SRBMiner Cryptonight AMD GPU Miner V1.9.3 - native algo switching

[doktor83]

--logfile log.txt does not seem to do anything if set in the start.bat command line when using auto-switching now. None of my rigs have created a file it seems. Is that something that can be fixed?

Don't know what is happening there because it created log when i was doing the development.
Just put the --logfile log-switching.txt in start.bat and it is used on every algo switch.

[dopper]

Dok, heeelp!
I still can't get the miner to restart windows when one card goes off (0 h/s). Miner restarts itself and freezes on initialising kernels.

start.bat

start SRBMiner-CN.exe --config configv7.txt --pools poolsocn.txt --usealgoswitching --sendallstales

config

{
"cryptonight_type" : "normalv7",
"intensity" : 112,
"worksize" : 16,
"double_threads" : true,
"min_rig_speed" : 3800,
"min_rig_speed_duration" : 30,
"retry_time" : 1,
"reboot_script" : "reboot-windows.bat",
}

reboot-windows.bat

shutdown.exe /r /t 3

algos

{
"algos" :
[
   {
      "algo" : "normalv7",
      "config" : "configv7.txt",
      "pools" : "poolsocn.txt",
      "hashrate" : 3850,
      "startup_script": "restart_gpu.bat"
   },
   {
      "algo" : "heavy",
      "config" : "confighv.txt",
      "pools" : "poolsocnhv.txt",
      "hashrate" : 3100,
      "startup_script": "restart_gpu.bat"
   },
   {
      "algo" : "haven",
      "config" : "confighav.txt",
      "pools" : "poolsocnhv.txt",
      "hashrate" : 3100,
      "startup_script": "restart_gpu.bat"
   },
   {
      "algo" : "bittubev2",
      "config" : "configbt.txt",
      "pools" : "poolsocnhv.txt",
      "hashrate" : 3100,
      "startup_script": "restart_gpu.bat"
   }   
]
}

How it works? When miner decides to restart itself and when to restart windows? I don't get it

[dragonmike]

You need to add the following line to each and all your config files (If you use multi-algo-switching, what you have in the main config file is not read):

"reboot_script" : "reboot-windows.bat",

(or whatever your reboot file is called).


So for example if you have two config files...

config-normalv7.txt
config-heavy.txt

...you will need to add the line above at the bottom of each.

[zeef]

Hi doktor,

I was using version 1.6.3 in my vega 56,

using the 1.6.6 i got less 30 hs

Using same settings,

On other rigs with rx 580 and vega 64 got the same hashes.

Just reporting  

Tell him which driver you using.

and which algo is it  

Hi doc,

Using the recommended 1.6.1,

The algo is Heavy, mining LOKI. ( 1 vega 56 mining solo)

I have other strange issue in other rig (9 vegas 64) in fast algo mining masari. When i start miner 1 gpu (Vega 64) stills on 36xx hs and the others 8 in 40xx hs. Is i restart is random the gpu that will hash less that others.  (But this behavior is not always, if i restart many times, one time will be ok , all performing the same.. very strange)

Thanks

[doktor83]

Hi doktor,

I was using version 1.6.3 in my vega 56,

using the 1.6.6 i got less 30 hs

Using same settings,

On other rigs with rx 580 and vega 64 got the same hashes.

Just reporting  

Tell him which driver you using.

and which algo is it  

Hi doc,

Using the recommended 1.6.1,

The algo is Heavy, mining LOKI. ( 1 vega 56 mining solo)

I have other strange issue in other rig (9 vegas 64) in fast algo mining masari. When i start miner 1 gpu (Vega 64) stills on 36xx hs and the others 8 in 40xx hs. Is i restart is random the gpu that will hash less that others.  (But this behavior is not always, if i restart many times, one time will be ok , all performing the same.. very strange)

Thanks

wait now, did you want to write 18.6.1 ? Or miner version 1.6.1

[zeef]

Hi doktor,

I was using version 1.6.3 in my vega 56,

using the 1.6.6 i got less 30 hs

Using same settings,

On other rigs with rx 580 and vega 64 got the same hashes.

