You can solo mine with bfgminer... I've done it.  In fact, I think its discussed back around page 122 or so, but see the Bfgminer README (from github), down around line 523... ah heck, here it is:

SOLO MINING

BFGMiner supports solo mining with any GBT-compatible bitcoin node (such as
bitcoind). To use this mode, you need to specify the URL of your bitcoind node
using the usual pool options (--url, --userpass, etc), and the --coinbase-addr
option to specify the Bitcoin address you wish to receive the block rewards
mined. If you are solo mining with more than one instance of BFGMiner (or any
other software) per payout address, you must also specify data using the
--coinbase-sig option to ensure each miner is working on unique work. Note
that this data will be publicly seen if your miner finds a block using any
GBT-enabled pool, even when not solo mining (such as failover). If your
bitcoin node does not support longpolling (for example, bitcoind 0.8.x), you
should consider setting up a failover pool to provide you with block
notifications. Note that solo mining does not use shares, so BFGMiner's adjusted
hashrate (third column) may suddenly drop to zero if a block you submit is
rejected; this does not indicate that it has stopped mining.

Example solo mining usage:

bfgminer -o http://localhost:8332 -u username -p password \
    --coinbase-addr 1QATWksNFGeUJCWBrN4g6hGM178Lovm7Wh \
    --coinbase-sig "rig1: This is Joe's block!"

Which readme?  I'm looking at the 3.10.0 bfgminer ones and not finding what your referencing...

Huh?  Every post I've seen in the past month from Luke has debunked that rumor.  He isn't personally doing the support, but the development team for BFGminer is fully behind Scrypt.  Heck, he still has CPU mining support in the codebase!

./bfgminer --help gives the basics, but the github README, in about the middle, tells the full story.

Anyhow, --balance causes my hashes to be split between the two pools rather than just having pool 2 as a failover site.  I actually have some miner rigs balanced between 6 pools.  That can be useful, for instance, if you want to explore some newbie coins - just allocate one rig to hit multiple newbie pools.

There is a --load-balance option as well, which you can use with quotas.  For instance, you might give your for-profit pool a quota of 70, and three exploratory pools a quota of 10 yeah, and a friends pool a quota of 5 just so you show up there occasionally for support!  Alas, I'm seeing some performance problems, or at least reporting problems, when I use quotas, but in theory they would be great for multi pool usage.

Trust free enterprise to do what is best.  The good ones will do well, the not-so-good will eventually shutdown.

What do you have the miners plugged into?  Most cheap USB strips do not have sufficient power to run more than one or two.

Figure 500ma per U1 if your overclocking to 2GHz.

Unix shell file, but shy the ./ you should be able to put that in a .bat file - would save typing.

./bfgminer -S antminer:all --set-device antminer:clock=x0981 --balance -o stratum+tcp://us-east.multipool.us:8888 -O Cassey.miner2am:x -o stratum+tcp://multicoin.zapto.org:3320 -O Cassey.miner2am:x

That is all you should need for antminer U1s.  The above is my complete command line, with no config file.  Feel free to use it exactly (I'll appreciate the extra hash credits <smile>)

1st)  Make sure your running bfgminer 3.10.0 - prior releases don't have antminer support.
2nd) ./bfgminer -S antminer:all --set-device antminer:clock=x0981  Should do it for you, overclocking them up to about 2 GH, where most are reasonably stable without mods (make sure you have enough power)
3rd)  This is a bfgminer thread, don't expect much cgminer support <smile>

GPU --scrypt mining.  The only difference in the startup scripts is the balance option chosen.  But I appreciate the thought!

Hi all.  I seem to be having an issue when I use --load-balance with quotas vs --balance.  I've tested this on 4 different machines and am getting similar results...

Here are the key stats, using --load-balance between 5 pools, with quotas at 70, 10, 10, 10, & 5 with respect to each pool.

 OCL 0: 50.0C | 290.0/ 77.9/ 70.1kh/s | A: 52 R:0+0(none) HW:0/none
 OCL 1: 40.0C | 182.0/ 49.4/ 57.2kh/s | A: 40 R:0+0(none) HW:0/none
 OCL 2:       |  96.1/ 25.5/ 16.7kh/s | A: 12 R:0+0(none) HW:0/none

vs. starting with the same script that just changes --load-balance to --balance and all the "--quota xx;" to "-o"

 OCL 0: 77.0C | 437.4/408.7/327.0kh/s | A: 54 R: 6+0(none) HW:0/none
 OCL 1: 60.0C | 273.1/265.0/257.4kh/s | A: 41 R: 7+0(none) HW:0/none
 OCL 2:       | 140.6/140.0/107.5kh/s | A: 14 R: 3+0(none) HW:0/none

Obviously, this is only after 10 minutes or so of running, but the difference is clear.

