Introducing CherryPicking - new Windows & Linux Pool Hopper

[scatterbrain]

i meant to say it started on the 24-25th since someone else reported that they had seen issues from the previous night

[Folax]

I can confirm that

Code:

Error reading miner output stream, terminating.

on Linux means that NO shares are being submitted. I will send log to Blood.

[Bloodred]

I have a possible fix that I'll release, but since I can't reproduce the bug on my systems (either Windows or Linux), it may or may not solve it. Did anybody else run into this problem on Linux? The more information I get the faster I'll be able to find a fix.

[scatterbrain]

this might have been brought up in an earlier post, but why isn't CP hopping btc guild. it appears that bithopper and another new poclbm hopper both hop it, please correct me if i'm wrong. are their stats delayed too long?

[Milkshanks]

this might have been brought up in an earlier post, but why isn't CP hopping btc guild. it appears that bithopper and another new poclbm hopper both hop it, please correct me if i'm wrong. are their stats delayed too long?

I'd also like to know that. Also, what about bitcoins.lc? Anyone else having that connection problem?

[scatterbrain]

if you mean the java security certificate issue with bitcoinlc, just remove the "s" after http in the cfg file

[pitabread]

if you mean the java security certificate issue with bitcoinlc, just remove the "s" after http in the cfg file

jinx 

[Milkshanks]

if you mean the java security certificate issue with bitcoinlc, just remove the "s" after http in the cfg file

LoL so simple. Thanks.

[Bloodred]

this might have been brought up in an earlier post, but why isn't CP hopping btc guild. it appears that bithopper and another new poclbm hopper both hop it, please correct me if i'm wrong. are their stats delayed too long?

Yeah, basically the stats are delayed too long. There are some methods that could help in "guessing" when to hop on btcguild (or deepbit) but they aren't very practical with CherryPicking since the whole point of how CherryPicking works was (is) to not mess with direct miner traffic. I have some ideas but I don't know exactly how practical they are (or if they'll make the release version of CP at all).

[scatterbrain]

i've been looking at my mining data and continue to be very impressed with CP. blood i must thank you for doing such a great job. with roughly a month of data to look at, 1 Ghash/s is averaging .75 BTC/day. the last 5 days alone 1 Ghash/s has been averaging almost .9 BTC/day. anyone else with some comparison numbers? it would be awesome if i was the low to mid end of the spectrum.

[organofcorti]

New "how to hop" blog post:

How to hop part2: More on score

This week we discover Slush-type score systems fatal flaw, an update on the Slush hop point (previous estimate nearly 50% out), and what 'c' really means, and how to make a great meatloaf using only left overs!

[kdf]

I am having a couple of issues in getting CP to work on a couple of pools.

One is NoFee and the other is BTCServer.  Neither one will pull stats and I made sure the config file is up to date with the information posted on this site.

Another issue I am trying to work through is if CP is have a problem running 5 different GPU's on one computer with Win7 x64.  I keep experiencing a system crash after about 2 hours of run time.  All my card temps are 61 - 65.  All are running stock voltage.

Card 1: 6950 shaders unlocked 850/400
Card 2: 6870 960/300
Card 3: 6870 960/300
Card 4: 5970 800/300 both gpus

I am running an Athalon X2 with affinity set to core 1
2gb RAM
PSU is Anatec 1200 pro High Current

I have one cable extension with a molex adapter.

I am running the cards a little slower than I did with guiminer...

Any ideas?

Thanks,

[scatterbrain]

I have 5 cards running on win7 x64 as well with no problems. My guess is the 6950 with unlocked shaders is crashing. I have 2 6950s that worked fine with unlocked shaders until I used them for mining, then they would crash after a few hours. I had to switch back to stock bios to keep them working while mining. Try running just your 6950 solo and see if your system still crashes. If not just go through your cards solo mining 1 by 1 until you find the one that's unstable

Nofee has never worked for me either. I can connect to them but all I get is rpc errors so I dropped them altogether. I think there is an updated cfg file for btcserv a few pages back that should work, but I still get a lot of connection problems with them too

[kdf]

Ok so it is good to know that it is not the number of GPU's.

I have been running that 6950 with the unlocked shaders for sometime now in this rig with out issue. Yesterday I took out a 5830 and replaced it with the 5970.  So that is the only turely untested card in the system.  I am doing another run right now and if I get another crash I might try and run the 5970 at stock MHZ but lower the mem clock to 300 and see what happens.

I also noticed that the BIOS on the MOBO had the cpu to run cool and quiet.  I disabled that as with the 100? core cpu usage may have been causing an issue.

One thing I have learned that when you mess with a working system it never just works right away.  You have to dial it in/break it in....

More testing....

[Bloodred]

New "how to hop" blog post:

How to hop part2: More on score

This week we discover Slush-type score systems fatal flaw, an update on the Slush hop point (previous estimate nearly 50% out), and what 'c' really means, and how to make a great meatloaf using only left overs!

Great article, thanks! The next patch of CherryPicking will definitely have the tweaks.

I am having a couple of issues in getting CP to work on a couple of pools.

One is NoFee and the other is BTCServer.  Neither one will pull stats and I made sure the config file is up to date with the information posted on this site.

Another issue I am trying to work through is if CP is have a problem running 5 different GPU's on one computer with Win7 x64.  I keep experiencing a system crash after about 2 hours of run time.  All my card temps are 61 - 65.  All are running stock voltage.

Card 1: 6950 shaders unlocked 850/400
Card 2: 6870 960/300
Card 3: 6870 960/300
Card 4: 5970 800/300 both gpus

I am running an Athalon X2 with affinity set to core 1
2gb RAM
PSU is Anatec 1200 pro High Current

I have one cable extension with a molex adapter.

I am running the cards a little slower than I did with guiminer...

Any ideas?

Thanks,

There's nothing in CherryPicking itself that could crash your system, absolute worst case scenario it could crash a miner and cause one or more of your GPUs to stop mining if it had a serious bug, but that hasn't happened yet. Since it's a system crash I suggest looking at your hardware and drivers, the problem would probably be there.

i've been looking at my mining data and continue to be very impressed with CP. blood i must thank you for doing such a great job. with roughly a month of data to look at, 1 Ghash/s is averaging .75 BTC/day. the last 5 days alone 1 Ghash/s has been averaging almost .9 BTC/day. anyone else with some comparison numbers? it would be awesome if i was the low to mid end of the spectrum.

Happy to hear you're getting good results. Which hopping algorithm have you been using this time?

[max in montreal]

I can confirm that

Code:

Error reading miner output stream, terminating.

on Linux means that NO shares are being submitted. I will send log to Blood.

I am having the exact same issue here.

[scatterbrain]

Been using dynamic fast multiplicative. I would like to try the others but i've been doing so well with this one I don't want to change it

[tal0n4]

I can confirm that

Code:

Error reading miner output stream, terminating.

on Linux means that NO shares are being submitted. I will send log to Blood.

I am having the exact same issue here.

Add me to the list... I just switched over to linux on my mining rig and I'm getting the same error. The error comes up occasionally and CP will continue to run but every so often it will go in to an endless loop with that error message.

[Bloodred]

Please send logs (upload them somewhere and send a PM or just post the link here), it's proving difficult to track.

Also, could anybody tell me if the error loop stops when CP attempts to switch pools again (even if it stops and then starts again)?

LE: What distro and what JRE is everybody using?

[max in montreal]

what is the log file?