[OS] nvOC easy-to-use Linux Nvidia Mining

[TenaciousJ]

fullzero - I just got the new TB250 12 x PCI slot mobo, testing 12 x Zotac Minis 1070s and Zotac 12 x 1060 Mining cards.
I see the new bash file now goes up to 15 cards so it should run.
Will post results and pix in a few days. Thanks

I'm interested to see how it goes; I am projected to receive my 12x biostar tomorrow; but tracking number shows no data which is questionable.  I might end up getting the ASRock 14x mobo first.

Its already working with 12 x GA-RX470s with smOS in the ETH farm.
Cant wait to get nvOC running on this board with the Zotac 1070 Minis and Zotac P106-100s.
I wish I had more 2400w power supplies  

I can now confirm that the AsRock H110 13 GPU mobo works with at least 9 Nvidia GPUs (8x 1070 1x 980 ti) running nvOC without any additional modifications to the o/s other than running update and dist-upgrade.  Asrock's marketing limited the nvidia gpus to 5, which must be the max for windows, but linux runs at least 9 without any trouble at all.  If I had more power and cards I'd test it up to 13, but that'll have to wait a while.

[1714240521]

1714240521

[1714240521]

1714240521

[fullzero]

fullzero - I just got the new TB250 12 x PCI slot mobo, testing 12 x Zotac Minis 1070s and Zotac 12 x 1060 Mining cards.
I see the new bash file now goes up to 15 cards so it should run.
Will post results and pix in a few days. Thanks

I'm interested to see how it goes; I am projected to receive my 12x biostar tomorrow; but tracking number shows no data which is questionable.  I might end up getting the ASRock 14x mobo first.

Its already working with 12 x GA-RX470s with smOS in the ETH farm.
Cant wait to get nvOC running on this board with the Zotac 1070 Minis and Zotac P106-100s.
I wish I had more 2400w power supplies  

I can now confirm that the AsRock H110 13 GPU mobo works with at least 9 Nvidia GPUs (8x 1070 1x 980 ti) running nvOC without any additional modifications to the o/s other than running update and dist-upgrade.  Asrock's marketing limited the nvidia gpus to 5, which must be the max for windows, but linux runs at least 9 without any trouble at all.  If I had more power and cards I'd test it up to 13, but that'll have to wait a while.

I have had other members confirm the ASRock mobo works with 9x GPUs as well, I will test with 13 + 1 via m2 adapter when I get mine.

[fullzero]

I keep getting a database error when I try to update the OP:  I have pm'ed the Admins about it.  For now here is a link:


THIS IS THE NEWEST VERSION

LOCALorREMOTE_oneBash + additional files (includes newest Watchdog, SRR,  switch_v3, reboot, AutoTEMP, Claymore 9.7) Link





This has the newest version of all the files; and a global switch ( LOCALorREMOTE ) which works with all COIN selections and SALFTER_NICEHASH_PROFIT_SWITCHING :

# LOCAL will attach the mining process to the gnome or guake terminal
# REMOTE will leave it unattached / ready for SSH

LOCALorREMOTE="LOCAL"       # LOCAL  or  REMOTE


salfter your newest version of switch was causing problems when run with a monitor connected (LOCAL); I would recommend relying on the:

IAmNotAJeep_and_Maxximus007_WATCHDOG

to handle miner crashes / 0 hashrates.

I spent a couple hours testing this, and it is very effective;

Also even if your crashes are perfectly handled; if your OC is so high it crashes every 7 minutes or less: you are losing more time restarting the mining process then you are gaining with a slightly higher hashrate.


I did notice a bug in the Maxximus007_AUTO_TEMPERATURE_CONTROL when it attempts to lower pl due to fan 100 and temp rising:

/home/m1/Maxximus007_AUTO_TEMPERATURE_CONTROL: line 119: [: missing `]'
/home/m1/Maxximus007_AUTO_TEMPERATURE_CONTROL: line 130: let: NEW_POWER_LIMIT=75.00-5: syntax error: invalid arithmetic operator (error token is ".00-5")
Missing value for -pl argument. Please run 'nvidia-smi -h' for help.