Just reporting  

Tell him which driver you using.

and which algo is it  

Hi doc,

Using the recommended 1.6.1,

The algo is Heavy, mining LOKI. ( 1 vega 56 mining solo)

I have other strange issue in other rig (9 vegas 64) in fast algo mining masari. When i start miner 1 gpu (Vega 64) stills on 36xx hs and the others 8 in 40xx hs. Is i restart is random the gpu that will hash less that others.  (But this behavior is not always, if i restart many times, one time will be ok , all performing the same.. very strange)

Thanks

wait now, did you want to write 18.6.1 ? Or miner version 1.6.1

Hi

The drivers is 18.6.1.

The miner hash works good on 18.6.3 and 18.6.4

On solo mining works good the 18.6.6 miner version and with better hashrates, but on rig no..


Thanks

[dopper]

You need to add the following line to each and all your config files (If you use multi-algo-switching, what you have in the main config file is not read):

"reboot_script" : "reboot-windows.bat",

I have this exact line at the bottom in all 4 config files. Still no luck.

{
"cryptonight_type" : "heavy",
"intensity" : 56,
"worksize" : 8,
"double_threads" : true,
"min_rig_speed" : 3000,
"min_rig_speed_duration" : 30,
"retry_time" : 1,
"reboot_script" : "reboot-windows.bat",
}

and so on..

[christiano88]

Something is wrong, when trying to force haven (only adding this algo to algos.txt) it says it cannot login the user to the pool (using monero ocean + enablealgoswitching parameter). With bittube and v7 it works fine. I see haven listed on their website, so why is it not working?

[dragonmike]

You need to add the following line to each and all your config files (If you use multi-algo-switching, what you have in the main config file is not read):

"reboot_script" : "reboot-windows.bat",

I have this exact line at the bottom in all 4 config files. Still no luck.

{
"cryptonight_type" : "heavy",
"intensity" : 56,
"worksize" : 8,
"double_threads" : true,
"min_rig_speed" : 3000,
"min_rig_speed_duration" : 30,
"retry_time" : 1,
"reboot_script" : "reboot-windows.bat",
}

and so on..

So what behaviour are you currently experiencing? Is the miner crashing and just restarts instead of rebooting?

[ikyo]

You need to add the following line to each and all your config files (If you use multi-algo-switching, what you have in the main config file is not read):

"reboot_script" : "reboot-windows.bat",

I have this exact line at the bottom in all 4 config files. Still no luck.

{
"cryptonight_type" : "heavy",
"intensity" : 56,
"worksize" : 8,
"double_threads" : true,
"min_rig_speed" : 3000,
"min_rig_speed_duration" : 30,
"retry_time" : 1,
"reboot_script" : "reboot-windows.bat",
}

and so on..

So what behaviour are you currently experiencing? Is the miner crashing and just restarts instead of rebooting?

I think what you are seeing is the miner restarting because the rig speed is falling before the minimum that you have set.  So I would create a startup script and put in the --resetvega so it will fully reset the cards. 

[dopper]

So what behaviour are you currently experiencing? Is the miner crashing and just restarts instead of rebooting?

When both cards work normal and hashrate drops below set parameter for 30 seconds, miner restarts and runs a startup script (to reset cards and apply overdriven settings). When one card goes off (0 h/s), I want computer to restart. Instead, after 30 sec of hashrate drop miner restarts itself and stops at "Loading kernels", because one card halted and needs to restart (i use switch-radeon-gpu instead of devcon).

Thats it. Windows running (i have hardware watchdog to restart it if pc freezes), srbminer process running (the same watchdog monitors it), pc hangs only after i try to end sbrminer task in taskmanager

[dopper]

I think what you are seeing is the miner restarting because the rig speed is falling before the minimum that you have set.  So I would create a startup script and put in the --resetvega so it will fully reset the cards. 

hm.. it should help, thank you.

so reboot_script to work i need to remove min_speed parameter?

