Closed Angel996 closed 5 years ago
From version 1.4.0 up excavator doesn't support command line parameters for algorithms and user anymore. You have to run excavator using config file: excavator -c path_to_config_file.json
Here is an example of config file to run equihash on first GPU: https://github.com/nicehash/excavator/blob/master/default_command_file.json
To run equihash on all 4 GPUs, your config file should look like this (if you want to run lyra2rev2, just change parameters in algorithm.add and worker.add methods from equihash to lyra2rev2):
[
{"time":0,"commands":[
{"id":1,"method":"subscribe","params":["nhmp.usa.nicehash.com:3200","btc-addr.workerName"]}
]},
{"time":2,"commands":[
{"id":1,"method":"algorithm.add","params":["equihash"]}
]},
{"time":3,"commands":[
{"id":1,"method":"worker.add","params":["equihash","0"]},
{"id":1,"method":"worker.add","params":["equihash","1"]},
{"id":1,"method":"worker.add","params":["equihash","2"]},
{"id":1,"method":"worker.add","params":["equihash","3"]}
]},
{"time":10,"loop":10,"commands":[
{"id":1,"method":"worker.print.speeds","params":[]},
{"id":1,"method":"algorithm.print.speeds","params":[]}
]}
]
Please take a look at API documentation: https://github.com/nicehash/excavator/tree/master/api#methods
Thanks for the info.
Maybe it would be a good idea to display some message to user in this case?
Whats nhmp subdomain on Nicehash? Auto-switch pool? I can't find any info on that on NH website. Thanks!
Maybe it would be a good idea to display some message to user in this case?
You're right, we'll add a cmd message in case of incorrect startup parameters.
Whats nhmp subdomain on Nicehash? Auto-switch pool?
From version 1.5.0 up we implemented our own stratum protocol - nhmp (nicehash miner protocol). Stratum servers that support nhmp protocol are located at nhmp.LOCATION.nicehash.com:3200. We're using the same port for all algorithms. It's similar to old stratum addresses which were algorithm specific (e.g equihash.LOCATION.nicehash.com:3357).
Ok, still nothing happens.
sudo excavator -c ./excavator-lyra2v2.json
Build time: 2017-11-21 16:44:49+00:00 Build number: 1 [18:44:15][0x00007f59e7235740][info] Log started [18:44:15][0x00007f59e7235740][info] core | Found CUDA device: GeForce GTX 1060 3GB [18:44:15][0x00007f59e7235740][info] core | Found CUDA device: GeForce GTX 1060 3GB [18:44:15][0x00007f59e7235740][info] core | Found CUDA device: GeForce GTX 1060 3GB [18:44:15][0x00007f59e7235740][info] core | Found CUDA device: GeForce GTX 1060 3GB [18:44:15][0x00007f59e7235740][info] core | Found CUDA device: GeForce GTX 1060 3GB [18:44:15][0x00007f59e7235740][info] core | Found CUDA device: GeForce GTX 1060 3GB [18:44:15][0x00007f59e7235740][info] api | Listening on 127.0.0.1:3456 [18:44:15][0x00007f59e7235740][info] core | Initialized! [18:44:18][0x00007f59e1862700][warning] core | Command exec: stoi
Contents of json file:
[
{"time":0,"commands":[
{"id":1,"method":"subscribe","params":["nhmp.eu.nicehash.com:3200","3P**********************************ux.RIG04"]}
]},
{"time":2,"commands":[
{"id":1,"method":"algorithm.add","params":["lyra2rev2"]}
]},
{"time":3,"commands":[
{"id":1,"method":"worker.add","params":["lyra2rev2","0"]},
{"id":1,"method":"worker.add","params":["lyra2rev2","1"]},
{"id":1,"method":"worker.add","params":["lyra2rev2","2"]},
{"id":1,"method":"worker.add","params":["lyra2rev2","3"]}
]},
{"time":10,"loop":10,"commands":[
{"id":1,"method":"worker.print.speeds","params":[]},
{"id":1,"method":"algorithm.print.speeds","params":[]}
]}
]
Hm, this should work just fine. Can you run excavator with -d 0 flag and c/p the cmd output here (excavator -c excavator-lyra2v2.json -d 0
) ?
@Angel996 I noticed you're using old version of excavator.
Build time: 2017-11-21 16:44:49+00:00 Build number: 1
Make sure you're running the 1.5.* version.
Got it working on the 6x 1060 rig, thanks!
Is the nhmp protocol intended for yet to be released auto-switch version of Linux miner for Nicehash? We'd love to have something like that on Linux. :)
Auto-switching on Linux will be supported through NiceHash OS (currently is available in beta testing).
Ubuntu 16.04 LTS. CUDA 9.2.
DL'ded the 1.5.5a today (vs. 1.3.0 used previously), here is the output:
Tried equihash and lyra2rev2. Just nothing happens. Power consumption stays low meaning GPUs aren't even getting initialized. The rig runs fine other miners (ethminer, dstm, bminer etc).