[ mining os ] nvoc - page 214. | Bitcointalksearch.org

leenoox

full member

Activity: 200

Merit: 101

Quote from: Zucca on September 25, 2017, 07:24:15 PM

I update the versión to 0019 now It doesnt start mining. It doesnt hangs, but It repetively tries to start mining but It reaches until Launching: Mining process.

I can manually start claymore and It works

Update to 0019-1.1 instructions were posted few pages back. You will need to re-edit your 1bash file again after the update.

leenoox

full member

Activity: 200

Merit: 101

Quote from: CryptAtomeTrader44 on September 25, 2017, 03:24:02 PM

Quote from: leenoox on September 25, 2017, 01:43:57 PM

Hey CryptAtomeTrader44,
I had some code that I was testing, however after updating to 19-1, the update has overwritten the watchdog file with all the changes I had in it and I had no backup of the modified wd Sad

I have to start over now, hopefully I will have working beta version by the end of the week. I will post once I am done.

Arf, the shoemaker's syndrome always badly shod! The computer technician (engineer?) Who does not back up his script before updating! hi hi ha ha oh oh! I know that too! I do not stop tinkering the Salfter files to have a little more details at my convenience on the MPH-Switch, but I do not know where the original files at all. But anyway, my changes are very minimal because I do not understand much about Python. Even some Bash script goes beyond me then ... For example I can not reproduce this manual command of nvidia-settings that would allow me to reduce a little fan speed when they make too much noise annoy me.

Post us your script with an explanatory comment on the changes made. At least it will be here and therefore saved in some sorts.

yup... lesson learned...

i will post once i have something modded, right now i am at square one.

here's the command to change fan speed on a single GPU:

Code:

nvidia-settings -a [gpu:5]/GPUFanControlState=1 -a [fan:5]/GPUTargetFanSpeed=70

This will set fan on GPU5 to 70%. Change as you see fit, however note that if you have Maxximus007_auto_temp control enabled, it will readjust it back on its regular cycle.

Zucca

newbie

Activity: 81

Merit: 0

I update the versión to 0019 now It doesnt start mining. It doesnt hangs, but It repetively tries to start mining but It reaches until Launching: Mining process.

I can manually start claymore and It works

leenoox

full member

Activity: 200

Merit: 101

Quote from: jmayniac on September 25, 2017, 06:02:22 PM

Quote from: Bigf00t on September 25, 2017, 04:05:18 PM

Quote from: jmayniac on September 25, 2017, 03:35:33 PM

the Equihash rates are less than with Windows 10, even with the same pool and OC settings.

In Linux you need to double your overclock numbers.

So for example +500 in Afterburner will be 1000 in Linux.

Got it adjusted, thank you, but man are my cards burning up. In Windows I can usually maintain most of the cards around 70C, but under nvOC they are up to 80C. Is there something Afterburner is doing to cool down the cards? Maybe variable fan speed. I turned of "Manual Fan Speed" in 1bash, but that didn't make a difference.

In linux core clock is same as in windows but for memory you have to double the number. So keep OC on core same as under windows but double the OC on memory only. Also note, power limit under windows is in percentage while under linux is set in watts. Adjust the power limit accordingly.

Regarding the temperature, enable maximus_auto_temp control and set your preffered temperature and it will auto adjust the fans to keep up your desired temp.

jmayniac

newbie

Activity: 77

Merit: 0

Quote from: Bigf00t on September 25, 2017, 04:05:18 PM

Quote from: jmayniac on September 25, 2017, 03:35:33 PM

the Equihash rates are less than with Windows 10, even with the same pool and OC settings.

In Linux you need to double your overclock numbers.

So for example +500 in Afterburner will be 1000 in Linux.

Got it adjusted, thank you, but man are my cards burning up. In Windows I can usually maintain most of the cards around 70C, but under nvOC they are up to 80C. Is there something Afterburner is doing to cool down the cards? Maybe variable fan speed. I turned of "Manual Fan Speed" in 1bash, but that didn't make a difference.

