[Guide] Dogie's Comprehensive ASICMiner Blade Setup

[sidehack]

If you're trying to connect to a remote pool's getwork, you'll need their server address in the Pool addresses field, their getwork port in the Pool ports field, and an accurate Gateway so outgoing internet traffic knows where to start routing.

My initial reply was to pluMmet, using the mining_proxy command line. If you're using the mining proxy on a local machine, it might help to change the default getwork port (-gp option) and make sure to match the setting with the Pool ports field in your blade. If for some reason a local bitcoind is running on your machine, or some other software is using port 8332, it could hose up communications. That's a simple thing to rule out.

[1714152531]

1714152531

[1714152531]

1714152531

[1714152531]

1714152531

[Theo1355]

If you're trying to connect to a remote pool's getwork, you'll need their server address in the Pool addresses field, their getwork port in the Pool ports field, and an accurate Gateway so outgoing internet traffic knows where to start routing.

My initial reply was to pluMmet, using the mining_proxy command line. If you're using the mining proxy on a local machine, it might help to change the default getwork port (-gp option) and make sure to match the setting with the Pool ports field in your blade. If for some reason a local bitcoind is running on your machine, or some other software is using port 8332, it could hose up communications. That's a simple thing to rule out.

Capiche.  thx, I'll give it a try.

[Theo1355]

Problem solved!  Here's what I did:

Instead of using getwork, I used the gbt server and port on eligius.  For some reason I just can't get getwork to work any more.  Input these, delete the cookies from the previous config and restart.  I'll update with my hashrates tomorrow.  Currently (a few minutes after startup I'm doing 20.4 gh/s).

Thanks to all who contributed.  This successfully proves the hypothesis that n00bs can, indeed, learn  

[pluMmet]

Problem solved!  Here's what I did:

Instead of using getwork, I used the gbt server and port on eligius.  For some reason I just can't get getwork to work any more.  Input these, delete the cookies from the previous config and restart.  I'll update with my hashrates tomorrow.  Currently (a few minutes after startup I'm doing 20.4 gh/s).

Thanks to all who contributed.  This successfully proves the hypothesis that n00bs can, indeed, learn  

Could you be very specific please...

What did you type into which fields?

[pluMmet]

okay I got it thx

9337,9337
gbt.mining.eligius.st, gbt.mining.eligius.st

[Theo1355]

Problem solved!  Here's what I did:

Instead of using getwork, I used the gbt server and port on eligius.  For some reason I just can't get getwork to work any more.  Input these, delete the cookies from the previous config and restart.  I'll update with my hashrates tomorrow.  Currently (a few minutes after startup I'm doing 20.4 gh/s).

Thanks to all who contributed.  This successfully proves the hypothesis that n00bs can, indeed, learn  

Could you be very specific please...

What did you type into which fields?

Sure.

IPs: 192.168.1.200  (.201 for the second blade)
Mask: default
Gateway: default
WEB port: 8000
Primary and secondary DNS: default
Ports: 9337,9337
Server addresses: gbt.mining.eligius.st,gbt.mining.eligius.st
user:pass:1K*my address*Q23:pwd,1K*my address*Q23:pwd

Hope this helps!  I reset both of mine to fac default, but I don't think I need to do that.

[Theo1355]

okay I got it thx

9337,9337
gbt.mining.eligius.st, gbt.mining.eligius.st

Great!  I left the detailed instructions up for someone else...

[pluMmet]

okay I got it thx

9337,9337
gbt.mining.eligius.st, gbt.mining.eligius.st

Great!  I left the detailed instructions up for someone else...

Yep, all friends here

[rozey]

Hey guys, having some troubles.
I have 3 blades set up.  They're all pointed at my computer, which is running slush's stratum proxy.  No problems here.

Now, I want to divide them up among different pools.
Trying to connect one blade to BTCguild what I tried is to create a .bat with the instructions provided by Dogie.  Still had all the blades pointing to my PC which was running now two different stratums.  What ended up happening is that the blades went nuts and all of them stopped mining the moment I launched the second stratum .bat file (launched it while slush's one was active).  I ended up restarting all of my blades and configuring them again.

So what I'd like to know is, whats the best way for me to achieve my goal of having each blade mine in a different pool?

[sidehack]

You need the -gp option on the stratum proxy, so each one is listening on a different pool.

mining_proxy.exe -o pool1 -p port1 -gp 8332
mining_proxy.exe -o pool2 -p port2 -gp 8333

Configure your pool1 blades to Ports 8332,8332 and the pool2 blades to Ports 8333,8333

I had a setup exactly like that running for a day earlier this week, mining for a friend; two stratum proxies on the same machine pointed to different pools.

