Topic: ethOS - Ethereum Mining Platform - page 8. (Read 94540 times)

Ethos LIVE Team Hangout – The Path To Mass Adoption .A LIVE chat with some of the Ethos team about how we're building the future of finance, Ethos Universal Wallet and some special announcements! Moderated by Amateo Ra as part of Crypto Virtual Summit.
https://www.youtube.com/watch?v=vrXt0iByxQc

Not sure what cards you are using, core and power settings vary between AMD and Nvidia, but for memclock you would definitely put:

globalmem 2000

Though I've found that varies too, for my cards it seems to be more effective to use full values for AMD and + or - for Nvidia, so if your base memory clock is, say, 1600, you would put:

globalmem +400

You can also break it down card to card, which is what the sample format shows, but I only do that if I have cards with different memory types in the rig, the variation isn't worth the trial and error for me (ymmv).

As far as power limiting goes, though, I can't give you specifics but for AMD you would play more with the core/globalcore setting, for Nvidia the globalpowertune. For ethash (Ethereum and similar coins), I have my 1060s (Nvidia) set like this:

globalmem +450                                                                                                                                          
globalpowertune 75

Which, on my Samsung memory cards, gives me about 22mh/s and 75W power draw.

For my 580s (AMD), my settings are:

globalcore 1200                                                                                                                                        
globalmem 2200                                                                                                                                          
globalpowertune 2

Which gives me about 28mh/s and rough estimate of 115W per card (harder to figure out because the dashboard only reports voltage for AMD).

These settings will MOST LIKELY NOT be optimal for your cards, but they are a place to start if you need one (though I suggest starting lower for memory, higher for power settings and doing some testing). Subsequently you can figure out your card's optimal settings using tools in windows and then use them as a starting point, and back off a little if there is any instability. Good luck!

EDIT: Also note that you will want to play with these settings if you switch algos, as each algo demands are different.

thanks for the link, but i couldn't manage it. tried already before.

could you please put it in ethos local.conf version?


I also read the pool.txt

as far as i read, there is no sub file since last few versions of ethos.

My claymore settings are like this.

could you please put it in ethos local.conf version?

Use the settings in your claymore stub in windows in your ethos config file instead.

If you have a computer on the same network you don't need it in ethOS, you just need it on that computer. I use OSX's desktop sharing or whatever it's called (built into screen sharing, just requires some tweaking) so when I leave town I just log into the gui of my server and then use the ethOS dashboard links to shell-in-a-box. Or you can just ssh into your ethOS machines from anywhere, I just find this method easier...

Yes

http://ethosdistro.com/kb/#overclocking

I am getting very good result with windows claymore. 28.5 MH/S stable hash rate.
But i get only 23 mh/s with ethos.

https://prntscr.com/imnsah

How should i do the overclock?

After update to 1.3.0  dpm state settings don't work anymore.
All gpus set to dpm state 2 but dmesg show dpm state 7. Even with forced voltage  dpm stays at 7.

Is there a way force dpm state 2 within amdgpu driver?

I tried to install google remote desktop but i didn't allow me to install .deb file.
Is google remote desktop possible with Ethos?

This mining method is really good, I need to learn to use.thanks!

Have you tried installing it? Not sure it would work with the config file, but it may. Use the Linux install instructions.

Ton of good changes, but STILL no bminer in ethos 

ethOS 1.3.0 changelog (released February 28th 2018)

Major Changes
Added support for certain AMDGPU RX series GPUs that displayed 511C and crashed via installing experimental kernels and drivers.
Added support for RX 550 via installing experimental kernels and drivers.
Added support for NVIDIA Titan V, GTX 1060 5GB, GTX 1060 Rev.2 via installing experimental kernels and drivers.
Added more support staff to #ethosdistro/Freenode IRC channel
Added /gethelp shortcut in on-rig IRC clients so that users can generate diagostics links by typing /gethelp directly in channel.
Updated all installed miners to their latest versions and added packages necessary to support them.
Created automatic driver packager for NVIDIA and created NVIDIA driver set for version 390.25.
Added CDN load balancer to ethOS download page in order to serve the ethOS download file from the server that is closest to the user.