Bigf00t

newbie

Activity: 60

Merit: 0

Quote from: jmayniac on September 25, 2017, 03:35:33 PM

the Equihash rates are less than with Windows 10, even with the same pool and OC settings.

In Linux you need to double your overclock numbers.

So for example +500 in Afterburner will be 1000 in Linux.

jmayniac

newbie

Activity: 77

Merit: 0

I'm a little confused by all the settings in '1bash'. I had, what I thought, was a good configuration, but when I boot into nvOC, the Equihash rates are less than with Windows 10, even with the same pool and OC settings.

Windows 10 - 2498 Sol/s
nvOC Linux - 2300 Sol/s

Basically all I changed was

1bash: https://pastebin.com/LDrWsgep

Code:

Coin="ZEC"
LOCALorREMOTE="LOCAL"
CUSTOM_WORKERNAME="omega"
TARGET_TEMP=85

ZEC_ADDRESS="t1Vbjn37Z76ARbg2NUiVsa1xkP4CLeUzBU2"
ZEC_POOL="us1-zcash.flypool.org"
ZEC_PORT="3333"

INDIVIDUAL_CLOCKS="YES" # YES NO

# Set individual clocks here if INDIVIDUAL_CLOCKS="YES"
__CORE_OVERCLOCK_0=100
MEMORY_OVERCLOCK_0=200 # 1060 3GB

__CORE_OVERCLOCK_1=100
MEMORY_OVERCLOCK_1=600 # 1070

__CORE_OVERCLOCK_2=100
MEMORY_OVERCLOCK_2=200 # 1060 6GB

__CORE_OVERCLOCK_3=100
MEMORY_OVERCLOCK_3=200 # 1080

__CORE_OVERCLOCK_4=100
MEMORY_OVERCLOCK_4=200 # 1080

__CORE_OVERCLOCK_5=100
MEMORY_OVERCLOCK_5=600 # 1070

CryptAtomeTrader44

full member

Activity: 340

Merit: 103

It is easier to break an atom than partialities AE

Quote from: leenoox on September 25, 2017, 01:43:57 PM

Hey CryptAtomeTrader44,
I had some code that I was testing, however after updating to 19-1, the update has overwritten the watchdog file with all the changes I had in it and I had no backup of the modified wd Sad

I have to start over now, hopefully I will have working beta version by the end of the week. I will post once I am done.

Arf, the shoemaker's syndrome always badly shod! The computer technician (engineer?) Who does not back up his script before updating! hi hi ha ha oh oh! I know that too! I do not stop tinkering the Salfter files to have a little more details at my convenience on the MPH-Switch, but I do not know where the original files at all. But anyway, my changes are very minimal because I do not understand much about Python. Even some Bash script goes beyond me then ... For example I can not reproduce this manual command of nvidia-settings that would allow me to reduce a little fan speed when they make too much noise annoy me.

Post us your script with an explanatory comment on the changes made. At least it will be here and therefore saved in some sorts.

leenoox

full member

Activity: 200

Merit: 101

Quote from: CryptAtomeTrader44 on September 22, 2017, 02:07:30 PM

Quote from: leenoox on September 17, 2017, 10:42:45 AM

Quote from: lards on September 16, 2017, 04:47:34 PM

Hi guys,

I have recently built a new rig with 13 P106-100 ASUS cards on nv0019 running -200cc 1550mc and PL90. I have a weird problem that after some time the hashrate on all cards drop by 50% or more and doesn't go back up, the miner doesn't restart and it just keeps running with lower hashrate. This time it did this after 7 hours , with 7 cards running it was up for over 12hours and no problems. Has anyone experienced this before and maybe you know where the problem is?

I am obviously running headless and using SSH to monitor the rig and adjust the settings.

PS. The miner or the OS doesn't restart the miner or reboot the system, just the hashrate drops pretty badly from 328 to 140 MH/s.

Thanks in advance to everyone who can help and also thanks to fullzero for the new version!

