[NemosMiner] multi algo profit switching NVIDIA/CPU miner

[millsys]

hi nemosminers,

Few days ago I reported about my 6 x gtx 1070 rig crashing and not being stable. I then started with adding and testing with 1 gpu, 2 gpu's, 3 gpus and found that as soon as I add a 4th gpu CPU usage increases a lot and having lower hash-rates then with 3 gpus. During benchmarking it will crash. It doesn't matter which of the 3 remaining gpus I add on any remaining pci slot. Result seems same. I'm using a g3900 CPU and 4gb ram. Could this be the bottleneck?

At home i'm running the same motherboard (msi z270 sli plus) with 6 1080ti's with G4400 cpu and 8gb ram now running fine with currently 5 active gpu's. I actually found that on that motherboard 1 pci slot is corrupt. Will test with m2 adapter tonight to see if it will run with 6

If your cpu usage is going up, check event viewer system log for PCI-E communication errors, could be a bad riser

Hey millsys, thank you for your reply. I have tested with 3 different gpu's having each their own riser. Chance that all 3 are bad is small but will have a look at it

Had same issue, ALL RISERS WERE BAD!!!  Seriously, I've RMAd over 18 risers already.   I've had rigs seem stable for 6 cards, then move those same 6 cards into a 12 card setup and the problem rears its ugly head because you now have 12 cards screaming PCI-E errors instead of just 6 and the system starts to grind to a halt.  Replaced risers with new $6 ebay ones and problem disappeared immediately (cpu usage, event viewer errors, stability).  Ebay auction 302483050496 is the only ones I trust now, have ordered over 100 from this source and they have been rock solid.  Cheap too, saved a few bucks sending back overpriced/bad ones to Amazon/NewEgg after replacement.

[1714600101]

1714600101

[1714600101]

1714600101

[1714600101]

1714600101

[1714600101]

1714600101

[cryptomoblo]

where to look exactly in the event viewer logs to find the PCI-E errors?

[jebidia]

I Have this issue with zpool 24hr...i downloaded it again to make sure it wasn't something i had messed up... still didn't work.   Ahash 24hr works flawlessly.   2.3cbeta

That's not the same problem.
That's an issue on zpool side, not NemosMiner - unfortunately.
Zpool's API has a limit on the number of requests per IP, so there's no solution for now.

use a vpn (for zpool) works for me...

Interesting,   i had actually considered a vpn on my rig before but worried it added latency.   Is this not an issue?

Thanks and I'll give it a shot today.

[bluedl]

Back on the nemosminer again. Using ahashpool 24 hr, running pretty good so far.
Just wanted to double check, my 1070's and 1080ti's tend to enjoy phi and skein. Is everybody else getting the same? I get some discrepancies across rigs. I have some 1070's that liked phi and some that liked skein, wondering if perhaps the benchmarking isn't that accurate?

[dragonmike]

hi nemosminers,

Few days ago I reported about my 6 x gtx 1070 rig crashing and not being stable. I then started with adding and testing with 1 gpu, 2 gpu's, 3 gpus and found that as soon as I add a 4th gpu CPU usage increases a lot and having lower hash-rates then with 3 gpus. During benchmarking it will crash. It doesn't matter which of the 3 remaining gpus I add on any remaining pci slot. Result seems same. I'm using a g3900 CPU and 4gb ram. Could this be the bottleneck?

At home i'm running the same motherboard (msi z270 sli plus) with 6 1080ti's with G4400 cpu and 8gb ram now running fine with currently 5 active gpu's. I actually found that on that motherboard 1 pci slot is corrupt. Will test with m2 adapter tonight to see if it will run with 6

If your cpu usage is going up, check event viewer system log for PCI-E communication errors, could be a bad riser

Where would I find those errors?
Cannot see anything under the Windows/System log in event viewer... And I doubt my risers are 100% reliable.

[millsys]

hi nemosminers,

Few days ago I reported about my 6 x gtx 1070 rig crashing and not being stable. I then started with adding and testing with 1 gpu, 2 gpu's, 3 gpus and found that as soon as I add a 4th gpu CPU usage increases a lot and having lower hash-rates then with 3 gpus. During benchmarking it will crash. It doesn't matter which of the 3 remaining gpus I add on any remaining pci slot. Result seems same. I'm using a g3900 CPU and 4gb ram. Could this be the bottleneck?