Any thoughts?  Maybe its just a reporting difference?  Like its just reporting on the last pool or something?  Hope so, because I love the --load-balance/quote feature!

Cassey

Oh, I didn't run it there for long, just long enough to try a few restarts of bfgminer.  I do suspect its interfering somehow.  Antminer came back immediately when directly plugged in.

Does work good on measuring voltage and current of my usb fan! <rofl>

Just a bit of news... I purchased one of those $6 USB power monitors that were mentioned here (or maybe over on the BFGMiner Antminer thread).  Arrived today.  Works great for UCB fans, but when an Antminer is plugged in, bfgminer doesn't see it anymore.

Haven't played with it much, but wanted to give folks a heads up.  It was $6, for that I'm happy the lights work, so not complaining...

Could someone post a brief tutorial on the different multi-pool options in bfgminer?

In my case, I have (4) pools I want to divide my hash rate into.  Three of them are supporter of my coin, so to show them support, and do so evenly, I'm trying to pump work to each.  The Fourth is my "for profit" pool.

I figured "--balance" was the best, and although I normally see some hashes in each pool, per their "Dashboards" and "My Worker" screens, work seems to roll between them, almost in waves.  Not sure if that is bfgminer, or MPOS reporting though.

I'm clueless how "--quota" would work, but it sounds interesting.

I thought I might be able to do a "--pool-priority", setting them all the same, but not sure if that would accomplish anything.

"--rotate" appears like it would do the job, but would spike each pool in turn, and I'd prefer to give each an even load.

And there there are more complex thoughts, like running bfgminer in a loop and use "--shares" to force balancing (and/or favor my fourth pool a bit with say, half, of the shares)

Of the (5) mining machines I'm working with, they vary from single GPU to triple GPU, and I'd really like to have them setup pretty much the same.  e.g.  Not running multiple  instances of bfgminer and dedicating GPUs to pools.

Any thoughts?

Sounds to me like X is finally running!  You need a windows manager, I use BlackBox when I need to test a Linux-QT wallet, but most of the time I just SSH in.

Good news, if you have X up, temperatures should start reporting.

I tried running mixed GPU and CPU mining in a single instance of bfgminer (-S opencl:auto -S cpu:auto --scrypt).  It appeared to be working fine with all devices found.

After running for 24 hours or so, I noticed the CPUs had not accepted a single share all all of them (across 4 machines) showed dozens of HW errors.  Not a single one had accepted a share.

I found that strange and thought it might be worth reporting.  Should be easy to repeat if anyone has any ideas. 

FYI - just looking to do this because 1)  It should work and 2)  It would be nice to have one less process to monitor (e.g. minerd).  I recognized that all of my CPUs combined don't come up to a single GPU, but that is not the point.

Thoughts?

Presuming your running Linux, yeah, that should work from Root.  You may need to install a package to get lspci, but that is not as critical as the amdconfig (or aticonfig - they are the same).

OK - lets start with the basics:

"lspci | grep VGA"
"amdconfig --list-adapters"

The first should respond with a line per adapter and the second should respond with model type information per adapter.

If they don't, then the OS isn't configured quite right (and that happens a lot to me).

Hmmm, tried this, and getting:

 [2014-02-23 22:09:47] OCL 0: set_device is not implemented (trying to apply rule: OCL:gpu-fan=100)
 [2014-02-23 22:09:47] OCL 0: set_device is not implemented (trying to apply rule: OCL:temp-overheat=90)
 [2014-02-23 22:09:47] OCL 0: set_device is not implemented (trying to apply rule: OCL:temp-cutoff=100)
 [2014-02-23 22:09:47] OCL 0: set_device is not implemented (trying to apply rule: OCL:temp-target=80)

./bfgminer -S opencl:auto --scrypt

That is the key string.  Just make sure your executing against the one you just compiles, not the system one.  A ./bfgminer --help will make that obvious.  If you don't see the CPU and --scrypt options in the help, you running the wrong version.