Author

Topic: [Mining OS] SimpleMining.net - Manage Your GPU farm the easy way! (30 days free) - page 534. (Read 835838 times)

sr. member
Activity: 652
Merit: 266
Do you support nvidia cards?
I already sent him my nvidia scripts, so expect soon support for nvidia Smiley
full member
Activity: 154
Merit: 100
Do you support nvidia cards?
sr. member
Activity: 652
Merit: 266
This problem can easily be solved by making an initrd environment and loading it to ram using ramfs where execution of scripts will not bother actual filesystem.
Yep, its already there but still, updates needs to be write to usb , and that is weak spot Smiley
Already using tmpfs for everything Smiley

miner@simpleminer:~$ df -h
Filesystem      Size  Used Avail Use% Mounted on
udev            3.8G     0  3.8G   0% /dev
tmpfs           767M   86M  682M  12% /run
/dev/sda2       6.5G  2.0G  4.5G  31% /
tmpfs           3.8G     0  3.8G   0% /dev/shm
tmpfs           5.0M     0  5.0M   0% /run/lock
tmpfs           3.8G     0  3.8G   0% /sys/fs/cgroup
tmpfs           300M   16K  300M   1% /tmp
tmpfs           3.8G   12K  3.8G   1% /var/log
tmpfs           3.8G   96M  3.7G   3% /var/tmp
tmpfs           3.8G     0  3.8G   0% /var/log/apt
/dev/sda1       8.9M  3.5K  8.9M   1% /mnt/user
tmpfs           767M     0  767M   0% /run/user/1000

I'd suggest different approach for this.
Switch to mLibc or ucLibc 64bit strap the default config file to initrd file and use dd to update to new images. I think that old projects like m0n0wall and pfSense are good starting point to do unified system. I am already onto this but only in my farm.
legendary
Activity: 2660
Merit: 1096
Simplemining.net Admin
This problem can easily be solved by making an initrd environment and loading it to ram using ramfs where execution of scripts will not bother actual filesystem.
Yep, its already there but still, updates needs to be write to usb , and that is weak spot Smiley
Already using tmpfs for everything Smiley

miner@simpleminer:~$ df -h
Filesystem      Size  Used Avail Use% Mounted on
udev            3.8G     0  3.8G   0% /dev
tmpfs           767M   86M  682M  12% /run
/dev/sda2       6.5G  2.0G  4.5G  31% /
tmpfs           3.8G     0  3.8G   0% /dev/shm
tmpfs           5.0M     0  5.0M   0% /run/lock
tmpfs           3.8G     0  3.8G   0% /sys/fs/cgroup
tmpfs           300M   16K  300M   1% /tmp
tmpfs           3.8G   12K  3.8G   1% /var/log
tmpfs           3.8G   96M  3.7G   3% /var/tmp
tmpfs           3.8G     0  3.8G   0% /var/log/apt
/dev/sda1       8.9M  3.5K  8.9M   1% /mnt/user
tmpfs           767M     0  767M   0% /run/user/1000
sr. member
Activity: 652
Merit: 266
Hi.
Yep, that is worth mentioning accually.
I always used more expensive usb thumbs.
a-data and corsair ones are best for me but also its worth to buy pendrives which have 10-20MB/s write speed at least.
Also if you have any problems, try inserting pendrive to usb 2.0 instead of 3.0
With pendrives there is also other problem. They have much shorter life period but with more expensive i had really few problems.
Anyway SSD is best choice but still ssd are much more expensive. (I am using all pendrives Smiley
Bet would be buing very cheap ssd, meaby i could find it and start selling to people with my OS. (still dont know if that would make economic sense)

This problem can easily be solved by making an initrd environment and loading it to ram using ramfs where execution of scripts will not bother actual filesystem.
legendary
Activity: 2660
Merit: 1096
Simplemining.net Admin
Hi.
Yep, that is worth mentioning accually.
I always used more expensive usb thumbs.
a-data and corsair ones are best for me but also its worth to buy pendrives which have 10-20MB/s write speed at least.
Also if you have any problems, try inserting pendrive to usb 2.0 instead of 3.0
With pendrives there is also other problem. They have much shorter life period but with more expensive i had really few problems.
Anyway SSD is best choice but still ssd are much more expensive. (I am using all pendrives Smiley
Bet would be buing very cheap ssd, meaby i could find it and start selling to people with my OS. (still dont know if that would make economic sense)
legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'

-is it RX OS ?
- what gpus ?
- what core and memory clocks ?
- do they have custom bios with modded timings ?
- which miner are you using ? optiminer ? mining zcash/komodo ?