I've had similar issue, when one or two cards would semi-freeze, bringing the whole rig to work extremely slow. I added whole bunch of debug options in the watchdog to figure out what was going on... it turns out watcdog enters into a loop for counting down errors before trying to restart the mining process and it takes about 5-10 min to change the count by -1 and it won't catch the low utilization to restart the rig until few hours later.

The mining doesn't stop but continues at reduced hash speed, some cards drop from 25 to 13 MH/s, some down to 0.

To make it worst, during this semi-hang the driver changes p-state due to low/high utilization so it ends up with some cards at p2, some at p8 and some at p0 which screws the OC and makes more cards to hang.

I have some ideas to modify or even rewrite the watchdog to address these issues and will post it once finished. I am quite busy atm to work on it but I can post logs from the added debug so that devs can look into it.

Anyway, it turns out all of this was caused by two cards (same brand/model/memory) that couldn't handle the overclocking as the rest of the cards. By reducing the memory overclock on these two from 1820 to 1700 it stabilized the rig, no more hangups.

For now, reduce your overclock and give it a try, see if it fixes the problem. Please post back with results.

Have you managed to improve your watchdog as you said in this previous post?

Hey CryptAtomeTrader44,
I had some code that I was testing, however after updating to 19-1, the update has overwritten the watchdog file with all the changes I had in it and I had no backup of the modified wd Sad

I have to start over now, hopefully I will have working beta version by the end of the week. I will post once I am done.

mikespax

full member

Activity: 362

Merit: 102

please disregard my earlier comments about not having temperature controls after updating to beta. Just had to slowly put my rig back together and it seems to be running stable on 13 cards.

Does anyone know how to mix amd and nvidia on linux?

Zucca

newbie

Activity: 81

Merit: 0

Quote from: papampi on September 25, 2017, 10:01:50 AM

Quote from: Zucca on September 25, 2017, 08:48:41 AM

Quote from: papampi on September 25, 2017, 08:09:51 AM

Quote from: Zucca on September 25, 2017, 07:48:55 AM

Hello! I have a MSI Z270-A pro. I can run nvOS by default with 4 gtx 1060 (I plug 6 but it only detects 4) and it works perfectly. If I enable above 4g memory, it detects all the 6 gpus and starts mining well. But the OS hangs after 5 minutes, every time I try.
I disabled HQ audio as well.

What can be wrong?

Whats your CPU?
Did you enable cpu mining ?
whats your nvOC version?

THank you for your reply,
I have a Celeron G3930, and I didn't enable CPU mining, I am trying with ETH + DCR. And it is currently running on nvOc v0018.
The hang problem started when I enabled Above 4g memory, before that it worked perfectly with 4GPUs

Is the bios updated to latest version?
And update to nv0019, nv0018 had swap problem which could cause some lags and long boot times.
You may also try changing peg0, peg1 values

I don't know what peg0 and peg1 values are

Zucca

newbie

Activity: 81

Merit: 0

Quote from: leenoox on September 25, 2017, 12:41:09 PM

Quote from: Zucca on September 25, 2017, 08:48:41 AM

Quote from: papampi on September 25, 2017, 08:09:51 AM

Quote from: Zucca on September 25, 2017, 07:48:55 AM

Hello! I have a MSI Z270-A pro. I can run nvOS by default with 4 gtx 1060 (I plug 6 but it only detects 4) and it works perfectly. If I enable above 4g memory, it detects all the 6 gpus and starts mining well. But the OS hangs after 5 minutes, every time I try.
I disabled HQ audio as well.

What can be wrong?

Whats your CPU?
Did you enable cpu mining ?
whats your nvOC version?

