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

[mdude77]

Anyone have a C1 pointing at p2pool?  Mine seems to behave like my S2 pre-fix ... abysmal.

I was expecting the C1 to act like a S3, but it doesn't, it acts like an S2.  Including the annoying "have to set queue to 0 every reboot" problem.

M

I have C1's using P2Pool.  Mine are getting the full 1000GH/s, but they do require the same re-setting of queue after a reboot.  I also set the pseudo share difficulty.  It has been discussed that setting pseudo difficulty does not effect the overall performance, however setting pseudo difficulty will make cgminer and the UI show the expected hashrate.  I use +1028 (If you use +1024, Best Share is not displayed.  But with +1028, best Share works.  I have no idea why.)  Regarding setting the queue, S2/S4/C1 all have the same OS configuration.

Thanks, I was using 518.  Are you using queue 0 or 1?

M

Doing as you said, setting queue to 0, gives me 965GH/s and 9% rejects.  Completely unreasonable, just like my S2.

M

Why am I not surprised by this... Bitmain claims to want to support p2pool... but the only products they've ever released that worked with p2pool out of the gate are the S1 and S3.  Even those products use forked versions of cgminer that until recently was based on 3.x.  Furthermore, their binaries throw away stale shares, which can lead to potential block solutions being tossed.

I'm not able to get my S2 working reasonably either, even with the latest firmware.

M

[1713906157]

1713906157

[1713906157]

1713906157

[1713906157]

1713906157

[MissouriMiner]

The number 1 issue is fixed with latest firmware, which runs cgminer 4.6.1.  When the work restart request comes in, I usually see the flush work within 1 second, then the new block task usually arrives in another second.  Often there are new accepted shares submitted within 2 seconds of the work restart request.  The new cgminer 4.6.1 has brought new life to the S2.  I used to point the S2's to a different pool, while all my other miners used P2Pool.  It's so nice to see them doing the full 1000 GH/s with P2Pool.  I have 2x S2's and 2x C1's, and all 4 of them run neck-and-neck.

Bear in mind if you are pointing an S1/2/3/4 to p2pool with bitmain's default firmware binaries, they throw out stale shares which is potentially disastrous on p2pool because that can throw out a valid block solve. Kano and I have pointed this out to them numerous times but that behaviour is still in their fork.

Is it enough to get rid of this prob if cgminer is replaced with http://ck.kolivas.org/apps/cgminer/antminer/s3/4.6.1-141020/ ?  Or some option in /etc/config/cgminer params line --submit-stales or similiar??

Any info where to get working or updated cgminers to other ants, heres few s1:s still making noise

Yes it's only fixed with my binaries which I've only made for s3 and 4 and kano's binary for s2. There is no option that will fix it in the bitmain fork.

When mining with P2Pool, I see the message below from cgminer 4.6.1 from Bitmain's firmware.  Does this mean that the stale shares are not thrown away?

[2014-11-30 02:45:57] Pool 2 stale share detected, submitting as user requested

[-ck]

When mining with P2Pool, I see the message below from cgminer 4.6.1 from Bitmain's firmware.  Does this mean that the stale shares are not thrown away?

[2014-11-30 02:45:57] Pool 2 stale share detected, submitting as user requested

No, bitmain decided to add extra block detection code themselves so they filter even more at the driver level.

[MissouriMiner]

When mining with P2Pool, I see the message below from cgminer 4.6.1 from Bitmain's firmware.  Does this mean that the stale shares are not thrown away?

[2014-11-30 02:45:57] Pool 2 stale share detected, submitting as user requested

No, bitmain decided to add extra block detection code themselves so they filter even more at the driver level.

Thanks for the info.

[-ck]

When mining with P2Pool, I see the message below from cgminer 4.6.1 from Bitmain's firmware.  Does this mean that the stale shares are not thrown away?

[2014-11-30 02:45:57] Pool 2 stale share detected, submitting as user requested

No, bitmain decided to add extra block detection code themselves so they filter even more at the driver level.

Thanks for the info.

Exact code if you're interested:
https://github.com/bitmaintech/cgminer/blob/master/driver-bitmain.c#L1294

[mahrens917]

I think your formulas are a bit arbitrary. For instance, getwork latency is given way too much importance in the pool formula. And everything is linear.. and so on (latencies are mostly ok  within a range before the problems they cause skyrocket).

