[OS] nvOC easy-to-use Linux Nvidia Mining - page 312.

ijduncan

newbie

Activity: 21

Merit: 0

Quote from: fullzero on July 22, 2017, 12:28:09 PM

Quote from: WoolF911 on July 22, 2017, 10:54:17 AM

Quote from: ijduncan on July 22, 2017, 10:43:08 AM

I am seeing a new error i have not seen before.
There is no screen to be resumed matching miner.
In the terminal window right before the miner of choice starts. And the mining still does seem to start in the background.

Yes, same problem.

Quote from: newmz on July 22, 2017, 11:53:49 AM

Quote from: ijduncan on July 22, 2017, 10:43:08 AM

I am seeing a new error i have not seen before.

There is no screen to be resumed matching miner.

In the terminal window right before the miner of choice starts. And the mining still does seem to start in the background.

Could it have something to do with a high resolution screen being present?

I have tried another image and the same thing occurred.

go to the quake terminal window and type "screen -r miner" and you should see your miner running

Have you set:

Code:

LOCALorREMOTE="REMOTE"

if so then this is expected

as newmz indicated:

open a terminal and enter

Code:

screen -r miner

to attach the mining process

The code is LOCAL. I have not adjusted the default.

fullzero

legendary

Activity: 1260

Merit: 1009

Quote from: ijduncan on July 22, 2017, 12:47:55 PM

Quote from: fullzero on July 22, 2017, 12:24:31 PM

Quote from: ijduncan on July 22, 2017, 12:36:27 AM

How about the option to set auto shutdown when the power button is pressed? I often need to move rigs and want to shut them down without plugging in or remoting in.

I have been running the following:

gsettings set org.gnome.settings-daemon.plugins.power button-power 'shutdown'

cheers

If you hold down the power button for ( I believe it is 5 seconds it should shutdown) . Is this shutdown different other than it acting immediately?

My understanding, and I may be completely wrong. Is that holding the power button is the same as turning off the power.

Holding down the power button is the equivalent of entering

Code:

sudo poweroff

in a terminal.

This is why I was asking if this modification is somehow different.

fullzero

legendary

Activity: 1260

Merit: 1009

Quote from: ijduncan on July 22, 2017, 12:42:09 PM

Quote from: newmz on July 22, 2017, 11:53:49 AM

Quote from: ijduncan on July 22, 2017, 10:43:08 AM

I am seeing a new error i have not seen before.

There is no screen to be resumed matching miner.

In the terminal window right before the miner of choice starts. And the mining still does seem to start in the background.

Could it have something to do with a high resolution screen being present?

I have tried another image and the same thing occurred.

go to the quake terminal window and type "screen -r miner" and you should see your miner running

So it's by design then?

This is what the:

Code:

LOCALorREMOTE

switch is for.

If you want the screen to attach to the gnome-terminal automatically use:

Code:

LOCALorREMOTE="LOCAL"

If you are using SSH; it is easier to have the screen detached and ready to be attached when you SSH in; so this is what REMOTE is for.

ijduncan

newbie

Activity: 21

Merit: 0

Quote from: fullzero on July 22, 2017, 12:24:31 PM

Quote from: ijduncan on July 22, 2017, 12:36:27 AM

How about the option to set auto shutdown when the power button is pressed? I often need to move rigs and want to shut them down without plugging in or remoting in.

I have been running the following:

gsettings set org.gnome.settings-daemon.plugins.power button-power 'shutdown'

cheers

If you hold down the power button for ( I believe it is 5 seconds it should shutdown) . Is this shutdown different other than it acting immediately?

My understanding, and I may be completely wrong. Is that holding the power button is the same as turning off the power.

fullzero

legendary

Activity: 1260

Merit: 1009

Quote from: hatch789 on July 22, 2017, 12:17:00 PM

Hi Guys,

There appears to be a bug with the Maxximus007_AUTO_TEMPERATURE_CONTROL. I verified in the nvoc0018 version and also in the updated patch to nv0018. I wanted to make sure the bug was still in the patched version before saying anything but now I have verified it is still there.