In v0018; at firstboot, oneBash will be copied from the windows partition to the home directory (folder)  dos2unix will run on it and the windows partition will be zeroized.

This version of oneBash and files are all in line with that setup.  I realize that 2unix will also need to be changed after adding the new files and oneBash.

In the home folder there is a file called:

2unix

find it and open it in gedit:

it should look like:

Code:

#!/bin/bash

cd /tmp

pasteBIN="NO"

pasteBASH="gjjUqLXd"

if [ $pasteBIN == "YES" ]
then
wget https://pastebin.com/raw/$pasteBASH
cp $pasteBASH '/media/m1/1263-A96E/oneBash'
rm $pasteBASH
fi

sudo dos2unix /media/m1/1263-A96E/oneBash

sleep 2

bash '/media/m1/1263-A96E/oneBash'

change this to:

Code:

#!/bin/bash

cd /tmp

pasteBIN="NO"

pasteBASH="z7SH8VzS"

if [ $pasteBIN == "YES" ]
then
wget https://pastebin.com/raw/$pasteBASH
cp $pasteBASH '/home/m1/oneBash'
rm $pasteBASH
sleep 10
fi

bash '/home/m1/oneBash'

and save; this should change to using the new oneBash you copied to the home directory (folder)

modifiy the new oneBash in home with your settings / addresses

and you should be good to go with the new files / oneBash

[kw1k]

I keep getting a database error when I try to update the OP:  I have pm'ed the Admins about it.  For now here is a link:


THIS IS THE NEWEST VERSION

LOCALorREMOTE_oneBash + additional files (includes newest Watchdog, SRR,  switch_v3, reboot, AutoTEMP, Claymore 9.7) Link





This has the newest version of all the files; and a global switch ( LOCALorREMOTE ) which works with all COIN selections and SALFTER_NICEHASH_PROFIT_SWITCHING :

# LOCAL will attach the mining process to the gnome or guake terminal
# REMOTE will leave it unattached / ready for SSH

LOCALorREMOTE="LOCAL"       # LOCAL  or  REMOTE


salfter your newest version of switch was causing problems when run with a monitor connected (LOCAL); I would recommend relying on the:

IAmNotAJeep_and_Maxximus007_WATCHDOG

to handle miner crashes / 0 hashrates.

I spent a couple hours testing this, and it is very effective;

Also even if your crashes are perfectly handled; if your OC is so high it crashes every 7 minutes or less: you are losing more time restarting the mining process then you are gaining with a slightly higher hashrate.


I did notice a bug in the Maxximus007_AUTO_TEMPERATURE_CONTROL when it attempts to lower pl due to fan 100 and temp rising:

/home/m1/Maxximus007_AUTO_TEMPERATURE_CONTROL: line 119: [: missing `]'
/home/m1/Maxximus007_AUTO_TEMPERATURE_CONTROL: line 130: let: NEW_POWER_LIMIT=75.00-5: syntax error: invalid arithmetic operator (error token is ".00-5")
Missing value for -pl argument. Please run 'nvidia-smi -h' for help.

In v0018; at firstboot, oneBash will be copied from the windows partition to the home directory (folder)  dos2unix will run on it and the windows partition will be zeroized.

This version of oneBash and files are all in line with that setup.  I realize that 2unix will also need to be changed after adding the new files and oneBash.