Thanks for the comments Naplam.  Based off of what you said I changed the criteria and reduced how much GWL matters to the p2pool score.  In regards to most of my equations being linear, do you have any suggestions on how to make the ranking more accurate?  Perhaps through tiering?

The latest criteria used can be found on the bottom of the p2pool node finder page, nodes.p2pool.co

[mdude77]

I think your formulas are a bit arbitrary. For instance, getwork latency is given way too much importance in the pool formula. And everything is linear.. and so on (latencies are mostly ok  within a range before the problems they cause skyrocket).

Thanks for the comments Naplam.  Based off of what you said I changed the criteria and reduced how much GWL matters to the p2pool score.  In regards to most of my equations being linear, do you have any suggestions on how to make the ranking more accurate?  Perhaps through tiering?

The latest criteria used can be found on the bottom of the p2pool node finder page, nodes.p2pool.co

I agree about the getwork latency comment, that's weighted too high IMHO.  My node (http://96.44.166.190:9332) is ranked at 51% solely because of the getwork latency.  Everything else is stellar including 111% efficiency.  I played with the bitcoind settings quite a bit before I settled on what it's set to now.  I decided getting more transactions was worth a slightly higher getwork latency, and as shown, it didn't affect the efficiency rating at all.

M

[yslyung]

does anyone have the latest guide for making own p2pool on linux ? tried some others dont seem quite to work

[PatMan]

does anyone have the latest guide for making own p2pool on linux ? tried some others dont seem quite to work

Are you using the script or installing everything manually?

[yslyung]

does anyone have the latest guide for making own p2pool on linux ? tried some others dont seem quite to work

Are you using the script or installing everything manually?

used the script provided here http://bitcoin.kyros.info/ but then i get stuck right at the end :
--2014-11-30 08:10:02--  (try: 3)  https://bitfetch.com/static/bootstrap.7z
Connecting to bitfetch.com (bitfetch.com)|176.9.50.103|:443...     

[PatMan]

does anyone have the latest guide for making own p2pool on linux ? tried some others dont seem quite to work

Are you using the script or installing everything manually?

used the script provided here http://bitcoin.kyros.info/ but then i get stuck right at the end :
--2014-11-30 08:10:02--  (try: 3)  https://bitfetch.com/static/bootstrap.7z
Connecting to bitfetch.com (bitfetch.com)|176.9.50.103|:443...     

Yeah, I'm not sure if the links still work as it's no longer being updated or maintained. Suggest you do the set-up manually instead?

[norgan]

geez the share diff is down. pointing some miners back here for a bit.

[yslyung]

does anyone have the latest guide for making own p2pool on linux ? tried some others dont seem quite to work

Are you using the script or installing everything manually?

used the script provided here http://bitcoin.kyros.info/ but then i get stuck right at the end :
--2014-11-30 08:10:02--  (try: 3)  https://bitfetch.com/static/bootstrap.7z
Connecting to bitfetch.com (bitfetch.com)|176.9.50.103|:443...     

Yeah, I'm not sure if the links still work as it's no longer being updated or maintained. Suggest you do the set-up manually instead?

yeps, learning still especially linux ... looking for a noob proof guide

[PatMan]

Yeah, I'm not sure if the links still work as it's no longer being updated or maintained. Suggest you do the set-up manually instead?

yeps, learning still especially linux ... looking for a noob proof guide

Try this one:

https://bitcointalk.org/index.php?topic=62842.0

geez the share diff is down. pointing some miners back here for a bit.

Yup, & 3 blocks today too - a good Sunday 

[windpath]

Was just about to suggest the guide PatMan did... It is a good one:

https://bitcointalk.org/index.php?topic=62842.0

[yslyung]

thx guys, managed to get to bitcoid & downloading blocks. will let it run n see how things progresses.

[mahrens917]

I think your formulas are a bit arbitrary. For instance, getwork latency is given way too much importance in the pool formula. And everything is linear.. and so on (latencies are mostly ok  within a range before the problems they cause skyrocket).

Thanks for the comments Naplam.  Based off of what you said I changed the criteria and reduced how much GWL matters to the p2pool score.  In regards to most of my equations being linear, do you have any suggestions on how to make the ranking more accurate?  Perhaps through tiering?

The latest criteria used can be found on the bottom of the p2pool node finder page, nodes.p2pool.co

I agree about the getwork latency comment, that's weighted too high IMHO.  My node (http://96.44.166.190:9332) is ranked at 51% solely because of the getwork latency.  Everything else is stellar including 111% efficiency.  I played with the bitcoind settings quite a bit before I settled on what it's set to now.  I decided getting more transactions was worth a slightly higher getwork latency, and as shown, it didn't affect the efficiency rating at all.