3x RX 480 8GB (2 powered and 1 unpowered risers)
Standard bios, clocks,...
Optiminer - zcash

The beta kernel is better, but not ideal. It may be a day fine, and the next day I must 2-5 times a day to restart the rig.
Today I switch from USB to SATA and 3 hours without errors for now.

Hey, I do not have any more problems after making the transition from USB to SSD, I have a beta kernel. (4 days for now, only 1 restart for OC)

Yep.  The reason is quite a bit of variance in USB 3.0 sticks.

I have sticks new sealed rated for up to 130 Mbps they wrote at 10 Mbps
I have sticks new sealed rated for up to 100 Mbps they wrote at 50 Mbps.

That is wrong and of course the underperforming sticks sucked at running the os.

With that in mind I went to all ssd and or pcie ssds.

I am happy with simplemining so.
full member
Activity: 194
Merit: 100

-is it RX OS ?
- what gpus ?
- what core and memory clocks ?
- do they have custom bios with modded timings ?
- which miner are you using ? optiminer ? mining zcash/komodo ?

3x RX 480 8GB (2 powered and 1 unpowered risers)
Standard bios, clocks,...
Optiminer - zcash

The beta kernel is better, but not ideal. It may be a day fine, and the next day I must 2-5 times a day to restart the rig.
Today I switch from USB to SATA and 3 hours without errors for now.

Hey, I do not have any more problems after making the transition from USB to SSD, I have a beta kernel. (4 days for now, only 1 restart for OC)
sr. member
Activity: 652
Merit: 266
Same here... i think it is a optiminer problem..

In my optionion, Optiminer or meaby alghoritm itself has problem (but optiminer more likely) and it just hangs gpu so it stays at 300mhz Tongue
I still dont have time to test this problems becasue i need to write inscructions for SRR becasue shipping is going right now Tongue
Accually all packages should be sent withing 7 days.



Try decreasing intensity by setting -i 4/3/2 until you find the sweet spot of stability. All my RX cards suffer from default intensity level, I have reduced it to 4 and it seems stable.
Until there is no open source miner noone can guarantee stability for your rigs. Also keep in mind that amdgpu/amdgpu-pro has poor powerplay(amd overdrive) support and I really don't advise you to clock the cards using sysfs, for now just do only bios mods without using sysfs dpm clocks.
legendary
Activity: 2660
Merit: 1096
Simplemining.net Admin
Same here... i think it is a optiminer problem..

In my optionion, Optiminer or meaby alghoritm itself has problem (but optiminer more likely) and it just hangs gpu so it stays at 300mhz Tongue
I still dont have time to test this problems becasue i need to write inscructions for SRR becasue shipping is going right now Tongue
Accually all packages should be sent withing 7 days.

full member
Activity: 237
Merit: 100
Ok guys, i figured it out. For those who have stuck gpus on optiminer showing  300mhz, what you have to do is run claymore first for couple of minutes where all gpus should run normal and then switch to optiminer. All gpus should work flawlesly. If you need to restart miner again on optiminer see if it shows 300mhz again. If it does just repeat the process.
Let me know if that worked for you

Same here... i think it is a optiminer problem..
legendary
Activity: 2660
Merit: 1096
Simplemining.net Admin
Would greatly appreciate you looking into the Optiminer issue.  Thanks so much for your hard work on all of this.

How about writing into optiminer thread - asking if they have similar issues ?
full member
Activity: 189
Merit: 100
Ok guys, i figured it out. For those who have stuck gpus on optiminer showing  300mhz, what you have to do is run claymore first for couple of minutes where all gpus should run normal and then switch to optiminer. All gpus should work flawlesly. If you need to restart miner again on optiminer see if it shows 300mhz again. If it does just repeat the process.
Let me know if that worked for you

I will need to sit on this and debug whats going on with this optiminer Tongue

is there a way to flash bios on my rx480´s rig? or any conf. to get 30mh from each card?
You should flash your rigs using windows.
In future there wil be flashing tool for linux.
Also there wil be in future option to set specific clocks in each gpu.