In the home folder there is a file called:

2unix

find it and open it in gedit:

it should look like:

Code:

#!/bin/bash

cd /tmp

pasteBIN="NO"

pasteBASH="gjjUqLXd"

if [ $pasteBIN == "YES" ]
then
wget https://pastebin.com/raw/$pasteBASH
cp $pasteBASH '/media/m1/1263-A96E/oneBash'
rm $pasteBASH
fi

sudo dos2unix /media/m1/1263-A96E/oneBash

sleep 2

bash '/media/m1/1263-A96E/oneBash'

change this to:

Code:

#!/bin/bash

cd /tmp

pasteBIN="NO"

pasteBASH="z7SH8VzS"

if [ $pasteBIN == "YES" ]
then
wget https://pastebin.com/raw/$pasteBASH
cp $pasteBASH '/home/m1/oneBash'
rm $pasteBASH
sleep 10
fi

bash '/home/m1/oneBash'

and save; this should change to using the new oneBash you copied to the home directory (folder)

modifiy the new oneBash in home with your settings / addresses

and you should be good to go with the new files / oneBash

Ive done all of this my mining process starts and then it leaves the screen and only shows nvidia-smi over and over and over

[ADN]

Hello,
I'm using a Z87 Fatal1ty Killer, the OS loads and then I'm getting a reboot loop caused by this line.

Code:

if [ $XORG == "FAIL" ]
then
sudo cp '/etc/X11/xorg.conf.backup' '/etc/X11/xorg.conf'
echo ''
echo "Xorg PROBLEM DETECTED"
echo ''
echo "Restoring Xorg"
echo ''
echo "Rebooting in 5"
sleep 5
sudo reboot
fi

Any ideea what is the cause? It should stop after the first reboot if I'm reading the script correctly.

[nklak]

Hello,
I'm using a Z87 Fatal1ty Killer, the OS loads and then I'm getting a reboot loop caused by this line.

Code:

if [ $XORG == "FAIL" ]
then
sudo cp '/etc/X11/xorg.conf.backup' '/etc/X11/xorg.conf'
echo ''
echo "Xorg PROBLEM DETECTED"
echo ''
echo "Restoring Xorg"
echo ''
echo "Rebooting in 5"
sleep 5
sudo reboot
fi

Any ideea what is the cause? It should stop after the first reboot if I'm reading the script correctly.

Yes,

please do not connect monitor to the motherboard...also make sure that in BIOS integrated graphic is disabled...this causes error. Please make new image on usb stick and connect monitor to only GPU in the main x16 slot (GPU0).

Best regards

[WarwickNZ]

Hey Everyone,

Just wondering what additional switches if any you use with the Genoil miner for the most consistent share count? So far for me I've only added "--farm-recheck 10000" since switching back from Claymore as I don't think it needs to check as frequently because It's pushed from the pool if i'm thinking correctly, seems a bit more stable.

Thanks as always, awesome project.

[ADN]

Hello,
I'm using a Z87 Fatal1ty Killer, the OS loads and then I'm getting a reboot loop caused by this line.

Code:

if [ $XORG == "FAIL" ]
then
sudo cp '/etc/X11/xorg.conf.backup' '/etc/X11/xorg.conf'
echo ''
echo "Xorg PROBLEM DETECTED"
echo ''
echo "Restoring Xorg"
echo ''
echo "Rebooting in 5"
sleep 5
sudo reboot
fi

Any ideea what is the cause? It should stop after the first reboot if I'm reading the script correctly.

Yes,

please do not connect monitor to the motherboard...also make sure that in BIOS integrated graphic is disabled...this causes error. Please make new image on usb stick and connect monitor to only GPU in the main x16 slot (GPU0).

Best regards

This is kind of a bad idea because GPU cards are known to fail occasionally, if the GPU card that I'm connected to failed it can cause a big problem managing the platform.

Any workaround around this limitation? It should be possible to use the on-board GPU without any issues.

[fullzero]

I keep getting a database error when I try to update the OP:  I have pm'ed the Admins about it.  For now here is a link:


THIS IS THE NEWEST VERSION

LOCALorREMOTE_oneBash + additional files (includes newest Watchdog, SRR,  switch_v3, reboot, AutoTEMP, Claymore 9.7) Link





This has the newest version of all the files; and a global switch ( LOCALorREMOTE ) which works with all COIN selections and SALFTER_NICEHASH_PROFIT_SWITCHING :

# LOCAL will attach the mining process to the gnome or guake terminal
# REMOTE will leave it unattached / ready for SSH

LOCALorREMOTE="LOCAL"       # LOCAL  or  REMOTE


salfter your newest version of switch was causing problems when run with a monitor connected (LOCAL); I would recommend relying on the:

IAmNotAJeep_and_Maxximus007_WATCHDOG

to handle miner crashes / 0 hashrates.

