BAMT - Easy persistent USB key based linux for dedicated miners/mining farms

[lodcrappo]

To all asking about wireless devices:  It's just not something I need, so it's not supported.  That isn't to say it won't work, but I don't have any wireless devices here so I won't be working on it. If you figure out how to do it, and there is something I can push out in a fix or add in future versions to make it work better, just let me know.

[1713586227]

1713586227

[1713586227]

1713586227

[1713586227]

1713586227

[1713586227]

1713586227

[jh1523]

Here's my solution to wireless: just attach the mining rig(s) to a cheapo wireless router running dd-wrt in repeater bridge mode. Especially advantageous if you run multiple rigs as you can connect them all to the same router.

[lagios]

i try to boot from a Kingston USB on a Gigabyte 990FXA-UD3 M/B but i only got the boot error.

Gigabyte won.

 

On the same machine i can boot Ubuntu Natty and Linuxcoin after i format the USB with the HP tool.

[lodcrappo]

i try to boot from a Kingston USB on a Gigabyte 990FXA-UD3 M/B but i only got the boot error.

Gigabyte won.

 

On the same machine i can boot Ubuntu Natty and Linuxcoin after i format the USB with the HP tool.

I will make BAMT work with any hardware you buy for me.  You can also donate BTC equal to the cost, let me know what hardware you want support for, and I will purchase the hardware myself and make it work for you.  There isn't much else I can do to help, since all my hardware works fine.

[jh1523]

I can't make Diapolo's latest kernels (https://bitcointalk.org/index.php?topic=25135.0) work with BAMT's mine script. The 8-4-11 and 8-11-11 kernels work if I start the phoenix miner manually after replacing the default kernel with the new ones, but if I use "/etc/init.d/mine start" the miners just don't do anything. Gpumon shows no activity and the pool doesn't receive any shares. No error messages either.

[tysat]

I can't make Diapolo's latest kernels (https://bitcointalk.org/index.php?topic=25135.0) work with BAMT's mine script. The 8-4-11 and 8-11-11 kernels work if I start the phoenix miner manually after replacing the default kernel with the new ones, but if I use "/etc/init.d/mine start" the miners just don't do anything. Gpumon shows no activity and the pool doesn't receive any shares. No error messages either.

What happens if you attach to a device in gpumon?  You should be able to see something then.

[mmartoccia]

i try to boot from a Kingston USB on a Gigabyte 990FXA-UD3 M/B but i only got the boot error.

Gigabyte won.

 

On the same machine i can boot Ubuntu Natty and Linuxcoin after i format the USB with the HP tool.

I too have a Gigabyte and having the same problem and I love BAMT.  I had to replace my original MB (EVGA) to work in more GPU's and now having the same (seems rampant with Gigabytes and Booting from USB) issue.

I've tried numerous workarounds: Using the HP format tool and then Win32diskInstaller.  Starting over and using unetbootin with Debian set and the .IMG as the .ISO and still no go.  In my research I came across this pagehttp://derfian.hamsterkollektivet.se/?p=319 which states:

The problem is that the BIOS Gigabyte uses simply doesn’t like partitions on USB sticks. A simple guide:

• Erase any existing partitions on the stick with gparted (although your partitioner of choice would likely work as well)
• Create vfat filesystem directly on USB stick
• Set up a Live image. I used unetbootin, livecd-creator would likely work as well.
  

Maybe there is a way to make BAMT a single partition over that of 3?  This may solve the Gigabyte issue for us Gigabyte users.  BTW, love love BAMT!
 

[lodcrappo]

Maybe there is a way to make BAMT a single partition over that of 3?  This may solve the Gigabyte issue for us Gigabyte users.  BTW, love love BAMT!
  

Unfortunately, the current bamt system does require multiple partitions on the usb key.  It is probably possible to do a single partition based system, but it would mean rearranging quite a bit of the underpinnings we use.

Based on the relatively pathetic amount of donations, I really don't think I'll be putting time into BAMT going forward, certainly not enough to justify the time it would take to redesign to a single partition.  It is really discouraging to see the huge number of downloads and hits on the fix server (which I provide at my own expense), while not a single donation in over a week.

I think there is potential in BAMT to be a great system, but I'm just not into working for free.  If someone else wants to carry on with it, please do.  I release all work I've done under the GPL v3.0.

[jh1523]

I can't make Diapolo's latest kernels (https://bitcointalk.org/index.php?topic=25135.0) work with BAMT's mine script. The 8-4-11 and 8-11-11 kernels work if I start the phoenix miner manually after replacing the default kernel with the new ones, but if I use "/etc/init.d/mine start" the miners just don't do anything. Gpumon shows no activity and the pool doesn't receive any shares. No error messages either.