[4ward]

Try Forager - Multi-algo switching miner for AMD/NVIDIA/CPU

Continuously developed, with largest selection and latest versions of miners and algos.
Highly configurable, Integrated Afterburner support for power usage and profit calculation

SRBMiner support included

[DeshaCAM]

Za doktora. Ograničen sam na dve poruke dnevno, pa moram ovde da napišem.
Ne prijavljije grešku, jednostvano se posle par sek ugasi majner, ne dođe ni do momenta učitavanja kernela ...

Ovaj deo u bat
SRBMiner-CN.exe --sendallstales --config config-normalv7.txt --pools pools-algoswitch.txt --usealgoswitching

... jel treba da stoji i --config config-heavy.txt    Jer je i heavy algo inicijalizovan u algos.txt

i da li za ovaj deo u bat  --pools pools-algoswitch.txt  treba da imam i takav txt fajl u folderu majnera koji bi se zvao pools-algoswitch.txt imam samo algos.txt

[jasonyslai]

Hi, Thanks for this awesome miner!
But two of rigs throws the following issue today and restarted :
"The Interactive Services Detection service terminated with the following error: Incorrect function."

- <EventData>
  <Data Name="param1">Interactive Services Detection</Data>
  <Data Name="param2">%%1</Data>
  <Binary>5500490030004400650074006500630074000000</Binary>
  </EventData>
  </Event>

Is that any solution for this issue =)?

[doktor83]

Something is wrong, when trying to force haven (only adding this algo to algos.txt) it says it cannot login the user to the pool (using monero ocean + enablealgoswitching parameter). With bittube and v7 it works fine. I see haven listed on their website, so why is it not working?

you are using your Monero wallet for login, not haven, right ?

[christiano88]

Something is wrong, when trying to force haven (only adding this algo to algos.txt) it says it cannot login the user to the pool (using monero ocean + enablealgoswitching parameter). With bittube and v7 it works fine. I see haven listed on their website, so why is it not working?

you are using your Monero wallet for login, not haven, right ?

Exactly. Since the payment are done in XMR, I figured it would be the right thing to do. Also, for bittube it works just fine. Also wrote to moneroocean team on discord and you replied, but that solution is already applied from the beginning.

L.E. Nevermind, it works now, it seems that it must first start in normalv7, then pool will request the most profitable coin, and it was able to switch to haven.

[doktor83]

Something is wrong, when trying to force haven (only adding this algo to algos.txt) it says it cannot login the user to the pool (using monero ocean + enablealgoswitching parameter). With bittube and v7 it works fine. I see haven listed on their website, so why is it not working?

you are using your Monero wallet for login, not haven, right ?

Exactly. Since the payment are done in XMR, I figured it would be the right thing to do. Also, for bittube it works just fine. Also wrote to moneroocean team on discord and you replied, but that solution is already applied from the beginning.

L.E. Nevermind, it works now, it seems that it must first start in normalv7, then pool will request the most profitable coin, and it was able to switch to haven.

You can start it with any algo, haven, stellite doesn't matter, it will switch to the algo the pool requested when connected.

[abg00]

Hi Dok
Can i use gpu0123 with normalv7 and gpu456 with heavy? need run two copy of miner? or add in config two section for different algo? use 1.6.6 now

[doktor83]

Hi Dok
Can i use gpu0123 with normalv7 and gpu456 with heavy? need run two copy of miner? or add in config two section for different algo? use 1.6.6 now

If you want to do algo switching then you must do it in config files, if you want to mine 'normally' you can do it either way.

SRBMiner-CN.exe --ccryptonighttype normalv7 --cgpuid 0,1,2,3 --cgpuintensity 0,0,0,0 --cgputhreads 2,2,2,2 --cpool POOL --cwallet WALLET --cpassword x --sendallstales

SRBMiner-CN.exe --ccryptonighttype heavy --cgpuid 4,5,6 --cgpuintensity 0,0,0 --cgputhreads 2,2,2 --cpool POOL --cwallet WALLET --cpassword x --sendallstales