I spent a couple hours testing this, and it is very effective;

Also even if your crashes are perfectly handled; if your OC is so high it crashes every 7 minutes or less: you are losing more time restarting the mining process then you are gaining with a slightly higher hashrate.


I did notice a bug in the Maxximus007_AUTO_TEMPERATURE_CONTROL when it attempts to lower pl due to fan 100 and temp rising:

/home/m1/Maxximus007_AUTO_TEMPERATURE_CONTROL: line 119: [: missing `]'
/home/m1/Maxximus007_AUTO_TEMPERATURE_CONTROL: line 130: let: NEW_POWER_LIMIT=75.00-5: syntax error: invalid arithmetic operator (error token is ".00-5")
Missing value for -pl argument. Please run 'nvidia-smi -h' for help.

In v0018; at firstboot, oneBash will be copied from the windows partition to the home directory (folder)  dos2unix will run on it and the windows partition will be zeroized.

This version of oneBash and files are all in line with that setup.  I realize that 2unix will also need to be changed after adding the new files and oneBash.

In the home folder there is a file called:

2unix

find it and open it in gedit:

it should look like:

Code:

#!/bin/bash

cd /tmp

pasteBIN="NO"

pasteBASH="gjjUqLXd"

if [ $pasteBIN == "YES" ]
then
wget https://pastebin.com/raw/$pasteBASH
cp $pasteBASH '/media/m1/1263-A96E/oneBash'
rm $pasteBASH
fi

sudo dos2unix /media/m1/1263-A96E/oneBash

sleep 2

bash '/media/m1/1263-A96E/oneBash'

change this to:

Code:

#!/bin/bash

cd /tmp

pasteBIN="NO"

pasteBASH="z7SH8VzS"

if [ $pasteBIN == "YES" ]
then
wget https://pastebin.com/raw/$pasteBASH
cp $pasteBASH '/home/m1/oneBash'
rm $pasteBASH
sleep 10
fi

bash '/home/m1/oneBash'

and save; this should change to using the new oneBash you copied to the home directory (folder)

modifiy the new oneBash in home with your settings / addresses

and you should be good to go with the new files / oneBash

Ive done all of this my mining process starts and then it leaves the screen and only shows nvidia-smi over and over and over

possible reasons:

you are using LOCAL and: SALFTER_NICEHASH_PROFIT_SWITCHING which attaches the screen in the guake terminal (press f12)

There is already an instance running in a screen that isn't attached; and thus the screen isn't attaching, but launching more instances instead.

You can kill all mining processes by entering this in a terminal:

Code:

pkill -e miner

alternatively you can reboot your rig.

you are using REMOTE and so the screen is not attached.

[fullzero]

This is kind of a bad idea because GPU cards are known to fail occasionally, if the GPU card that I'm connected to failed it can cause a big problem managing the platform.

If a GPU hangs; there is a problem regardless of how your monitor is attached.

nvOC can be used full headless; you don't need to use a monitor.

In v0018 nvOC will have; mining client error mitigation, GPU soft crash mitigation, and if you have an SRR: hard crash mitigation.  As far as I know this is the most advanced of any mining OS.

Any workaround around this limitation? It should be possible to use the on-board GPU without any issues.

This has previously been requested by another member, and it has been added to the list.  I will make a version that supports integrated graphics eventually; but it is not a priority as it is not really a limitation.  It is only a limitation to those who are inclined to use the integrated graphics.

A real limitation is: only being able to support 8x GPUs when there are 13x GPU motherboards.

BTW I just tested the:

ASRock H110 PRO BTC+ and have confirmed that v0017 supports 13x GPUs out the box with no bios changes. 

As the Biostar 12x GPU mobo uses the same chipset I would expect it to be fully supported as well.