What happens if you attach to a device in gpumon?  You should be able to see something then.

gpumon is acting weird for me even in normal situations. It doesn't display miners correctly even with the default kernel. Example:
http://img263.imageshack.us/img263/9836/gpumon1.png
Looks like miner #0 is not working, right? Well that's not true, miner #0 is in fact working fine and pumping shares. If I attach to it I can see it working, but gpumon for some reason thinks it doesn't. That's with the default kernel.

However with Diapolo's new kernels what I get is: mine restart apparently works and doesn't generate any error message; in gpumon both my miners are red; if I attach to either there is a screen that briefly flashes then reverts to the main gpumon screen. If I keep 0 and respectively 1 pressed so I see what's on that screen (I get a continuous flickering between gpumon's main screen and the miner screen) I can distinguish that there's a succession of messages "There is no screen to be resumed matching gpu0." (for 0, and similar with gpu1 for 1). Again with the same new kernels if I start phoenix.py from the command line directly it works fine.

[The--Captain]

I can't make Diapolo's latest kernels (https://bitcointalk.org/index.php?topic=25135.0) work with BAMT's mine script. The 8-4-11 and 8-11-11 kernels work if I start the phoenix miner manually after replacing the default kernel with the new ones, but if I use "/etc/init.d/mine start" the miners just don't do anything. Gpumon shows no activity and the pool doesn't receive any shares. No error messages either.

What happens if you attach to a device in gpumon?  You should be able to see something then.

gpumon is acting weird for me even in normal situations. It doesn't display miners correctly even with the default kernel. Example:
[IMAGE REDACTED]
Looks like miner #0 is not working, right?

Edit: No, it looks like gpumon didn't pull *any* stats at all from that card.  Shouldn't be too hard to poke around in the gpumon scripts and figure out why.

Well that's not true, miner #0 is in fact working fine and pumping shares. If I attach to it I can see it working, but gpumon for some reason thinks it doesn't. That's with the default kernel.

However with Diapolo's new kernels what I get is: mine restart apparently works and doesn't generate any error message; in gpumon both my miners are red; if I attach to either there is a screen that briefly flashes then reverts to the main gpumon screen. If I keep 0 and respectively 1 pressed so I see what's on that screen (I get a continuous flickering between gpumon's main screen and the miner screen) I can distinguish that there's a succession of messages "There is no screen to be resumed matching gpu0." (for 0, and similar with gpu1 for 1). Again with the same new kernels if I start phoenix.py from the command line directly it works fine.

Uncomment out the line that prints the phoenix cmdline in wrapper.pl, restart mining, then attach using 'screen -d -r gpu0' to see the cmdline that is being used to start phoenix.  Compare that with the command line you have found that works.  You might be forgetting to specify a WORKSIZE, which that kernel seems to need.  Now donate to lodcrappo.

Edit:  As far as those who are complaining about their broken hardware not being able to boot BAMT - you should be yelling at your hardware vendor for a firmware update if not a refund. lodcrappo has nothing to do with it.  You can't expect everyone to supply workarounds for your broken equipment (unless maybe you're Cisco).  If you bought a car that couldn't drive uphill, would you complain to the government for building roads that aren't completely flat, or would you return the car to the dealership?

[drgr33n]

Maybe there is a way to make BAMT a single partition over that of 3?  This may solve the Gigabyte issue for us Gigabyte users.  BTW, love love BAMT!
  

Unfortunately, the current bamt system does require multiple partitions on the usb key.  It is probably possible to do a single partition based system, but it would mean rearranging quite a bit of the underpinnings we use.

Based on the relatively pathetic amount of donations, I really don't think I'll be putting time into BAMT going forward, certainly not enough to justify the time it would take to redesign to a single partition.  It is really discouraging to see the huge number of downloads and hits on the fix server (which I provide at my own expense), while not a single donation in over a week.

I think there is potential in BAMT to be a great system, but I'm just not into working for free.  If someone else wants to carry on with it, please do.  I release all work I've done under the GPL v3.0.

+1

[jh1523]

Uncomment out the line that prints the phoenix cmdline in wrapper.pl, restart mining, then attach using 'screen -d -r gpu0' to see the cmdline that is being used to start phoenix.  Compare that with the command line you have found that works.  You might be forgetting to specify a WORKSIZE, which that kernel seems to need.  Now donate to lodcrappo.