M

OK, I removed GWL from the score and your node is doing much better now.  I agree it is efficiency and latency which matters and not internal measurements of the system.  I hope I have the proper balance between efficiency and latency.  I am looking to ensure the scoring system is the fairest out there and a true representation of the best node to connect to.

Thanks for the feedback.

[nreal]

> Error submitting primary block: (will retry)
2014-12-01 08:58:52.269638 > Traceback (most recent call last):
2014-12-01 08:58:52.269676 >   File "/usr/lib64/python2.7/site-packages/twisted/internet/defer.py", line 490, in _startRunCallbacks
2014-12-01 08:58:52.269704 >     self._runCallbacks()
2014-12-01 08:58:52.269738 >   File "/usr/lib64/python2.7/site-packages/twisted/internet/defer.py", line 577, in _runCallbacks
2014-12-01 08:58:52.269780 >     current.result = callback(current.result, *args, **kw)
2014-12-01 08:58:52.269832 >   File "/usr/lib64/python2.7/site-packages/twisted/internet/defer.py", line 1155, in gotResult
2014-12-01 08:58:52.269856 >     _inlineCallbacks(r, g, deferred)
2014-12-01 08:58:52.269880 >   File "/usr/lib64/python2.7/site-packages/twisted/internet/defer.py", line 1099, in _inlineCallbacks
2014-12-01 08:58:52.269913 >     result = g.send(result)
2014-12-01 08:58:52.269954 > --- <exception caught here> ---
2014-12-01 08:58:52.269987 >   File "/mnt/60gb/p2pool/p2pool/util/deferral.py", line 41, in f
2014-12-01 08:58:52.270028 >     result = yield func(*args, **kwargs)
2014-12-01 08:58:52.270059 >   File "/mnt/60gb/p2pool/p2pool/bitcoin/helper.py", line 67, in submit_block_p2p
2014-12-01 08:58:52.270112 >     factory.conn.value.send_block(block=block)
2014-12-01 08:58:52.270136 >   File "/mnt/60gb/p2pool/p2pool/util/p2protocol.py", line 102, in <lambda>
2014-12-01 08:58:52.270169 >     return lambda **payload2: self.sendPacket(command, payload2)
2014-12-01 08:58:52.270201 >   File "/mnt/60gb/p2pool/p2pool/util/p2protocol.py", line 93, in sendPacket
2014-12-01 08:58:52.270224 >     raise TooLong('payload too long')
2014-12-01 08:58:52.270247 > p2pool.util.p2protocol.TooLong: payload too long

Whatsup?


It became orphan because payload too long? Too many users on pool or too many transactions in a block or what?

[jedimstr]

I think your formulas are a bit arbitrary. For instance, getwork latency is given way too much importance in the pool formula. And everything is linear.. and so on (latencies are mostly ok  within a range before the problems they cause skyrocket).

Thanks for the comments Naplam.  Based off of what you said I changed the criteria and reduced how much GWL matters to the p2pool score.  In regards to most of my equations being linear, do you have any suggestions on how to make the ranking more accurate?  Perhaps through tiering?

The latest criteria used can be found on the bottom of the p2pool node finder page, nodes.p2pool.co

I agree about the getwork latency comment, that's weighted too high IMHO.  My node (http://96.44.166.190:9332) is ranked at 51% solely because of the getwork latency.  Everything else is stellar including 111% efficiency.  I played with the bitcoind settings quite a bit before I settled on what it's set to now.  I decided getting more transactions was worth a slightly higher getwork latency, and as shown, it didn't affect the efficiency rating at all.

M

OK, I removed GWL from the score and your node is doing much better now.  I agree it is efficiency and latency which matters and not internal measurements of the system.  I hope I have the proper balance between efficiency and latency.  I am looking to ensure the scoring system is the fairest out there and a true representation of the best node to connect to.

Thanks for the feedback.

The latest algorithm change definitely boosted my node's standing (currently 112% score in your list).

Could you change the listing from the IP  71.172.92.76 to the host name galactica.geekgalaxy.com?

Also, the algorithm FAQ at the bottom still lists GWL.

[PatMan]

I was comparing your results with http://p2pools.org/btc & I noticed forre.st node had been down for over 50 days - has he completely abandoned his node now as well? Or has he just changed his domain?