Topic: [Mining OS] SimpleMining.net - Manage Your GPU farm the easy way! (30 days free) - page 272. (Read 836208 times)

How to save miner logs in SMOS ?

edit your /root/start.sh

modify line 16 like this :
   screen -L -dm -S miner bash -c "/root/utils/update_register.sh" &

now, all your screen output will be redirected to /home/miner/screenlog.0 so type :
  tail -f /home/miner/screenlog.0

and you will see the live output. in colors by the magic of gnu. Ctrl+C will get you back to shell.

Beware the file will grow fast, scp it somewhere and delete it periodically

You may have to edit the /root/start.sh file after SMOS updates

I can confirm that rigs with more than 8 GPUs behave a bit suspicious compared to my rocksolid 6-8 GPU rigs

is it possible to mine BTX?
Have 9x Radeon cards but i don't find any miner with timetravel10 in simpleos and Radeon-Cards

paD

Bug report:

tytanick, I have 11 rigs with 12/13 cards each. All of them Nvidia and all of them misbehaving in terms of processor load.
I believe I've figured it out: The majority of people won't notice it, but a LOT of the instability is caused by cronjobs executing before the previous one has finished.

When you have less than 10 cards, the nvidia-smi and related crontabs are executed and finished within the time before the next cron runs.
When you have > 10 cards, it takes a HECK of a long time for the crons to finish, there by putting tremendous load both on the CPU (and consuming boatloads of wattage / power) and the GPU's to continuously pull stats.

I highly recommend you write a master script / function that does a process listing prior to executing and only execute if the old process is done executing.

Hey guys I've been having two issues with SMOS.

First off, I'm going to say this out directly, all Hardware has been tested and stable for over 3 Days 24/7 operation or a week plus. I know there are NO hardware faults, no bad risers causing issues, etc. Ill list Specs below.

Rigs
G3930
H110 PRO BTC
EVGA BT 450w(For motherboard power)
Delta 2400w Server PSU (GPU/Riser Power)
Riser 007c,(LED lights including working)
Nvidia GPUS/AMD RX580's.
M.2 SSD. /Also tested with WD Raptor HDDs/16Gb 3.0 USB Drive
Overclocks all stable

H110 Motherboard is powered by BOTH molex connections AND a SATA connection for PCI-E lanes.

Here is my issue,

With Nvidia, NO MORE THAN, 11 GPUs will run stable for 24 hours. I've tried 13 GPUs but constant crashes, restarts etc every few hours. I've tried EVERYTHING hardware related. To test my theory I split my 13 GPU GTX 1070 rig into 6 and 7 rig(same risers, same motherboard, same overclocks etc) both ran beautifully for 48+ hours zero restarts. I recombined them back into 13 GPU rig, ran 23+ hours fine, then it started constantly crashing, restarting etc.

Here what I've tried
Flashing new USB Drive, Swapping to HDD, M.2 SSD.

I put down to 11, back to 48+ hours stablity, hell ran over 2 week stable with 11 GPUs.

I Got some GTX 1060's in, and decided to test again, 13+ GPUs. Absolute no go. Suffered same issues, crashes, restarts every few hours etc.

As of right now I'm blaming terrible NV drivers for Linux. I'm wondering if anyone else has 13+ Nvidia GPU rig? Has yours been running stable for over 48 hours with the same setup? For the life of me cannot get 13 GPUs to run with an Nvidia Rig.

On the other hand, I just purchased and setup a RX580 13 GPU rig mining ETH totally stable for 3+ days now. It's running great. My problem is when it Mines Monero.

Mining ETH I get 3+ days of stability. Mining Monero I'm luck to even get one day stable. Everything is EXACLTY the same. Same clocks, Same cards, same everything. Using Claymore ETH miner for ETH and Claymore XMR AMD for Monero. Everytime I run Monero, it will crash after a 6~ hours. When I restart the miner via SMOS. I will be down a card. It seems like a card will lose power or just stop running. I have to completely shut down the computer and restart it, then when the miner reload all cards will show up. mind you I didn't touch anything all i did do a cold boot for the miner rig.

I appreciate any idea or anyone having similar issues. I'm just stumped.

Have you checked the screen at the begining. You must see all expansion slots in green if not check wires risers.
I prefer use https://hiveos.farm with this moterboard I find too much fail now with SMOS.
On ETH for memory 2250 is TOO HIGH in my opinion I put 2100/2150 maximum
For weeks when SMOS doesn't reboot or reload when there is a problem. All is green but the rig is freeezing you must look at the console to see time is 2 or 3 hours sooner.

I ordered some B250 minig expert to migrate my other rigs on hiveos.
I have 5 rigs on SMOS at the moment 2 on HIVEOS. Now I have another problem with smos some rigs have ZERO h but it is OK for smos all is green so I must reboot manually the rigs with hiveos you can tuned your rig and if the rig is under the amount you put it reload or reboot automatically.

Tytanick I apreciate your OS it was the first I used but now there are too much fails and the gui is not friendly to use to change the pool, the adress, etc.. with hiveos you can mixed nvidia and amd on the same rig. Look at hiveos there are a lot of good ideas. Of course it was a very bad idea to limit report on the gpu we need to see all the information on the gpu.

I had also the fail with 511° and I must reboot.

Now I have a zero h on one rig



the dashboard on HIVEOS where I said 10 GPU but there is only 9 ;-)



I can use usb keys with hiveos on all my MB, on some MB with smos I must use SSD. On hiveos no fail "no screen attached"

explanation for WatchDog

WD = watch dog enable
M3 = if no activity 3 minutes miner reload
R5 = if no activity 5 minutes rig reboot
claymore x = 680 if one GPU h under 680 miner reload

What problems are you seeing?   I thought my problems were all Saphire 570 Nitro related....seems these new ELPIDA cards are tough to get right on the bios.  But I do have this TB250 Pro (only one)  all my other systems are ASUS Z270-A...........

Well, in my vast amount of wisdom and knowledge I discovered I didn't reconnect the ethernet cable. So I'm up and going fine now.

And now I have this error.

[ 305.438985] EXT4-fs error (device sda2): ext4_find_entry:1465: inode #14 comm accounts-daemon: reading lblock 0
[ 350.909357] EXT4-fs error (device sda2): ext4_find_entry:1465: inode #370 comm (ystemctl): reading directory lblock 0

Is it just me or is the Biostar TB250+ and Pro just not supported?  TB85 works like a champ

Well after seeing that screen for about 10 minutes and a couple restarts it just stopped. Although now it seems as though it is not communicating with SMOS. If anyone has an explanation as to why that would be great.

Now I'm getting the error as well though.

could not resolve hostname zen.suprnova.cc
Miner ended/crashed. Restarting miner in 10 seconds

It seems as though it is detecting everything fine and OC is set to 0/0 at the moment.


EDIT: And a restart started the initial problem again.
https://imgur.com/a/fktVb
Thanks for any help.

I am not a linux user.  But may I ask how to view the console screen  (running miner) via ssh
I tried ssh with user:root (or miner) and typed 'screen -r', nothing happened.

Thank you.

So I just got my 1000w PSU in and shut my system down, installed it as well as my other 2 1070ti's, totaling 3 now.

Started my rig back up and now it is giving me the following...

Started session *** of user miner.
Started user manager for UID 1000...
Stopping user manager for UID 1000.
Removed slice User Slice of miner.
Created slice user slice of miner.


I'm not sure what is causing this.

were you able to get yours back up and running?

Having issues with my rig and looking for some help, not familiar with Linux at all.

Currently mining Monero with 5x Sapphire Tech Nitro+ Rx 580 and after about 5 minutes or so getting GPU #2 hangs in Open CL. For the past few days have been using it with 4 cards and didn't have any problems. Any suggestions? Also, the fifth card was added into the back end rather than in between other cards if that matters..

Any help is appreciated.

When restarting two rigs they started download an "update" and then i got "there is no screen to be attached matching miner" problem...wtf

Yes I have 10+ systems with SMOS and am now experiencing this error "there is no screen to be attached matching miner" with some of the systems

Did something change with the code?

How is it possible to find which card got stopped the mining Rig in the SMOS? While the rig will be stopped so there is no SSH access to it until reboot it and after that, there is no log (SMOS will clear any log after each reboot...!!)

3930 and 4GB are enough, but I've only heard of people running up to 13 per rig on SmOS, B250 or not. With the B250, you must run 6 x P106 mining cards to get up to 19, and due to overall power complications, even with this possibility, success is rare.

Sure why not?! That's good while you don't need great CPU for GPU mining!
I've CPU G4400, 4GB DDR4 Ram, 12x Sapphire RX 470 Mining Edition 4GB, 2x PSU 1200W 80 Plus Platinum and at the end i've around 350 MH/s with Ethereum on Ethermine.org pool!