Would greatly appreciate you looking into the Optiminer issue.  Thanks so much for your hard work on all of this.
legendary
Activity: 2660
Merit: 1096
Simplemining.net Admin

Update v1064:
- fixed "GPU_MAX_ALLOC_PERCENT is not specified!"
legendary
Activity: 2660
Merit: 1096
Simplemining.net Admin
Ok guys, i figured it out. For those who have stuck gpus on optiminer showing  300mhz, what you have to do is run claymore first for couple of minutes where all gpus should run normal and then switch to optiminer. All gpus should work flawlesly. If you need to restart miner again on optiminer see if it shows 300mhz again. If it does just repeat the process.
Let me know if that worked for you

I will need to sit on this and debug whats going on with this optiminer Tongue

is there a way to flash bios on my rx480´s rig? or any conf. to get 30mh from each card?
You should flash your rigs using windows.
In future there wil be flashing tool for linux.
Also there wil be in future option to set specific clocks in each gpu.
full member
Activity: 237
Merit: 100
is there a way to flash bios on my rx480´s rig? or any conf. to get 30mh from each card?
newbie
Activity: 6
Merit: 0
Ok guys, i figured it out. For those who have stuck gpus on optiminer showing  300mhz, what you have to do is run claymore first for couple of minutes where all gpus should run normal and then switch to optiminer. All gpus should work flawlesly. If you need to restart miner again on optiminer see if it shows 300mhz again. If it does just repeat the process.
Let me know if that worked for you
legendary
Activity: 1834
Merit: 1080
---- winter*juvia -----
<>

@Citronik - Are your Sapphire RX480 the reference style ones?  If those are stable, I might convert those over to Simple OS, So far I've only tested non ref RX470 / 480  

Sapphire RX480s are all reference cards (bought 2 weeks after launched).
Its running very stable but fans are blower style -- so fans gets loud during ETH mining.
But it works for me because I hang them vertically open cage rigs style, with blowers pointed up.
Temps and wattage - not that bad during ZEC and XMR mining.

My GPU farm currently does the following format -- all 32 rigs are on smOS, 1 x Win10 and 1 x PandaOS for the Pandaminer.

1. All Sapphire RX480s (66 x GPUs) -- ETH mining, Claymore v7.4, with -ethi 16 and -fanmax 99 settings

2. All Non-RX480 (45 x Sapphire R9-Nanos, 25 x MSI-R9-390s) -- ZEC mining, Claymore v11.1, -i 8 and -fanmax 99 settings, and XMR mining with sgminer-gm

3. All Development rigs (Asus 290s, 290x, 370s and MSI-980Ti) -- 3 x dev rigs, 2 on smOS and 1 on Windows 10.

4. 1 x Pandaminer, with Eliovp PandaOS and custom ROM. Currently mining XMR/sgminer-gm

5. All RX480s and R9-Nanos, have their roms modded with Eliovp low energy custom roms. 290s on Stilt roms.

6. smOS rigs divided into 2 groups/2 emails; RX and Non-RX so I can mix and match mining priorities.

7. For now, all my rigs are selling hash at Nicehash..... until coin profitability improves.
sr. member
Activity: 414
Merit: 251
Here is 1 issue, I simply cannot figure out.  Rig 105 is showing as offline but have been hashing still for the past 3 days straight.  If I go and delete it, it will come back as Noname so I know it's the correct rig.  It would also go online for 5-30mins if I reboot but then will show offline like this but will continue to mine for days on pool side,


Does this ris in on OTHER internet connection than the rest ?
Now there are some dns/network route problems to OVH i suppose.

Edit: . That means, miner cant connect to simplemining and send console and other updates, so you see it offline.
But it is hashing anyway.
Even if simplemining server is down, rigs still are hashing !
But as now there is propably route problem in network.

Rig sit in the same site location as the rest of the other rigs.  All are Fury rigs as that's the only image that is stable for me right now.  Yes, it's just a minor annoyance but it's still hashing.  Furies are solid on Simple Mining OS. 

@Citronik - Are your Sapphire RX480 the reference style ones?  If those are stable, I might convert those over to Simple OS, So far I've only tested non ref RX470 / 480  
full member
Activity: 189
Merit: 100
Also guys i still cannot figure out why are you mining ZEC on RX470/RX480

ETH/ETC is more profitable and causes less problems..

Or meaby you are mining or 380 , then its propalby better to miner ZEC, ok


EDIT. Also guys renember, if most people dont have problem, and you have problem, then its propably riser,gpu,power,timings,undervolt,pool etc etc Grin
Always check stability on lower clocks and no undervolting.
And renember that optiminer causes many problems and propably claymore-zet too as it must have less memory clocks set to work stable.

I'm not sure it matters what coin we are mining.  Some of us may want to mine Zec regardless as the masses move to other coins the difficulty decreases and profitability follows suit.
Jump to: