[GUIDE] GridSeed GC3355 5 Chip Setup/power/windows/linux/rpi by UnicornHasher

[miaviator]

Which cgminer are you running on the Pi?

just thought i would share this with you all
i have 18 Gridseed miners running on a Ras-pi
using a 49 port Hub

Hi its 3.7.2 i got it from this thread
the Rpi is running Scripta Os

Well It's either the wheezy image I'm using or the pi won't do 10 instances.   Either way the pi just dies after launching 10 cgminer instances.

err why are you running 10 instances ??
and the image is not wheezy ? what ever that is

I'm running 10 because my cases and cords were delayed in shipping.  I need to run at least 32.

On API port per grid seed.  These are my hosted and leased units.

I'll just run 100 on a Linux PC until I can test on the next batch. 

Unless you wanna try running 10 instances

[pjcltd]

Which cgminer are you running on the Pi?

just thought i would share this with you all
i have 18 Gridseed miners running on a Ras-pi
using a 49 port Hub

Hi its 3.7.2 i got it from this thread
the Rpi is running Scripta Os

Well It's either the wheezy image I'm using or the pi won't do 10 instances.   Either way the pi just dies after launching 10 cgminer instances.

err why are you running 10 instances ??
and the image is not wheezy ? what ever that is

I'm running 10 because my cases and cords were delayed in shipping.  I need to run at least 32.

On API port per grid seed.  These are my hosted and leased units.

I'll just run 100 on a Linux PC until I can test on the next batch. 

Unless you wanna try running 10 instances

so you run 1 instance per gridseed miner?
why would you want to do that ?
or am i missing something ?

[rob1313]

Which cgminer are you running on the Pi?

just thought i would share this with you all
i have 18 Gridseed miners running on a Ras-pi
using a 49 port Hub

Hi its 3.7.2 i got it from this thread
the Rpi is running Scripta Os

Well It's either the wheezy image I'm using or the pi won't do 10 instances.   Either way the pi just dies after launching 10 cgminer instances.

err why are you running 10 instances ??
and the image is not wheezy ? what ever that is

I'm running 10 because my cases and cords were delayed in shipping.  I need to run at least 32.

On API port per grid seed.  These are my hosted and leased units.

I'll just run 100 on a Linux PC until I can test on the next batch. 

Unless you wanna try running 10 instances

so you run 1 instance per gridseed miner?
why would you want to do that ?
or am i missing something ?

yes because he lease the miners and host them for people

[suchmoon]

I'm running 10 because my cases and cords were delayed in shipping.  I need to run at least 32.

On API port per grid seed.  These are my hosted and leased units.

I'll just run 100 on a Linux PC until I can test on the next batch. 

Unless you wanna try running 10 instances

so you run 1 instance per gridseed miner?
why would you want to do that ?
or am i missing something ?

If I understand this correctly - he's hosting/renting/leasing these to customers and each needs to be able to set their own pools etc. Probably not possible to do without starting a separate instance for each.

I tried running 6 cgminers for a different reason (for benchmarking) and had all sorts of issues - miners not recognized, slow hashing, etc - but I assumed there was an issue with the USB hub. Now I'm thinking maybe not. Maybe multiple cgminer instances just don't play nice sharing USB. Runs fine with 2 currently.

[DPoS]

with bfgminer on windows, it is actually better to run them all on their own instance and point to their own worker in a pool.
You can always run a batch script to kick them all off and then restart the ones that trip out of the gate.

What we would need is a simple way to restart each miner from bfgminer to be able to run them all on one instance

[DPoS]

What about instead of using a Raspberry, would it not be an idea to run these on like the cheapest Intel NUC ?
those NUC like run something like 15watt

The NUCs don't come with anything..  I am rolling the dice with getting a zbox id81 plus

http://www.zotacusa.com/zbox-id81-plus.html

Found them for $199 so seems worth it instead of slapping an old power hungry pc together from spare parts


I'll report back once I get it and set up 20 gridseeds with it

[wolfey2014]

Hello,
can someone who knows what they are talking about, tell me / us what each line means in the cpuminer window running for each miner instance?

Please explain it to the rest of us who don't know?

Some of it is obvious like "Dispatching new work to GC3355 LTC core" but the rest isn't.

But the rest of it needs to be explained.
I'd love to fully understand and comprehend what I am reading as each set of lines passes.

The red lines...
The green lines...
The blue lines...
The white lines...
The yellow lines...

Please be very detailed, literal, concise and complete in your explanation.