THank you for your reply,
I have a Celeron G3930, and I didn't enable CPU mining, I am trying with ETH + DCR. And it is currently running on nvOc v0018.
The hang problem started when I enabled Above 4g memory, before that it worked perfectly with 4GPUs

As suggested above, update bios to the latest version.
You have to enable Above 4g in the bios.
If the system is running for about 5 minutes with all 6 cards then hangs, you probably overclocked them too much, try lowering the OC.
Are all 6 cards same model/version and have same memory manufacturer (eg. Samsung, Micron)? If not, execute this in terminal if you haven't already:

Code:

sudo echo 'export CUDA_DEVICE_ORDER=PCI_BUS_ID' | sudo tee -a /etc/profile.d/cuda-device-order.sh

and set individial OC for each card, try with low values and increase slowly until your system remains stable

There are 4 identical Zotac and 2 different EVGA video cards. I tried with no OC (100 MC and 100 CC) and it didn't work. I will try later to update version of BIOS and nvOc.
Oh and it also hangs with only one video card connected and of course above 4g memory enabled

leenoox

full member

Activity: 200

Merit: 101

Quote from: Zucca on September 25, 2017, 08:48:41 AM

Quote from: papampi on September 25, 2017, 08:09:51 AM

Quote from: Zucca on September 25, 2017, 07:48:55 AM

Hello! I have a MSI Z270-A pro. I can run nvOS by default with 4 gtx 1060 (I plug 6 but it only detects 4) and it works perfectly. If I enable above 4g memory, it detects all the 6 gpus and starts mining well. But the OS hangs after 5 minutes, every time I try.
I disabled HQ audio as well.

What can be wrong?

Whats your CPU?
Did you enable cpu mining ?
whats your nvOC version?

THank you for your reply,
I have a Celeron G3930, and I didn't enable CPU mining, I am trying with ETH + DCR. And it is currently running on nvOc v0018.
The hang problem started when I enabled Above 4g memory, before that it worked perfectly with 4GPUs

As suggested above, update bios to the latest version.
You have to enable Above 4g in the bios.
If the system is running for about 5 minutes with all 6 cards then hangs, you probably overclocked them too much, try lowering the OC.
Are all 6 cards same model/version and have same memory manufacturer (eg. Samsung, Micron)? If not, execute this in terminal if you haven't already:

Code:

sudo echo 'export CUDA_DEVICE_ORDER=PCI_BUS_ID' | sudo tee -a /etc/profile.d/cuda-device-order.sh

and set individial OC for each card, try with low values and increase slowly until your system remains stable

papampi

full member

Activity: 686

Merit: 140

Linux FOREVER! Resistance is futile!!!

Quote from: Zucca on September 25, 2017, 08:48:41 AM

Quote from: papampi on September 25, 2017, 08:09:51 AM

Quote from: Zucca on September 25, 2017, 07:48:55 AM

Hello! I have a MSI Z270-A pro. I can run nvOS by default with 4 gtx 1060 (I plug 6 but it only detects 4) and it works perfectly. If I enable above 4g memory, it detects all the 6 gpus and starts mining well. But the OS hangs after 5 minutes, every time I try.
I disabled HQ audio as well.

What can be wrong?

Whats your CPU?
Did you enable cpu mining ?
whats your nvOC version?

THank you for your reply,
I have a Celeron G3930, and I didn't enable CPU mining, I am trying with ETH + DCR. And it is currently running on nvOc v0018.
The hang problem started when I enabled Above 4g memory, before that it worked perfectly with 4GPUs

Is the bios updated to latest version?
And update to nv0019, nv0018 had swap problem which could cause some lags and long boot times.
You may also try changing peg0, peg1 values

Zucca

newbie

Activity: 81

Merit: 0

Quote from: papampi on September 25, 2017, 08:09:51 AM

Quote from: Zucca on September 25, 2017, 07:48:55 AM

Hello! I have a MSI Z270-A pro. I can run nvOS by default with 4 gtx 1060 (I plug 6 but it only detects 4) and it works perfectly. If I enable above 4g memory, it detects all the 6 gpus and starts mining well. But the OS hangs after 5 minutes, every time I try.
I disabled HQ audio as well.

What can be wrong?

Whats your CPU?
Did you enable cpu mining ?
whats your nvOC version?