SO the bug is that whatever your individual power limit (MAIN ONE) is set to, is used only right in the beginning when the rig starts up. After the Maxximus007_AUTO_TEMPERATURE_CONTROL takes over it starts using whatever power limits are set in the INDIVIDUAL section.

The only way to fix this right now is set each individual power limit to whatever you want.

Obviously you have to restart 1bash for the new changes to take effect or reboot the rig.

With the big fix 1bash and files you should not need to: restart 1bash for the new changes to take effect or reboot the rig.

It is the case that when using Maxximus007_AUTO_TEMPERATURE_CONTROL:

Code:

POWERLIMIT="YES"     

POWERLIMIT_WATTS=75

only are initially applied

and then autotemp uses the powerlimits from:

Quote

# Note you MUST set individual powerlimits if using Maxximus007_AUTO_TEMPERATURE_CONTROL

INDIVIDUAL_POWERLIMIT_0=75

INDIVIDUAL_POWERLIMIT_1=75

INDIVIDUAL_POWERLIMIT_2=75

INDIVIDUAL_POWERLIMIT_3=75

INDIVIDUAL_POWERLIMIT_4=75

INDIVIDUAL_POWERLIMIT_5=75

INDIVIDUAL_POWERLIMIT_6=75

INDIVIDUAL_POWERLIMIT_7=75

INDIVIDUAL_POWERLIMIT_8=75

INDIVIDUAL_POWERLIMIT_9=75

INDIVIDUAL_POWERLIMIT_10=75

INDIVIDUAL_POWERLIMIT_11=75

INDIVIDUAL_POWERLIMIT_12=75

INDIVIDUAL_POWERLIMIT_13=75

I realize this is overly complicated; I could change this in the next version.

Right now I am leaning towards an individual GPU settings 1bash and a general GPU settings 1bash for the next version; which would both be in a folder in the windows partition. Then members can use the the general unless they have a mixed rig.

ijduncan

newbie

Activity: 21

Merit: 0

Quote from: newmz on July 22, 2017, 11:53:49 AM

Quote from: ijduncan on July 22, 2017, 10:43:08 AM

I am seeing a new error i have not seen before.

There is no screen to be resumed matching miner.

In the terminal window right before the miner of choice starts. And the mining still does seem to start in the background.

Could it have something to do with a high resolution screen being present?

I have tried another image and the same thing occurred.

go to the quake terminal window and type "screen -r miner" and you should see your miner running

So it's by design then?

fullzero

legendary

Activity: 1260

Merit: 1009

WarwickNZ

newbie

Activity: 21

Merit: 0

Would it be possible/easy to please add/explain how I can add a Genoil switch for EXP? It seems to default to claymore...

Thanks!

fullzero

legendary

Activity: 1260

Merit: 1009

Quote from: newmz on July 22, 2017, 11:51:57 AM

Hello, is anyone successfully using the Asrock Z87-Killer Socket 1150 motherboard with nvOC?

It has 7 PCIe sockets so I bought one hoping I could expand my 5 GPU rig to a 7 (Nvidia) GPU rig and get rid of some of my power hungry AMD R9 290s.

Any reason why it wont work?

Most likely you need the system files for the z87 chipset. I don't have one; so the image doesn't have those files.

fullzero

legendary

Activity: 1260

Merit: 1009

Quote from: WoolF911 on July 22, 2017, 10:54:17 AM

Quote from: ijduncan on July 22, 2017, 10:43:08 AM

I am seeing a new error i have not seen before.
There is no screen to be resumed matching miner.
In the terminal window right before the miner of choice starts. And the mining still does seem to start in the background.

Yes, same problem.