Thanks!
Wolfey2014

[jamieb81]

What about instead of using a Raspberry, would it not be an idea to run these on like the cheapest Intel NUC ?
those NUC like run something like 15watt

The NUCs don't come with anything..  I am rolling the dice with getting a zbox id81 plus

http://www.zotacusa.com/zbox-id81-plus.html

Found them for $199 so seems worth it instead of slapping an old power hungry pc together from spare parts


I'll report back once I get it and set up 20 gridseeds with it

True the NUC doesn't come with anything except the processor, but trow in a 64gig ssd, some ram and you have decent something.

but anyway i've got a brand new laptop that I can put to use, with an i3 and 15" and it also just uses 15w and paid 280$ for it.

Just less practical than a box that you can put in a corner

[flounderella]

with bfgminer on windows, it is actually better to run them all on their own instance and point to their own worker in a pool.
You can always run a batch script to kick them all off and then restart the ones that trip out of the gate.

What we would need is a simple way to restart each miner from bfgminer to be able to run them all on one instance

Run in own instance: Is there an easy way to do this for 10x GS (batch line arg etc)? Right now if you fire up bfgminer, it finds all 10 and then goes off working.

Thanks

[flounderella]

Okay, thanks.
So the question still remains. Is it advantageous to do it that way or to have each miner reporting separately to the pool to there be added up and tallied?
I mean, is it more profitable to combine total hash rates etc. at the origin point (local) or at the receipt point? Or does it matter at either way?
Does one method make more money than the other?
That's the point of this discussion, after all Wink
Wolfey2014

it's all going into the same wallet eventually so I don;t think there's any advantage to having a worked for each GS instance - unless you like to see the difference in has rates and it can be dramatic - I have 20 works set up for my 20 GS's and that has rates can run from 260 up to 415 sometimes.

I noticed having one worker for all Gridseeds tends to raise the pool difficulty and then starts starving some GS, e.g. for DOGE. For lower difficulty coins/pools, they work just fine.

[wolfey2014]

Okay, thanks.
So the question still remains. Is it advantageous to do it that way or to have each miner reporting separately to the pool to there be added up and tallied?
I mean, is it more profitable to combine total hash rates etc. at the origin point (local) or at the receipt point? Or does it matter at either way?
Does one method make more money than the other?
That's the point of this discussion, after all Wink
Wolfey2014

it's all going into the same wallet eventually so I don;t think there's any advantage to having a worked for each GS instance - unless you like to see the difference in has rates and it can be dramatic - I have 20 works set up for my 20 GS's and that has rates can run from 260 up to 415 sometimes.

"I noticed having one worker for all Gridseeds tends to raise the pool difficulty and then starts starving some GS, e.g. for DOGE. For lower difficulty coins/pools, they work just fine."

Now, that's the kind of information I'm looking for! That is definately important to know!
Thanks!
Wolfey2014

[TrollboxChamp]

I have 80 workers for my 80 gridseeds.

this way i know if each unit is getting work or not. Too many on one worker and things get fucked

[DPoS]

with bfgminer on windows, it is actually better to run them all on their own instance and point to their own worker in a pool.
You can always run a batch script to kick them all off and then restart the ones that trip out of the gate.

What we would need is a simple way to restart each miner from bfgminer to be able to run them all on one instance

Run in own instance: Is there an easy way to do this for 10x GS (batch line arg etc)? Right now if you fire up bfgminer, it finds all 10 and then goes off working.

Thanks

on windows instead of :all you put \\.\COMn  where n is the COM# listed in your device manager (e.g COM48)

so you make a shortcut file with different startup info for each

[flounderella]

Ahh cool ... 10 windows but at least this way, can manage the difficulty better.

Easy to check but if you know already: do the COM ports assigned to the GS stay the same, even after hub restart. Wondering if I should stick a label on each GS, so in case of bumps, can just restart the affected GS instead of reseting all from the psu/hub

[flounderella]

So this is what I've done now:

Create a launch.bat with:

Code:

echo on
my_minions.bat <URL> <USERNAME> <PASS> 850
pause

And then create the helper batch file (my_minions.bat) with the correct COM ports from device manager. Assuming you are not too OCD and use simple user names of 1-10 and just a single password for all workers :p

Code:

start bfgminer.exe --scrypt -o %1 -u %2.1 -p %3 -d gridseed -S gridseed:\\.\COM1 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.2 -p %3 -d gridseed -S gridseed:\\.\COM2 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.3 -p %3 -d gridseed -S gridseed:\\.\COM3 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.4 -p %3 -d gridseed -S gridseed:\\.\COM4 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.5 -p %3 -d gridseed -S gridseed:\\.\COM5 --set-device gridseed:clock=%4

start bfgminer.exe --scrypt -o %1 -u %2.6 -p %3 -d gridseed -S gridseed:\\.\COM6 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.7 -p %3 -d gridseed -S gridseed:\\.\COM7 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.8 -p %3 -d gridseed -S gridseed:\\.\COM8 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.9 -p %3 -d gridseed -S gridseed:\\.\COM9 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.10 -p %3 -d gridseed -S gridseed:\\.\COM10 --set-device gridseed:clock=%4

It seems to be doing better. Previously with one worker, the pool was ramping up the difficulty very quickly to 2000+, but now I see each worker getting a diff of about 100-200. Will let it run for a while and see how that works out.