Bug Fixes
Fixed cases where ethminer hashrate under 10 megahashes would not display hashrate correctly.
Added blake2s and keccak coins to dualminer reporting.
Added better support for updating from very old ethOS versions.
Added necessary fglrx and nvidia patches to kernel 4.15
Added an automated CPU performance manager in order to reduce CPU load when running many NVDIA GPUs.
Fixed cases where ethminer proxy would not restart correctly.
Fixed many cases of false positive displays of "gpu clock problem".
Fixed a bug that caused hash collection to time out if system load was too high.
Fixed a bug that prevented stats from reporting if too many rigpools were specified.
Allowed ethOS config parameters to work even if they have preceding spaces.
Fixed a bug that prevented poolemail from applying correctly when dual mining in certain cases.
Added driver patch for USB NIC into kernel sources.
Added local JSON stats dump to /var/run/ethos/stats.json
Added install-amd-beta and install-nv-beta commands to the pull latest beta drivers from the ethOS repository.
Added hotfix command for use with future hotfixes.
Sanitized overclock values in overclock scripts.
Fixed a case where the display monitor shows yellow text and blank screen during the boot process.
Fixed regex issues associated with the poolemail config option, which had parsing problems under certain conditions.
Fixed a bug in the AMDGPU (RX) overclock script that could cause crashes if applied after booting.
Updated amdgpu-pro OPENCL stack to latest stable version.
Made configmaker.com perform more reliably by upgrading its server.
Added support for failover pools in dstm-zcash

Cosmetic Changes and Cleanup
Added the ethOS license agreement to the ethOS desktop and MOTD, which disappears once accepted.
Added prompt if user attemps to run ethos-update without a working Internet connection.
Added "possible miner stall" message to ethOS status if miner has been working for a while, but hash is zero.
Added list of any crashed GPUs into the ethOS status message.
Corrected the bash prompt hashrate display to display "0" if the miner stalled or was otherwise not mining.
Removed color codes from miner output files in gethelp diagnostic files, so that logs are easier to read.
Added warnings and advice if user attempted to use an ethOS config option as a linux command.
Allowed show miner log to automatically clear if the miner changed.
Fixed cases where the NVIDA gpulist pci bus identification was not accurate.
Fixed the "lockscreen" button on the right hand side of the taskbar in the ethOS desktop.
Revamped ethOS desktop menus and added a "hard-reboot" option into the ethOS desktop interface.
Added Shutdown and Restart buttons to ethOS desktop, which were accidentally removed in a previous version.
Added notices to ethOS desktop and ethOS status if an invalid miner is specified.
Removed silentarmy miner due to no active development from author.
Added collection method for unique rig UUID.
Added collection method for ccminer flags and dualminer_pool
Added ethos-smi output to gethelp diagnostics link.
Added motherboard BIOS version to gethelp diagnostics link, ethOS panel, and JSON API.
Changed "#" ethOS bash prompt to "$" ethOS bash prompt (to see clear difference between root and ethos users).
Added list of common ethOS commands to the ethOS Knowledge Base.
Added pci.ids for GTX 1060 5GB and GTX 1060 Rev.2
Added ethOS license agreement to ethOS download page.

I highly recommend the 1.3.0 upgrade (or upgrading miners independently). They upgraded all the miners; I'm getting +1-2/card on ethminer since the update...

Great. Thanks for that. I would give it a try when it goes to throttle next time. It was puzzling me for a while and I’m glad that I asked here thanks again.

Try "clear-thermals" it should reset everything.

Temperature control continues to be my biggest headache right now and few of my GPU goes into throttle mode within few hours. Is there a way I can reset the GPU from throttle mode to normal mode without restarting the rig and miner ?

It is not a good idea to get the OS from unreliable source. You can start mining for different person without any warning.

where did you find it
I really want to test it before buy.

Yes, it seems installation issue. The downgraded OS works just fine with same configurations. I also updated my other rig into the latest version and that also works fine and only individual GPUs goes into throttle mode. Thanks @greyday for your help. Cheers