Topic: BAMT version 0.5 - Easy USB based mining Linux with farm wide management tools - page 30. (Read 324185 times)

did you put a
in the cgminer command in the bamt conf?

Hints on upgrading to current version of cgminer here:
https://bitcointalksearch.org/topic/m.873655
https://bitcointalksearch.org/topic/m.972078

Take your pick, I used the first link and just had to update the sources to reflect current versions. It's been running about a month now without issue for me anyway. Here are some hints for the cgminer options to add in the BFL's:
https://bitcointalksearch.org/topic/m.989549

If this works for you, please remember to donate to lodcrappo to keep supporting BAMT.

even if you get the devices working in cgminer, they will not show up in any of the bamt tools, are not monitored or graphed or managed at all by bamt.  bamt does not support nongpu mining.

Some but not all.  It's an older version of cgminer, 2.3.1 I believe, not 100% sure.  So anything supported in that is good to go out of the box.  You can build an newer version of cgminer on BAMT if you have enough Linux knowledge.  Or, you can bug lodcrappo to update the version and push it out on the fixer, good luck on that.  

cgminer supports fpga.  bamt does not.

Doesnµt the included cgminer support fpga devices?

yes

nice one thanks!

BAMT can support all devices that cgminer can.

Is BAMT 0.5 only for GPU rigs?

There is a known problem with BAMT and some machines keeping a DHCP address.  Assigning a permanent lease on your router is still a dynamic address and won't solve the problem.

A fix is unlikely, since the problem is relegated to a few boards/configurations and it's hard to reproduce apparently.

Just assign a static IP to your machine and be done with it.  Worked for me.

I am having a lot of problems with the loss of network issue on my rigs.  I understand 1) it is not really a BAMT problem and more a Linux Network manager problem, 2) only happens on some hardware and 3) several fixes have been discussed on this thread.

I am a Linux newbie and so my learning curve seems slow.  Please be patient.  I have 4 nearly identical rigs all with MSI 890fxa-gd70’s, Saphire 5830’s and a Cooler Master 1200w.  I lost my network connection with all of them.  (I still have internet to my laptop).  I have a verizon router in my house and a switch box in my garage where the rigs are.  The address to one of my rigs was (before the disconnect) 192.168.1.23. 

Question 1.  I am standing in front of the rig with a monitor, keyboard and mouse, what can I do (what command is entered) to re-establish the connection?  How do I force a new lease?

Here is what I have tried.
1)   I took BitMinerN8’s suggestion to use an older NIC.  No luck I still lost connection.
2)   I entered “ifdown eth0” logged in as root.  I got the response “eth0” not configured.  Eth0 was the port that had been used.
3)   I shut down the rig and then restarted it.  This works occasionally (but very rare).

Question 2.  I tried the more permanent fix suggested in the bitcointrading.com link.  Here is what I did:
1)   I logged in as root
2)   Then “nano /etc/network/interfaces”
3)   The configuration file had:
auto loiface
lo inet loopback
4)   I added the following under the first two lines:  (I put a “1” in the 3rd position of the address, netmask, broadcast, and gateway.  I think this is right because I go through a router and switch.  Is this right?
iface eth0 inet static
address 192.168.1.23
netmask 255.255.255.0
network 192.168.1.0
broadcast 192.168.1.255
gateway 192.168.1.1
5)   Then: apt-get remove network-manager* AND /etc/init.d/networking restart
This did not fix the problem.  Then I “nano /etc/network/interfaces” and my changes were not there.  Instead I had  "auto lo / iface lo inet loopback", and "allow-hotplug eth0 / iface eth0 inet dhcp"

How do I know what address do I use?  I used 192.168.1.23

Please help.  There is nothing more frustrating than staring at a rig (or four) that is not mining.

Problem solved.  It seems that the problem was that the bios would ONLY allow Sapphire Trixx to overclock the cards.  Flashed the bios with a "bios overclock" and everything works fine now.  Apparently the difference in CCC 11.6 and 12.4 was the differential between the poclbm and phatk2 problem leading me to believe poclbm is better with CCC 12.4 while phatk2 is better with 11.6.

Everything is running 100% stable now with 6 5830's running at 1020Mhz/360Mhz/1.168V/80% fan.  Average 6x320MH/s=1.92GH/s per rig.  Not sure on power consumption (I left my Kill-a-watt in Missouri and I have free power this summer for my internship so I really don't care), but the fans on the power supplies are pumping some pretty warm air out.  Rig is headless on its' own 20amp breaker.

Cards are running at about 70C in an ambient of about 69F.

I'll be sure and donate some BTC to the BAMT guys for this awesome OS.

They have a warranty (all cards from Newegg). I will make sure I handle it for you. Not a problem.  If you want to go the 72BTC route it would be shipped to you with your name and you could handle, up to you.

Not sure then, I have not run into that, but I don't push them that far either. On one of my rigs I have (5) 5830s, but I only take them to 995/280 (core_speed_2: 995, leave 0 and 1 commented out) and just left all the core_voltage_x commented out. So maybe try backing down on the OC and see if the Hhash goes up on it.

Good luck.

  is empty

See if there are any files located in this directory: /live/image/BAMT/CONTROL/ACTIVE/

That dir should be empty, so if you see anything like a noOCGPUx in there, remove it, then restart mining.

Hey guys, I just set up a new rig with BAMT.  I'm running 6 5830's in it, all running stable @ ~=63C.  Problem is there are two different models of Sapphire 5830 Xtremes.  They look the exact same, but the components on the PCB are different and in different locations. 

I've had been using phatk2 with phoenix on my 5970 with great success.  phatk2 with phoenix on the one model of 5830 gets me a solid 320MH/s @ 1000:300:1.168V.  On the other model, same clocks/voltage, I get 250MH/s.

So here is the funny thing.  I switched the 250MH/s cards into my gaming computer and I get the same thing with phatk2, but if I use guiminer with poclbm, I get an easy 310MH/s....which is ass backwards if you ask me.

So I'm trying to run 3 cards on phatk2 and 3 on poclbm in bamt with no success.  The phatk2's start mining no problem, but the poclbm's don't start and the cpu peaks and stays peaked trying to start them. 

As far as software.  I installed bamt 0.5c.  Ran apt-get update (Don't do upgrade it screws EVERYTHING up).  And ran the fixer.  I tried installing pyopencl and it seemed to screw thing up too.

Hardware: sempron 140, msi 890fxa-gd70, 1gb kingston, 2x cooler master 750w supplies, 3 extenders are powered by molex.

Please help guys!  How do I get this poclbm working?  Thanks in advance.