Modular Python Bitcoin Miner - Official Thread

[TheSeven]

Hm, that looks like a Python or OpenUSB/LibUSB/PyUSB bug. You might want to try a different Python/PyUSB version.

[1714008203]

1714008203

[coblee]

Hm, that looks like a Python or OpenUSB/LibUSB/PyUSB bug. You might want to try a different Python/PyUSB version.

Ok, just FYI. This worked with the previous version of MPBM. So the new web-based version is doing something different.

[coblee]

Hm, that looks like a Python or OpenUSB/LibUSB/PyUSB bug. You might want to try a different Python/PyUSB version.

Ok, just FYI. This worked with the previous version of MPBM. So the new web-based version is doing something different.

I tried both pyusb-1.0.0-a0 and pyusb-1.0.0a2 and neither worked. I switched back to v0.0.4 and my x6500 boards are working again. Any ideas?

Honestly, I like v0.0.4 better. It's much easier to spot if anything is wrong with my miners. The latest has a lot of stats, but most of them don't mean much to me. So it makes it harder for me to find the important ones like hashrate and stales. You should consider make it adjustable so I could hide the stats I don't care about.

[TheSeven]

Hm, that looks like a Python or OpenUSB/LibUSB/PyUSB bug. You might want to try a different Python/PyUSB version.

Ok, just FYI. This worked with the previous version of MPBM. So the new web-based version is doing something different.

I tried both pyusb-1.0.0-a0 and pyusb-1.0.0a2 and neither worked. I switched back to v0.0.4 and my x6500 boards are working again. Any ideas?

Honestly, I like v0.0.4 better. It's much easier to spot if anything is wrong with my miners. The latest has a lot of stats, but most of them don't mean much to me. So it makes it harder for me to find the important ones like hashrate and stales. You should consider make it adjustable so I could hide the stats I don't care about.

If it works with v0.0.4 this points towards an issue with forking. In v0.0.x everything was running in a single process, since v0.1.x the computationally intensive parts of some worker modules (x6500 and ztex) have been moved to out-of-process board proxys, which are created by forking the main process. This was neccessary to allow MPBM to use multiple processor cores for bigger clusters.

I know that the UI needs some improvement, however I'm pretty busy these days and thus focus on a stable backend for now. Feel free to submit a patch

[freshzive]

Hm, that looks like a Python or OpenUSB/LibUSB/PyUSB bug. You might want to try a different Python/PyUSB version.

Ok, just FYI. This worked with the previous version of MPBM. So the new web-based version is doing something different.

I tried both pyusb-1.0.0-a0 and pyusb-1.0.0a2 and neither worked. I switched back to v0.0.4 and my x6500 boards are working again. Any ideas?

Honestly, I like v0.0.4 better. It's much easier to spot if anything is wrong with my miners. The latest has a lot of stats, but most of them don't mean much to me. So it makes it harder for me to find the important ones like hashrate and stales. You should consider make it adjustable so I could hide the stats I don't care about.

No matter what I do, I can't get this to work correctly on a mac. Tried different versions of python/pyUSB/libusb, but nothing. 0.4 runs beautifully out of the box, guess I'll stick with that for now.

[coblee]

Hm, that looks like a Python or OpenUSB/LibUSB/PyUSB bug. You might want to try a different Python/PyUSB version.

Ok, just FYI. This worked with the previous version of MPBM. So the new web-based version is doing something different.

I tried both pyusb-1.0.0-a0 and pyusb-1.0.0a2 and neither worked. I switched back to v0.0.4 and my x6500 boards are working again. Any ideas?

Honestly, I like v0.0.4 better. It's much easier to spot if anything is wrong with my miners. The latest has a lot of stats, but most of them don't mean much to me. So it makes it harder for me to find the important ones like hashrate and stales. You should consider make it adjustable so I could hide the stats I don't care about.

No matter what I do, I can't get this to work correctly on a mac. Tried different versions of python/pyUSB/libusb, but nothing. 0.4 runs beautifully out of the box, guess I'll stick with that for now.

Yes, I'm on a Mac too. And I've reverted back to 0.0.4. TheSeven, let me know if there's anything we can do to help get this fixed for the Mac.

[hashking]

It seems like Im getting a very high rate of rejects on gpumax.  Over 10%.  Is this normal.

[TheSeven]

It seems like Im getting a very high rate of rejects on gpumax.  Over 10%.  Is this normal.

It definitely isn't.
There were some GPUMAX issues that have been fixed in the past, but it looks like we missed some :/
However, as the GPUMAX guys refuse to send me an invite since months, I can't really do anything about that.

[LoWang]

Yes, I'm on a Mac too. And I've reverted back to 0.0.4. TheSeven, let me know if there's anything we can do to help get this fixed for the Mac.

Hello Coblee I wrote you several times in the offical LTC thread, then I PM you but you still don't respond. I wrote you private message on #litecoin IRC but you ignored it. Are you that busy or is there some other reason why you ignore me? Thank you

[coblee]

Yes, I'm on a Mac too. And I've reverted back to 0.0.4. TheSeven, let me know if there's anything we can do to help get this fixed for the Mac.