[Avarets]

Waiting for Biostar tb250 pro and 12 Zotac P106-100 to start mining using this OS.
Is it possible to quickly update to the latest version if it is installed on hdd/ssd?

[fullzero]

Waiting for Biostar tb250 pro and 12 Zotac P106-100 to start mining using this OS.
Is it possible to quickly update to the latest version if it is installed on hdd/ssd?

I will release v0018 soon.  Currently you must reimage to upgrade; I will add an updater in a later version.

You can change v0017 into almost v0018 by following this post:

https://bitcointalk.org/index.php?topic=1854250.msg20099702#msg20099702

However; v0018 will support any storage type; USB key, SSD, HDD, m2 SSD ect. So, I would recommend waiting to download v0018 if you don't have your components yet and plan on using a HDD or SSD.

[ADN]

Thank you.

Can you add an additional condition regarding GPU temp?

if (CURRENT_TEMP == TARGET_TEMP and CURRENT_FAN_SPEED == MAX) then POWER_LIMIT = POWER_LIMIT-10% ?

Or something similar, do prevent GPU's from crashing it they reach max temperature and fans can no longer catch up. I realize that it involves multiple variables not just the simple condition above, but should be a good start.

P.S. Are you planing to release v18 today?

[dittie]

Can you add SIBCOIN ?

Thanks!

[idle_shell]

Waiting for Biostar tb250 pro and 12 Zotac P106-100 to start mining using this OS.
Is it possible to quickly update to the latest version if it is installed on hdd/ssd?

I will release v0018 soon.  Currently you must reimage to upgrade; I will add an updater in a later version.

You can change v0017 into almost v0018 by following this post:

https://bitcointalk.org/index.php?topic=1854250.msg20099702#msg20099702

However; v0018 will support any storage type; USB key, SSD, HDD, m2 SSD ect. So, I would recommend waiting to download v0018 if you don't have your components yet and plan on using a HDD or SSD.

Excellent!  Was just going to post about some challenges I was having getting the 0017 image running (oneBash loops repeatedly before ever starting the miner).  But sounds like we're close to 0018--which I hope solves my issues.

Sent you some hashes last night testing gear.  You're doing awesome work here!

[vlasic]

Add, pls, SIB (Sibcoin) and CHC (Chaincoin) support.

THNX

[Maxximus007]

Can you add SIBCOIN ?

Thanks!

As far as I know that is X11, not in nvOC. Might be better to use an ASIC for that?

[Maxximus007]

Thank you.

Can you add an additional condition regarding GPU temp?

if (CURRENT_TEMP == TARGET_TEMP and CURRENT_FAN_SPEED == MAX) then POWER_LIMIT = POWER_LIMIT-10% ?

Or something similar, do prevent GPU's from crashing it they reach max temperature and fans can no longer catch up. I realize that it involves multiple variables not just the simple condition above, but should be a good start.

P.S. Are you planing to release v18 today?

Something similar will be working in V0018 I believe.

[jlbaseball11]

Does anyone know why all shares are being rejected because of "low difficulty share?"  It only happens with DMD and ZCOIN.  Every other coin works fine.  I've tried different pools and oc settings with no change.

[TenaciousJ]

If a GPU hangs; there is a problem regardless of how your monitor is attached.

nvOC can be used full headless; you don't need to use a monitor.

In v0018 nvOC will have; mining client error mitigation, GPU soft crash mitigation, and if you have an SRR: hard crash mitigation.  As far as I know this is the most advanced of any mining OS.

I was just about to ask you if it would be possible to add some detection so that if your miner performance drops to 0 sol/s, or falls below some threshold setting, that it would automatically restart..  had a system wide gpu reset sometime in the middle of the night and the whole system just posted zeros for hours.

BTW I just tested the:

ASRock H110 PRO BTC+ and have confirmed that v0017 supports 13x GPUs out the box with no bios changes. 

As the Biostar 12x GPU mobo uses the same chipset I would expect it to be fully supported as well.

Sweet.... now if I could only find some GPUs to buy lol