At home i'm running the same motherboard (msi z270 sli plus) with 6 1080ti's with G4400 cpu and 8gb ram now running fine with currently 5 active gpu's. I actually found that on that motherboard 1 pci slot is corrupt. Will test with m2 adapter tonight to see if it will run with 6

If your cpu usage is going up, check event viewer system log for PCI-E communication errors, could be a bad riser

Where would I find those errors?
Cannot see anything under the Windows/System log in event viewer... And I doubt my risers are 100% reliable.

This is what Event Viewer looks like on all rigs I've built with bad risers

https://imgur.com/a/ed0f3

[cryptomoblo]

Thanks for creating that image! For me not seeing this kind of errors, so would suggest that its not riser related in my case. Tomorrow extra bank of ram comes and different cpu. If it wont solve my issue will send it back, will keep you guys keeps updated once if've installed and been running stable (or not).

[starrbuck]

Had same issue, ALL RISERS WERE BAD!!!  Seriously, I've RMAd over 18 risers already.

Wow, that sux!  The first seven risers I have tried have all been good.  I guess I am lucky!

[X299]

I don't see in Nemosminer this algo "Ethash" (Ethereum coin), for miningpoolhub.com
Is it possible to see it in the future in the next version Nemosminer 2.3d or 2.4?

[jebidia]

I Have this issue with zpool 24hr...i downloaded it again to make sure it wasn't something i had messed up... still didn't work.   Ahash 24hr works flawlessly.   2.3cbeta

That's not the same problem.
That's an issue on zpool side, not NemosMiner - unfortunately.
Zpool's API has a limit on the number of requests per IP, so there's no solution for now.

use a vpn (for zpool) works for me...

Interesting,   i had actually considered a vpn on my rig before but worried it added latency.   Is this not an issue?

Thanks and I'll give it a shot today.

Actually, just gave this a shot and it still gives an error trying to connect with the Zpool 24hr nemo file...I use PIA VPN

[knarkers]

Hi!
I got a small rig running with 4x1060 3GB and a single 1080ti and I'm having trouble with the benchmark.
The 4x1060 is running on one instance of nemos and the 1080ti on another.
However the benchmark results seem to get mixed up. Isn't it possible to run multiple instances?

Cheers!

[kalud]

Hi!
I got a small rig running with 4x1060 3GB and a single 1080ti and I'm having trouble with the benchmark.
The 4x1060 is running on one instance of nemos and the 1080ti on another.
However the benchmark results seem to get mixed up. Isn't it possible to run multiple instances?

Yes, use the -SelGPUDSTM and -SelGPUCC to select the right GPUs for each instances and I would use two different folders for the installed nemosminer instances. There is a stats folder and both instances overwrites into it. Using two different nemosminer install folders would solves that.

[primo1983]

Where do I see which is the best Algorithm to be executed in the day, what else have gains? In case I use ahaspoool ?
-Algorithm equihash,tribus,Lyra2RE2

[knarkers]

Hi!
I got a small rig running with 4x1060 3GB and a single 1080ti and I'm having trouble with the benchmark.
The 4x1060 is running on one instance of nemos and the 1080ti on another.
However the benchmark results seem to get mixed up. Isn't it possible to run multiple instances?

Yes, use the -SelGPUDSTM and -SelGPUCC to select the right GPUs for each instances and I would use two different folders for the installed nemosminer instances. There is a stats folder and both instances overwrites into it. Using two different nemosminer install folders would solves that.

I've done exactly that.
Selected the gpu's and running from separate folders. So not exactly sure why it isn't working...

[pttrade]

I Have this issue with zpool 24hr...i downloaded it again to make sure it wasn't something i had messed up... still didn't work.   Ahash 24hr works flawlessly.   2.3cbeta

That's not the same problem.
That's an issue on zpool side, not NemosMiner - unfortunately.
Zpool's API has a limit on the number of requests per IP, so there's no solution for now.

use a vpn (for zpool) works for me...

