Every time I tried to tweak the performance in XMR-STAK in web panel, I got errors. And I had to return to default setting which give no errors.
Every time I tweaked as the following:
“gpu_threads_conf” :
[
// gpu: GeForce GTX 1070 Ti architecture: 61
// memory: 7971/8119 MiB
// smx: 19
{ “index” : 0,
“threads” : 60, “blocks” : 57,
“bfactor” : 0, “bsleep” : 0,
“affine_to_cpu” : false, “sync_mode” : 3,
“mem_mode” : 1,
},
// gpu: GeForce GTX 1080 architecture: 61
// memory: 7983/8119 MiB
// smx: 20
{ “index” : 1,
“threads” : 80, “blocks” : 60,
“bfactor” : 0, “bsleep” : 0,
“affine_to_cpu” : false, “sync_mode” : 3,
“mem_mode” : 1,
},
// gpu: GeForce GTX 1060 6GB architecture: 61
// memory: 5990/6078 MiB
// smx: 10
{ “index” : 2,
“threads” : 16, “blocks” : 30,
“bfactor” : 0, “bsleep” : 0,
“affine_to_cpu” : false, “sync_mode” : 3,
“mem_mode” : 1,
},],
XMR-STAK through HIVEOS returned an error.
And then I had to go it manually without the help of HiveOS.
Lots of glitches during a month…
And…
And today, I got this error all of sudden:
[2018-10-31 17:34:44] : Mining coin: cryptonight_bittube2
[2018-10-31 17:34:44] : NVIDIA: try to load library ‘xmrstak_cuda_backend_cuda10_0’
WARNING: NVIDIA cannot load backend library: libxmrstak_cuda_backend_cuda10_0.so: cannot open shared object file: No such file or directory
[2018-10-31 17:34:44] : NVIDIA: try to load library ‘xmrstak_cuda_backend_cuda9_2’
WARNING: NVIDIA cannot load backend library: libxmrstak_cuda_backend_cuda9_2.so: cannot open shared object file: No such file or directory
[2018-10-31 17:34:44] : NVIDIA: try to load library ‘xmrstak_cuda_backend’
I hoped to install only those library with minimal effort. But I can not find an answer to this.
And then I had to burn new image of HIVEOS (Bleeding Edge version accidentally) and restarted it.
However, the web panel seems disconnected to the rig for good.
And all the OC settings can be seen on booting but of no avail. -> hash rate dropped to half.
some of GPUs are gone. and so much more of problems.
I had to revert it by re-use the previous image. I didn’t erase it! Thanks God.
Now XMR-STak with old image of HIVEOS is not working.
But new image give me chaos, hodge-podge of endless glitches and problems.
How can this be? half of GPUs are gone from mining (all GPUs are detected on booting.)
hash rates of all GPUs dropped to half of previous one. OC-settings on nVIDIA and AMD GPUs have no effect. some command are not working. miner start|stop is unstable.
Web panel is turned off, showing nothing active. while half of GPUs are actually mining.
I don’t understand why Web panel knows nothing about the rig with new image.
RIG ID is identical. only fan speed on each GPU are correctly seen from Web panel.