[rozey]

Thanks very much, it works great!  From your experience, is it better mining with different pools?  I thought it to be logical, having larger exposure- but I'd like to hear what others have to say about it .

[InfiniteGrim]

So I have had two blades running in my desktop PC.  They ran flawlessly. My 3 other blades, backplane, and PSU have finally come in. I have them all hooked up and I am running into an issue. Randomly 1-2 blades will "reset" to factory settings. They will be on Hashing for hours, but for ASIC test setup or whatever the server and worker is.

At first I thought it was my cheap ethernet switch, but that should NOT cause the blades to factory reset. I just switched the position in the slots of the backplane to see if that helps. The blades are not moving either, so I doubt they are getting "loose'.

Any ideas?  The power supply is the long thin one that is the "factory default" one I believe. I also have a box fan blowing on the blades 24/7 so they are not overheating.

Thanks!

[matt4054]

So I have had two blades running in my desktop PC.  They ran flawlessly. My 3 other blades, backplane, and PSU have finally come in. I have them all hooked up and I am running into an issue. Randomly 1-2 blades will "reset" to factory settings. They will be on Hashing for hours, but for ASIC test setup or whatever the server and worker is.

At first I thought it was my cheap ethernet switch, but that should NOT cause the blades to factory reset. I just switched the position in the slots of the backplane to see if that helps. The blades are not moving either, so I doubt they are getting "loose'.

Any ideas?  The power supply is the long thin one that is the "factory default" one I believe. I also have a box fan blowing on the blades 24/7 so they are not overheating.

Thanks!

If you have a multimeter, you should check the voltage of each power lane, they should be 1.06 or 1.2V depending on your clock needs.

Also, the random resets could be caused by a faulty control (Ethernet) board. Try exchanging the control board with a Blade that you know to be OK, and if the problems follows the board, you know where it's coming from.

[sidehack]

He's probably running V2 blades, which means "1.06 or 1.2V depending on your clock needs" is irrelevant - 1.05 to 1.06 volts *only* for the single clock speed. Also everything is built on a single board so there is no interchangable ethernet board.

[matt4054]

He's probably running V2 blades, which means "1.06 or 1.2V depending on your clock needs" is irrelevant - 1.05 to 1.06 volts *only* for the single clock speed. Also everything is built on a single board so there is no interchangable ethernet board.

True, if he's got V2 blades. It sucks. That's why I ordered V1 instead of V2's actually

[sidehack]

I like the V2s. They're fun to mess with.

[InfiniteGrim]

So I have had two blades running in my desktop PC.  They ran flawlessly. My 3 other blades, backplane, and PSU have finally come in. I have them all hooked up and I am running into an issue. Randomly 1-2 blades will "reset" to factory settings. They will be on Hashing for hours, but for ASIC test setup or whatever the server and worker is.

At first I thought it was my cheap ethernet switch, but that should NOT cause the blades to factory reset. I just switched the position in the slots of the backplane to see if that helps. The blades are not moving either, so I doubt they are getting "loose'.

Any ideas?  The power supply is the long thin one that is the "factory default" one I believe. I also have a box fan blowing on the blades 24/7 so they are not overheating.

Thanks!

If you have a multimeter, you should check the voltage of each power lane, they should be 1.06 or 1.2V depending on your clock needs.

Also, the random resets could be caused by a faulty control (Ethernet) board. Try exchanging the control board with a Blade that you know to be OK, and if the problems follows the board, you know where it's coming from.

These are V2 Blades though.

[user27]

I have done it for a few friends and now I'm looking at starting a 72 hour V2 blade overclocking service; what's the level of interest for this?

Thanks


edit: forgot to mention that the overclock I have been doing goes to about 15gh per blade; 16gh is doable but risky.

[InfiniteGrim]

What are your guarantees you will not screw up someone's blades?

Also I thought 13GH was the limit on these, or is that V1?

[sidehack]

I've found 15GH is hard for most people to keep cool, takes some heatsink improvements and really good fans. What's your price tag?
I overclock them to 14.5GH with next-business-day turnaround and return priority shipping included for 0.085BTC (done about 15 with zero casualties, but if I wreck it I replace it), and DIY kits for 13GH, 14.5GH and 15GH for 0.025BTC. Haven't worked up any V1 blades, but I have one coming in soon to test with so that'll be available soon. Also done general repair/maintenance for a few people.


InfiniteGrim - is it the same blades resetting, or different ones in the bunch?