THank you for your reply,
I have a Celeron G3930, and I didn't enable CPU mining, I am trying with ETH + DCR. And it is currently running on nvOc v0018.
The hang problem started when I enabled Above 4g memory, before that it worked perfectly with 4GPUs

papampi

full member

Activity: 686

Merit: 140

Linux FOREVER! Resistance is futile!!!

Quote from: Zucca on September 25, 2017, 07:48:55 AM

Hello! I have a MSI Z270-A pro. I can run nvOS by default with 4 gtx 1060 (I plug 6 but it only detects 4) and it works perfectly. If I enable above 4g memory, it detects all the 6 gpus and starts mining well. But the OS hangs after 5 minutes, every time I try.
I disabled HQ audio as well.

What can be wrong?

Whats your CPU?
Did you enable cpu mining ?
whats your nvOC version?

Zucca

newbie

Activity: 81

Merit: 0

Hello! I have a MSI Z270-A pro. I can run nvOS by default with 4 gtx 1060 (I plug 6 but it only detects 4) and it works perfectly. If I enable above 4g memory, it detects all the 6 gpus and starts mining well. But the OS hangs after 5 minutes, every time I try.
I disabled HQ audio as well.

What can be wrong?

rayd89

newbie

Activity: 28

Merit: 0

Tried to use the "auto update" to get to 0019-1.1 and my machine failed to load, was stuck on black screen forever.
had to re-image back to 0019

hendivalentino

newbie

Activity: 31

Merit: 0

Quote from: damNmad on September 24, 2017, 05:07:56 PM

Quote from: hendivalentino on September 24, 2017, 05:04:07 PM

Quote from: damNmad on September 24, 2017, 04:51:35 PM

Quote from: hendivalentino on September 24, 2017, 04:44:26 PM

oh, i added in 1bash, sudo nano 1bash
and my version is v0019

Code:

UBQ_WORKER="$WORKERNAME"
UBQ_ADDRESS="my ubq wallet"
UBQ_POOL="stratum+tcp://us.ubiqpool.io:8008"

Code:

DOT_POOL_FORMAT_or_FORWARD_SLASH_POOL_FORMAT="DOT"

These lines already exist in 1bash; please check (you can see it after EXP coin details). You need to delete the newly added lines and edit the existing ones.

If you are using claymore the above settings should work.

If you still facing issue, when you see this error

Code:

ETH AUTHORIZATION FAILED.. stratum reading socket failed

please run this command:

Code:

ps aux | grep miner

& paste the result here; I can help you from that.

I don't see anything after EXP... that is why i added them manually..

That's odd!!!

If it wasn't there, then how come you were getting Authorization failure error??!!!

OK then please start mining UBQ (though you getting error) and paste the result of 'ps aux | grep miner' command here.

just send you a message

damNmad

full member

Activity: 378

Merit: 104

nvOC forever

Quote from: hendivalentino on September 24, 2017, 05:04:07 PM

Quote from: damNmad on September 24, 2017, 04:51:35 PM

Quote from: hendivalentino on September 24, 2017, 04:44:26 PM

oh, i added in 1bash, sudo nano 1bash
and my version is v0019

Code:

UBQ_WORKER="$WORKERNAME"
UBQ_ADDRESS="my ubq wallet"
UBQ_POOL="stratum+tcp://us.ubiqpool.io:8008"

Code:

DOT_POOL_FORMAT_or_FORWARD_SLASH_POOL_FORMAT="DOT"

These lines already exist in 1bash; please check (you can see it after EXP coin details). You need to delete the newly added lines and edit the existing ones.

If you are using claymore the above settings should work.

If you still facing issue, when you see this error

Code:

ETH AUTHORIZATION FAILED.. stratum reading socket failed

please run this command:

Code:

ps aux | grep miner

& paste the result here; I can help you from that.

I don't see anything after EXP... that is why i added them manually..

That's odd!!!

If it wasn't there, then how come you were getting Authorization failure error??!!!

OK then please start mining UBQ (though you getting error) and paste the result of 'ps aux | grep miner' command here.

Topic: [ mining os ] nvoc - page 214. (Read 418546 times)