Interesting,   i had actually considered a vpn on my rig before but worried it added latency.   Is this not an issue?

Thanks and I'll give it a shot today.

I made "special" proxy to prevent zpool cap on calls by ip address. But it needs to change the pool script a little. Let me know if you are interested on trying.

[airwalker_]

Hi!
I got a small rig running with 4x1060 3GB and a single 1080ti and I'm having trouble with the benchmark.
The 4x1060 is running on one instance of nemos and the 1080ti on another.
However the benchmark results seem to get mixed up. Isn't it possible to run multiple instances?

Yes, use the -SelGPUDSTM and -SelGPUCC to select the right GPUs for each instances and I would use two different folders for the installed nemosminer instances. There is a stats folder and both instances overwrites into it. Using two different nemosminer install folders would solves that.

I've done exactly that.
Selected the gpu's and running from separate folders. So not exactly sure why it isn't working...

If you are running multiple instances you need to benchmark them separately. When done, you need to set the miner stats for each folder to read-only. Just the miner stats, don't touch the pool stats.

[knarkers]

Hi!
I got a small rig running with 4x1060 3GB and a single 1080ti and I'm having trouble with the benchmark.
The 4x1060 is running on one instance of nemos and the 1080ti on another.
However the benchmark results seem to get mixed up. Isn't it possible to run multiple instances?

Yes, use the -SelGPUDSTM and -SelGPUCC to select the right GPUs for each instances and I would use two different folders for the installed nemosminer instances. There is a stats folder and both instances overwrites into it. Using two different nemosminer install folders would solves that.

I've done exactly that.
Selected the gpu's and running from separate folders. So not exactly sure why it isn't working...

If you are running multiple instances you need to benchmark them separately. When done, you need to set the miner stats for each folder to read-only. Just the miner stats, don't touch the pool stats.

Thanks! I'll try that

[minerx117]

NemosMiner-v2.4

zpool  hashrefinery  ahashpool  miningpoolhub  nicehash

https://github.com/nemosminer/NemosMiner-v2.4/releases/download/v2.4/NemosMiner-v2.4.zip

Added:

XMRig (cryptonight)

mBTC/perday along side BTC/perday

-workername (nicehash)

Lyra2z --submit-stale (miningpoolhub)

updated: ccminers (Tpruvot & KlausT)

Increased Benchmark time 8min per Algo (takes awhile but very Accurate)

Added: -ActiveMinerGainPct (percent of advantage that active miner has over candidates in term of profit (default 3%) this will help reduce unnecessary switching

bug fix: for 8 card Rigs (not closing miner)

Tested with 6x10603gb & 6x1080ti

[HardBlockMiner]

NemosMiner-v2.4

zpool  hashrefinery  ahashpool  miningpoolhub  nicehash

https://github.com/nemosminer/NemosMiner-v2.4/releases/download/v2.4/NemosMiner-v2.4.zip

Added:

XMRig (cryptonight)

mBTC/perday along side BTC/perday

-workername (nicehash)

Lyra2z --submit-stale (miningpoolhub)

updated: ccminers (Tpruvot & KlausT)

Increased Benchmark time 8min per Algo (takes awhile but very Accurate)

Added: -ActiveMinerGainPct (percent of advantage that active miner has over candidates in term of profit (default 3%) this will help reduce unnecessary switching

bug fix: for 8 card Rigs (not closing miner)

Tested with 6x10603gb & 6x1080ti

Thanks for keeping this going.  It really makes everything so much easier.

[primo1983]

After downloading and running zpool is giving this error what is it?

[2018-01-18 22:22:57] Starting on stratum+tcp://:
[2018-01-18 22:22:57] NVAPI GPU monitoring enabled.
[2018-01-18 22:22:57] 1 miner thread started, using 'x17' algorithm.
[2018-01-18 22:22:58] Stratum connection failed: Failed to connect to  port 80: Connection refused
[2018-01-18 22:22:58] ...retry after 30 seconds
[2018-01-18 22:23:29] Stratum connection failed: Failed to connect to  port 80: Connection refused
[2018-01-18 22:23:29] ...retry after 30 seconds
[2018-01-18 22:24:00] Stratum connection failed: Failed to connect to  port 80: Connection refused