Topic: OLD: BFGMiner 3.10.0: modular ASIC+FPGA, GBT+Strtm, RPC, Mac/Lnx/W64, AntU1, DRB - page 127. (Read 1193231 times)

i`ve been back and forth too, trying to get my BE working with CG but they just won`t (maybe due to my inexperience compiling via mingw or something)

BFGminer just works straight away so I`m sticking.

Been back and forth testing CGminer and BFGminer. Since I got my BFL Asics I've notice the CGminer builds take about 90-95% cpu usage on my Rpi and BFGminer gets same performance at ~20% CPU usage. The verdict is in BFGminer is much more efficient for my Rpi.

After I decide to write stratum proxy support (or someone else contributes it), at least 

when to expect stratum proxy enabled for OpenWRT?

If you are using the latest version of bfg, I dont think you need "--icarus-options 115200:1:1 --icarus-timing 3.0=100", you can delete that.

I had a BEE/BES issue like that, but it was a couple of versions back.  Have had zero problems with version 3.2.0

Thanks for getting back so quick Luke.  Whenever I try with -S erupter:all    NO    --icarus     
     
          -I start getting sick/dead within 10-20 minutes.

Whenever I try    -S erupter:noauto     
     
        - then only the blade shows up   ----No erupters

I got --scan-serial erupter:*   from the  ASIC readme
   
          -I guess I misunderstood the usage.

The only way I've been able to get them running close to as stable as before is by using a similar command to the one I have been using since I started setting up these AsicMiner devices on BFGminer......

bfgminer.exe -o 127.0.0.1:8332 -u worker99 -p pass -o gbt.mining.eligius.st:9337 -u ITSYOURS -p pass --failover-only -G -S all --icarus-options 115200:1:1 --icarus-timing 3.0=100 --coinbase-addr 14qi3gVPuDzcuTiJpdHCoTETFFFvG4SeD3 --coinbase-sig "BUYER99 SOLO MINE"

I'm mining some AltCoins via P2Pools and would like to disable the automatic increasing of queue size.

Is there a way to do that?

When the Pi is heavily loaded it needs a good clean power supply.
Mine was forever stalling and restarting.
I now have it connected through a blackberry charger, on its own power outlet, and it has behaved for over a week without crashing.

Luke,

I was wondering if you could add JSONP commands to your APIs.
JSONP extension of your APIs would allow to access bfgminer directly from any static (local to browser) page.
AJAX calls can be made from the browser directly to API port.  No cross-domain issues.

I got this idea while extending my akbash interface to support JSONP (in next version) to monitor multiple instances of bfgminer on different computers/sites.

If bfgminer would return JSONP, one static/AJAX page can aggregate/format output from multiple bfgminer instances from remote sites.
Such page can be loaded in the browser via "file://".  No need to run web servers, or third party apps.

I think it will be a nice extension to the API.

af_newbie

This is pretty odd. They really should all show up as BES (Block Erupter Sapphire), but I'm curious as to why that is creating problems for you.
What options are you using? Can you try with just -S erupter:all and not any --icarus-* options, in the same instance?
Temporary workaround in any case: -S erupter:noauto will disable the autodetection of autodetect-capable Erupters (likely the 3 that show up as BES for you).

* is not a special character with --scan-serial, and will be ignored.

-S erupter:all

None. 


I doubt it - considering the little TP-Link I've got running bfgminer with 11 Erupters, with it's Casio calculator CPU, I think the dual core RPi will run both BFG and the getwork server no problems.

I relied on my friend who just said "same with BFGminer".
Another 60GH/s Single works just fine on the same computer, that's why I assumed it was a device problem.

Video works fine for me, haha.

Which switch are you using?  "-S all"   or   "-S erupter:all"   or   "--scan-serial erupter:*"      

Icarus flags or no?

Now that I moved the 3 problem children to the windows box and using   -S all    they do show up as ICA on there and have given me zero problems..

It has always showed me 3 as BES and the rest as ICA on the pi.

Now that I have moved the 3 "oddballs" off of the pi everything has been rolling smootly but then, all of sudden about 15 of them went to ERR at the same time and all restarted.. They did ALL restart fine though.

I realize it could be power or hubs or cooling or one bad erupter or a driver or etc etc... Just trying to find out more if I can since it has been so rock solid up until this point. I have been swapping ports & hubs & power adapters & erupters.

Maybe running the getwork server and the erupters is a little too much for the pi?

As for the queue they both (win7 miner and pi miner) have queued up, but only a little  2-4.

Wrong SDK versions.  You need to wipe out your ATI drivers totally, and install 12.8 fresh.

My sticks never come up as ICA - they either come up as BEE or BES.  BES works best, I usually expect them to stop working when they report as BEE, and when they do it's almost always cured by changing the USB port the hub is plugged in to.

Under Ubuntu 13.04 Server, the hubs worked best plugged in to the USB3 ports - under Windows best in USB2 ports.  Same hubs, same machine.

The staged underrun thing seems normal, I get that here too.  Mine usually ends up queuing 4 or 8.

Very cool, I've been reading your posts hoping it would work out.

Congratulations HellDiver!


Luke, I want to thank you for what you have done and have sent a small donation your way. I recently rebuilt BFGminer on my pi with 3.2.0 and now with the new getwork option I have my USB's and my blade all happy together in one interface!

My only issue at this point is for some reason 3 out of my 25 sticks show up as BES instead of ICA and are giving me sick/dead restart problems. Before I changed to this version, I have been stable since I set it up. Impressive to say the least. I definitely want my blade connected through BFGminer... So I have isolated the 3 erupters that show up differently and pulled them, now BFG is running like a champ. Not a big deal since my windows machine is always running with my node and so I can just put the 3 "BES"sticks on there. Just thought I would add that information in case someone else is having a similar issue.

 FYI...

I have tried with -G switch and without.
I have tried with the icarus timings and without
 I have tried -S erupter:all   and   --scan-serial erupter:*  

-S all   with the icarus timings  seems to give me the best results, but after a while the 3 sticks which show up as BES go to sick and sometimes dead, also randomly causing problems with other sticks in the process of restarting themselves etc.  Staged work underrun will not increase automatically pass 36 keeps coming up every 2-3 min..  So far since I have removed those 3 from the equation, no messages about staged overrun and it looks like all is well. Hardware errors are also very low... Thanks for everything you do!

Edit: Still no issues, stable like it always was but just got this message "staged work underrun; increasing queue minimum to 2" Doesn't seem to be a problem, I'm sure it has said this many times before (but it's been running so good I haven't been watching it at all) and sorry for being such a newb but google doesn't want to tell me what it means.

  I run bfgminer and cgminer on new install and when i start it it comes up with bfgminer has experince an error and needs to close any idea on why this is, I have install bitminter and it runs fine. Cards run fine on linux just can't get them to work with windows. windows will give me a 120 k/h increase in hash rate. cards have worked before with windows. no way i flashed pcb board to only function with linux is there?

That worked.  Pleased to announce BFG is happily mining away with 11 Block Erupters on a TL-WR703N with a 4GB USB stick.

RAM is a bit tight, but I think it'd run the HTTP proxy thingymabob for the Blades, too, if a nice developer could compile it in. 

opkg install kmod-usb-serial-cp210x