Quote from: newmz on July 22, 2017, 11:53:49 AM

Quote from: ijduncan on July 22, 2017, 10:43:08 AM

I am seeing a new error i have not seen before.

There is no screen to be resumed matching miner.

In the terminal window right before the miner of choice starts. And the mining still does seem to start in the background.

Could it have something to do with a high resolution screen being present?

I have tried another image and the same thing occurred.

go to the quake terminal window and type "screen -r miner" and you should see your miner running

Have you set:

Code:

LOCALorREMOTE="REMOTE"

if so then this is expected

as newmz indicated:

open a terminal and enter

Code:

screen -r miner

to attach the mining process

fullzero

legendary

Activity: 1260

Merit: 1009

Quote from: ijduncan on July 22, 2017, 12:36:27 AM

How about the option to set auto shutdown when the power button is pressed? I often need to move rigs and want to shut them down without plugging in or remoting in.

I have been running the following:

gsettings set org.gnome.settings-daemon.plugins.power button-power 'shutdown'

cheers

If you hold down the power button for ( I believe it is 5 seconds it should shutdown) . Is this shutdown different other than it acting immediately?

fullzero

legendary

Activity: 1260

Merit: 1009

Quote from: engineerking on July 22, 2017, 12:26:02 AM

I am experiencing the exact same problems almost verbatim as the poster above. I was thinking it is related to a bad flash drive possibly. I am trying a different one. Also version 14 and 15 seem to work better than 17 and 18. I am getting lag and freezes and random reboots just trying to use you ubuntu with no mining software running.

Let me know if you have the same problems with a different USB key or SSD.

fullzero

legendary

Activity: 1260

Merit: 1009

Quote from: WoolF911 on July 22, 2017, 12:09:22 AM

Quote from: fullzero on July 21, 2017, 09:44:35 PM

I would change

Code:

ZEC_PORT="3337"

to

Code:

ZEC_PORT="3333"

But your ZEC config looks like it should work with DwarfPool.

It is also worth trying removing the workername:

Code:

ZEC_WORKER="nv$IP_AS_WORKER"

to

Code:

ZEC_WORKER=""

With:

Also I have the error "Utilization is too low: restart 1bash" and multiple "Maxximus007_AUTO_TEMPERATURE_CONTROL" tabs. What it's mean? Maybe this is a reason of problem?

Due to some changes I made in the last 1bash; currently when in LOCAL mode: each restart of 1bash will kill the existing Maxximus007_AUTO_TEMPERATURE_CONTROL process, but not kill the tab containing it; then a new Maxximus007_AUTO_TEMPERATURE_CONTROL tab will be created with a new process inside.

The problem is likely that you are not connecting to the pool for some reason; when this happens the watchdog will detect low utilization and restart 1bash.

You can see if you can reach the pool by opening a new guake tab and entering the following cmd:

Code:

ping zec-ru1.dwarfpool.com

I would try to use a different pool and see if you can connect. For now while troubleshooting; I would set :

Code:

IAmNotAJeep_and_Maxximus007_WATCHDOG="NO"

fullzero, thank you for your reply. I'm sure that is no problem with connection to the pool, also ping is ok. I think it's a some problem with powerlimit, because when I disabled watchdog and trying to change some parameters, include powerlimit, my hashrate is growing to 200 Sol/s, also when I changed some parameters I seen my hashrate grow to 400 Sol/s, but I don't remember parameters what I changed. So at this moment I have hashrate 200 Sol/s with powerlimit 200, when I trying to change it to 250 I have lower hashrate. Cc 150, mc 900 as you wrote in the first message here. Also console showed power consumption lower than I set powerlimit. Any idea?

If you are using the Maxximus007_AUTO_TEMPERATURE_CONTROL ensure you are also changing the individual powerlimits for each GPU:

Code:

INDIVIDUAL_POWERLIMIT_0=75

INDIVIDUAL_POWERLIMIT_1=75

INDIVIDUAL_POWERLIMIT_2=75

for a 1080ti I recommend using between a 200 and 225 powerlimit.