Hello Coblee I wrote you several times in the offical LTC thread, then I PM you but you still don't respond. I wrote you private message on #litecoin IRC but you ignored it. Are you that busy or is there some other reason why you ignore me? Thank you

Sorry for ignoring you, I will send you PM.

[hashking]

It seems like Im getting a very high rate of rejects on gpumax.  Over 10%.  Is this normal.

It definitely isn't.
There were some GPUMAX issues that have been fixed in the past, but it looks like we missed some :/
However, as the GPUMAX guys refuse to send me an invite since months, I can't really do anything about that.

I can let you borrow my gpumax account so you can fix the issues with it.

[Garr255]

I already offered, he seems convinced to convince the gpumax guys to give him one of his own

[hashking]

This would be a very important item to fix for this miner.  Would a bounty to fix this speed up things.  

[BR0KK]

Im preparing to get one of the newly announced Ztex 1.15y Quad Spartan Boards. If i have it in my hands you could have a look at it when u like (for implementing support into MPBM) As i described i could offer u Remote Access to all of my Boards via Teamviewer etc

€:
Got one yesterday and its hashing nicely since then

Could you get anything new about Ztex and your licensing problem?

[hashking]

Here is the error I keep getting with GPUMAX.  Can you please fix this.  Thanks.


2012-05-01 21:30:15.310   [200]   GPUMAX: Error while fetching job: Traceback (most recent call last):
  File "C:\Users\Desktop\X6500_Easy_Package_For_Windows\mpbm-v0.1.0beta\core\actualworksource.py", line 155, in get_job
    jobs = self._get_job()
  File "C:\Users\Desktop\X6500_Easy_Package_For_Windows\mpbm-v0.1.0beta\modules\theseven\bcjsonrpc\bcjsonrpcworksource.py", line 120, in _get_job
    conn.request("POST", self.settings.path, req, headers)
  File "c:\Program Files (x86)\Python32\lib\http\client.py", line 964, in request
  File "c:\Program Files (x86)\Python32\lib\http\client.py", line 1002, in _send_request
  File "c:\Program Files (x86)\Python32\lib\http\client.py", line 960, in endheaders
  File "c:\Program Files (x86)\Python32\lib\http\client.py", line 805, in _send_output
  File "c:\Program Files (x86)\Python32\lib\http\client.py", line 743, in send
  File "c:\Program Files (x86)\Python32\lib\http\client.py", line 721, in connect
  File "c:\Program Files (x86)\Python32\lib\socket.py", line 386, in create_connection
socket.gaierror: [Errno 11004] getaddrinfo failed

[TheSeven]

Here is the error I keep getting with GPUMAX.  Can you please fix this.  Thanks.


2012-05-01 21:30:15.310   [200]   GPUMAX: Error while fetching job: Traceback (most recent call last):
  File "C:\Users\Desktop\X6500_Easy_Package_For_Windows\mpbm-v0.1.0beta\core\actualworksource.py", line 155, in get_job
    jobs = self._get_job()
  File "C:\Users\Desktop\X6500_Easy_Package_For_Windows\mpbm-v0.1.0beta\modules\theseven\bcjsonrpc\bcjsonrpcworksource.py", line 120, in _get_job
    conn.request("POST", self.settings.path, req, headers)
  File "c:\Program Files (x86)\Python32\lib\http\client.py", line 964, in request
  File "c:\Program Files (x86)\Python32\lib\http\client.py", line 1002, in _send_request
  File "c:\Program Files (x86)\Python32\lib\http\client.py", line 960, in endheaders
  File "c:\Program Files (x86)\Python32\lib\http\client.py", line 805, in _send_output
  File "c:\Program Files (x86)\Python32\lib\http\client.py", line 743, in send
  File "c:\Program Files (x86)\Python32\lib\http\client.py", line 721, in connect
  File "c:\Program Files (x86)\Python32\lib\socket.py", line 386, in create_connection
socket.gaierror: [Errno 11004] getaddrinfo failed

This looks like a GPUMAX-side or configuration problem that I can't do anything about.
It basically means that the DNS name that the work source is using could not be resolved.
Judging from the GPUMAX thread it has been pretty much dead recently, with most IP addresses locked out, so could it be just that?

[nbtcminer]

Can anyone help me out with some reqs / initial setup instructions on how to get a BFL single to work with MPBM 0.1.0? Thanks in advance for the help!

[TheSeven]

Can anyone help me out with some reqs / initial setup instructions on how to get a BFL single to work with MPBM 0.1.0? Thanks in advance for the help!

Which operating system?

On Linux it's mostly "plug the BFLs, unzip MPBM, start it, configure the pools etc.".

[nbtcminer]

Can anyone help me out with some reqs / initial setup instructions on how to get a BFL single to work with MPBM 0.1.0? Thanks in advance for the help!

Which operating system?

On Linux it's mostly "plug the BFLs, unzip MPBM, start it, configure the pools etc.".

Sorry forgot to add that important part in; Windows 7 Pro X64

[TheSeven]

Install the FTDI drivers (just like for any other BFL miner software), and then either:
1. install python and pyserial, clone mpbm git
2. use prebuilt package from the first post of this thread (a bit outdated right now)