static from outside.

Yes, thanks a lot, greetings from Poland.

Panda Mouse.

[DeathAndTaxes]

Donation Sent.
tx id: 3b728fab90a603507b35acfca3cf9bd96c67ac02825793108c8bfe3963c45b87

Glad to see p2pool made it through the deadline without issue.

A minor request if I can make one. 
On the graphs page would it be possible to add "current" next to the mean value.  Even better would be a min, max, avg, and current values.

Code:

Local Hashrate:
Current: 11.8 GH/s
Min: 10.4 GH/s
Max: 12.8 GH/s
Mean: 11.6 GH/s

Also adding an expected daily payout after Current Payout Value would be a nice addition.
Essentially it would be (Current Payout Value)*(24 hours) / (expected block time in hours)

[kano]

...
Glad to see p2pool made it through the deadline without issue.
...

It didn't - see block 174053 (~2.5 hours ago)
Of course, the meaning of that is that someone using P2Pool hasn't upgraded.
So yeah it's a bit hard to say "p2pool made it through" since that isn't true until everyone using p2pool has upgraded.

[twmz]

...
Glad to see p2pool made it through the deadline without issue.
...

It didn't - see block 174053 (~2.5 hours ago)
Of course, the meaning of that is that someone using P2Pool hasn't upgraded.
So yeah it's a bit hard to say "p2pool made it through" since that isn't true until everyone using p2pool has upgraded.

270 GH/s or 75% of the pre-change mining hashpower made it through.  I think 75% is plenty enough to call it "made it through".  And block 174053 is just proof that isolating those that have abandoned their mining rigs was the right thing to do.  If they hadn't been isolated those miners would have created a useless block while still getting paid in other blocks by the other 75% of the p2pool miners for their useless shares.

Heck, there is probably a small peer to peer network of people that never upgraded after the last required upgrade.  Getting everyone to upgrade is an unachievable goal.  Getting "most" upgraded is enough.

[tenzor]

What happens to this block?
http://blockchain.info/block-height/174118

It appear for a minute in p2pool.info, but...


EDIT: oh, its a unupdated fork...

[Panda Mouse]

I have big problem now, after install AMD 12.3 for windows, all looks ok but in run_p2pool.exe (0.10.3) shares are still 0, efficiency .

What is the problem? Driver?

Thanks in advance Panda Mouse.

[Soros Shorts]

No blocks solved for 24 hours?

(I am on the 275GH/s fork)

[bitpop]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Yeah I havent gotten a payout either but before that we were on a roll
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.17 (MingW32)
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=VEsS
-----END PGP SIGNATURE-----