If you have already done this:

It is also possible there are some bad sectors on your USB key and they are causing problems. I have had 5x USB keys go bad so far; although I do reimage a lot with test rigs.

It might be a good idea to try using another USB key or SSD.

hatch789

member

Activity: 66

Merit: 10

Hi Guys,

There appears to be a bug with the Maxximus007_AUTO_TEMPERATURE_CONTROL. I verified in the nvoc0018 version and also in the updated patch to nv0018. I wanted to make sure the bug was still in the patched version before saying anything but now I have verified it is still there.

SO the bug is that whatever your individual power limit (MAIN ONE) is set to, is used only right in the beginning when the rig starts up. After the Maxximus007_AUTO_TEMPERATURE_CONTROL takes over it starts using whatever power limits are set in the INDIVIDUAL section.

The only way to fix this right now is set each individual power limit to whatever you want.

Obviously you have to restart 1bash for the new changes to take effect or reboot the rig.

newmz

sr. member

Activity: 372

Merit: 250

The road of excess leads to the palace of wisdom

Quote from: fullzero on July 21, 2017, 05:01:26 PM

Quote from: neot76 on July 21, 2017, 01:21:23 PM

Why did my hash rate mining ETH on Claymore 9.7 go from 23.5 on Windows 7 down to 18.9 on nvOC? I have the power set to 94 and the mc set to 825. Is the latest driver that bad for 1060 cards? I am pretty sure I have the latest driver on Windows.

The OC curve for Linux is different than Windows. You need to set a higher OC to get the same results in linux. Also in Windows power limit is in % TDP; in linux it is in watts.

Try a mc of 1000, then keep bumping it up until you see a soft crash.

If your 1060s have the standard TDP of 125: to set an equalivent power limit you need to use 0.87 * 125 (or whatever the exact TDP of your model is) = 109 watts

I have 2 1070s and 3 1060s. When mining eth, I set core overclock to -200 and memory overclock to 1300. I get approx 135mh/s whereas stock it's somewhere in the region of 95mh/s.

newmz

sr. member

Activity: 372

Merit: 250

The road of excess leads to the palace of wisdom

Quote from: ijduncan on July 22, 2017, 10:43:08 AM

I am seeing a new error i have not seen before.

There is no screen to be resumed matching miner.

In the terminal window right before the miner of choice starts. And the mining still does seem to start in the background.

Could it have something to do with a high resolution screen being present?

I have tried another image and the same thing occurred.

go to the quake terminal window and type "screen -r miner" and you should see your miner running

newmz

sr. member

Activity: 372

Merit: 250

The road of excess leads to the palace of wisdom

Hello, is anyone successfully using the Asrock Z87-Killer Socket 1150 motherboard with nvOC?

It has 7 PCIe sockets so I bought one hoping I could expand my 5 GPU rig to a 7 (Nvidia) GPU rig and get rid of some of my power hungry AMD R9 290s.

Any reason why it wont work?

newmz

sr. member

Activity: 372

Merit: 250

The road of excess leads to the palace of wisdom

WoolF911

newbie

Activity: 9

Merit: 0

Quote from: ijduncan on July 22, 2017, 10:43:08 AM

I am seeing a new error i have not seen before.
There is no screen to be resumed matching miner.
In the terminal window right before the miner of choice starts. And the mining still does seem to start in the background.

Yes, same problem.

ijduncan

newbie

Activity: 21

Merit: 0

I am seeing a new error i have not seen before.

There is no screen to be resumed matching miner.

In the terminal window right before the miner of choice starts. And the mining still does seem to start in the background.

Could it have something to do with a high resolution screen being present?

I have tried another image and the same thing occurred.

Topic: [OS] nvOC easy-to-use Linux Nvidia Mining - page 312. (Read 418257 times)