Since its harder to monitor 10 open windows, so now just going by pool worker stats to keep an eye. The good thing is at pools like scryptguild you can set per worker notifications when one has been idle for x minutes, so maybe easier to manage that way.

[wolfey2014]

So this is what I've done now:

Create a launch.bat with:

Code:

echo on
my_minions.bat <URL> <USERNAME> <PASS> 850
pause

And then create the helper batch file (my_minions.bat) with the correct COM ports from device manager. Assuming you are not too OCD and use simple user names of 1-10 and just a single password for all workers :p

Code:

start bfgminer.exe --scrypt -o %1 -u %2.1 -p %3 -d gridseed -S gridseed:\\.\COM1 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.2 -p %3 -d gridseed -S gridseed:\\.\COM2 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.3 -p %3 -d gridseed -S gridseed:\\.\COM3 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.4 -p %3 -d gridseed -S gridseed:\\.\COM4 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.5 -p %3 -d gridseed -S gridseed:\\.\COM5 --set-device gridseed:clock=%4

start bfgminer.exe --scrypt -o %1 -u %2.6 -p %3 -d gridseed -S gridseed:\\.\COM6 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.7 -p %3 -d gridseed -S gridseed:\\.\COM7 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.8 -p %3 -d gridseed -S gridseed:\\.\COM8 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.9 -p %3 -d gridseed -S gridseed:\\.\COM9 --set-device gridseed:clock=%4
start bfgminer.exe --scrypt -o %1 -u %2.10 -p %3 -d gridseed -S gridseed:\\.\COM10 --set-device gridseed:clock=%4

It seems to be doing better. Previously with one worker, the pool was ramping up the difficulty very quickly to 2000+, but now I see each worker getting a diff of about 100-200. Will let it run for a while and see how that works out.

Hey.
Cool experiment!
I'm very interested in the results. Would be great to have a way to get better odds of making a bit more profit from the work your miners do, eh?

What comm port drivers are you using? STM? Is there a link to where I can get the right ones?
What are the basic instructions for setting up bfgminer.
I'd be converting over from cpuminer power saver version. I'm running Scrypt only.
How fine are the overclock steps in bfgminer?
Thanks
Wolfey2014

[SVK]

Guys,
I'm absolute retard from Raspberry pi and all I can do with it is put SD card in and power it up.

I have done this

I configured a few things on first boot:
Expand the image to fill the card
Change password
Change startup method (Text only)

Now in PUTTY it is asking me for "LOG IN AS"  
What is login name - I didn't set any login name and "raspberry" or "root" doesn't work.

I have changed password to "0000" but I didn't set any login name.

Can somebody help me please ?
Thank you very much

[pjcltd]

Guys,
I'm absolute retard from Raspberry pi and all I can do with it is put SD card in and power it up.

I have done this

I configured a few things on first boot:
Expand the image to fill the card
Change password
Change startup method (Text only)

Now in PUTTY it is asking me for "LOG IN AS"  
What is login name - I didn't set any login name and "raspberry" or "root" doesn't work.

I have changed password to "0000" but I didn't set any login name.

Can somebody help me please ?
Thank you very much

Hi where are i can help I'm in the UK

[SVK]

Guys,
I'm absolute retard from Raspberry pi and all I can do with it is put SD card in and power it up.

I have done this

I configured a few things on first boot:
Expand the image to fill the card
Change password
Change startup method (Text only)

Now in PUTTY it is asking me for "LOG IN AS"  
What is login name - I didn't set any login name and "raspberry" or "root" doesn't work.

I have changed password to "0000" but I didn't set any login name.

Can somebody help me please ?
Thank you very much

Hi where are i can help I'm in the UK

UK

[pjcltd]

Guys,
I'm absolute retard from Raspberry pi and all I can do with it is put SD card in and power it up.

I have done this

I configured a few things on first boot:
Expand the image to fill the card
Change password
Change startup method (Text only)

Now in PUTTY it is asking me for "LOG IN AS"  
What is login name - I didn't set any login name and "raspberry" or "root" doesn't work.

I have changed password to "0000" but I didn't set any login name.

Can somebody help me please ?
Thank you very much

Hi where are i can help I'm in the UK

UK

ok pm me a number i can call you on
or we can use Skype
pjcltd telford