I am specifying a WORKSIZE. And I have donated to lodcrappo the day after I started using BAMT, over 2 weeks ago.
After uncommenting the CMD printing and restarting mining, 'screen -d -r gpu0' returns 'There is no screen to be detached matching gpu0.' Similar for gpu1.
(if I use the same steps with the older kernel, I do get the command line printed before the normal miner output - but with the new kernel it looks it doesn't even get there).

[zx9r]

I have installed a new version of phatk kernel and I got a little improvement, from 389 to 394 Mhash per each 5850

May be a patch upgrade would be fine for this

[bitminers]

I have installed a new version of phatk kernel and I got a little improvement, from 389 to 394 Mhash per each 5850

May be a patch upgrade would be fine for this

Thats cool! how did you do that can you list here the steps? That would be really great!

[mmartoccia]

Maybe there is a way to make BAMT a single partition over that of 3?  This may solve the Gigabyte issue for us Gigabyte users.  BTW, love love BAMT!
  

Unfortunately, the current bamt system does require multiple partitions on the usb key.  It is probably possible to do a single partition based system, but it would mean rearranging quite a bit of the underpinnings we use.

Based on the relatively pathetic amount of donations, I really don't think I'll be putting time into BAMT going forward, certainly not enough to justify the time it would take to redesign to a single partition.  It is really discouraging to see the huge number of downloads and hits on the fix server (which I provide at my own expense), while not a single donation in over a week.

I think there is potential in BAMT to be a great system, but I'm just not into working for free.  If someone else wants to carry on with it, please do.  I release all work I've done under the GPL v3.0.

+1

I've got Linuxcoin running and I'm verifying the multiple partition issue with Gigabyte by implementing persistance.  If persistence works its not the partitions but more how the image is created ala unetbootin.  I'll report back.


EDIT: Multiple Partitions are in fact working.  The difference here is using the .IMG compared to unetbootin.   How difficult would it be to make BAMT work with unetbootin?  Can BAMT be made into an ISO?

    

[gnar1ta$]

I have 2 Gagabyte boards (which I'm planning on setting on fire this weekend) that boot BAMT and Linuxcoin.  I used dd command in linux. If you have a linux box run:

Code:

dd if=<BAMT>.img of=/dev/sd[x]

Or I'm sure you can use qemu to convert to ISO, not sure of the syntax though.

[hugolp]

I have 2 Gagabyte boards (which I'm planning on setting on fire this weekend) that boot BAMT and Linuxcoin.  I used dd command in linux. If you have a linux box run:

Code:

dd if=<BAMT>.img of=/dev/sd[x]

Or I'm sure you can use qemu to convert to ISO, not sure of the syntax though.

Apparently the problem is not in all Gygabyte boards but only in a few models. I can confirm the Gigabyte 990fx UD5 is affected.

[mmartoccia]

I have 2 Gagabyte boards (which I'm planning on setting on fire this weekend) that boot BAMT and Linuxcoin.  I used dd command in linux. If you have a linux box run:

Code:

dd if=<BAMT>.img of=/dev/sd[x]

Or I'm sure you can use qemu to convert to ISO, not sure of the syntax though.

Did BAMT work right out the box for you?  What MB's are you using and would you mind sharing the BIOS versions?  Maybe this could all go away with a BIOS downgrade.

EDIT I can do a similar action in terminal on OS X

Code:

dd if=bamt_v0.4b.img of=/dev/disk1 bs=8192

My question is, where does bamt_v0.4b.img need to reside to be seen? Sorry, Dee de dee.

I have 2 Gagabyte boards (which I'm planning on setting on fire this weekend) that boot BAMT and Linuxcoin.  I used dd command in linux. If you have a linux box run:

Code:

dd if=<BAMT>.img of=/dev/sd[x]

Or I'm sure you can use qemu to convert to ISO, not sure of the syntax though.

Apparently the problem is not in all Gygabyte boards but only in a few models. I can confirm the Gigabyte 990fx UD5 is affected.

I'm using the exact same MB.  

[gnar1ta$]

Did BAMT work right out the box for you?  What MB's are you using and would you mind sharing the BIOS versions?  Maybe this could all go away with a BIOS downgrade.

Sorry, I've switched MB's recently...those are being shipped to my brothers. I'll try to get specs for you once they get there.

[mmartoccia]

Rebuilding the USB via Terminal on OS X:

Code:

dd if=bamt_v0.4b.img of=/dev/disk1 bs=8192

Same result as doing it with Win32 Disk Imager:

Code:

Loading Operating System......
Boot Error

Need to determine what the boot differences are between Linuxcoin and that of BAMT.