@Genoil It didn't crash. It looked like it just paused itself. The screen showed the the usual mining on work info, received new job etc.. Checked the stats on miningpoolhub and confirm the worker is not running. Still manage to ctrl-c to close. Restarted and currently running. Will let you know if it happens again, will run it overnight. It is weird cause everything was running smoothly.
Just tested 1.1 It ran for about 10-15 minutes, after that, the GPU cooled down and fans stopped (idle temp), but the miner kept running as if it is was still mining?
m 21:01:01|main Mining on PoWhash #cebd65b7 : 117.21MH/s [A84+0:R0+0:F0] i 21:01:03|gpuminer2 Solution found; Submitting to europe1.ethereum.miningpoolhub.com ... i 21:01:03|gpuminer2 Nonce: 0x5e4cb10702f0c9fa i 21:01:03|stratum B-) Submitted and accepted. m 21:01:03|main Mining on PoWhash #cebd65b7 : 117.21MH/s [A85+0:R0+0:F0] m 21:01:05|main Mining on PoWhash #cebd65b7 : 108.83MH/s [A85+0:R0+0:F0] m 21:01:07|main Mining on PoWhash #cebd65b7 : 113.02MH/s [A85+0:R0+0:F0] m 21:01:09|main Mining on PoWhash #cebd65b7 : 113.02MH/s [A85+0:R0+0:F0] i 21:01:10|stratum Received new job #6acc i 21:01:10|gpuminer0 set work to: #7594f752, target #00000002ef59a5da i 21:01:10|gpuminer1 set work to: #7594f752, target #00000002ef59a5da i 21:01:10|gpuminer2 set work to: #7594f752, target #00000002ef59a5da i 21:01:11|gpuminer3 set work to: #7594f752, target #00000002ef59a5da m 21:01:11|main Mining on PoWhash #7594f752 : 39.38MH/s [A85+0:R0+0:F0] m 21:01:13|main Mining on PoWhash #7594f752 : 113.02MH/s [A85+0:R0+0:F0] i 21:01:13|stratum Received new job #6acd i 21:01:13|gpuminer0 set work to: #578f3dc1, target #00000002ef59a5da -- still 'mining' and submitting shares that don't appear on the pool, hashrate is 0 --
Also on the pool shares didn't go up and hashrate dropped to 0. After restarting multiple times it ran for about 5/10 minutes, fans cool down and stop, and miner keeps 'mining' or freezes...
i 21:37:58|stratum Received new job #6b68 i 21:37:58|gpuminer0 set work to: #3fd535cc, target #000000022e85e0cc i 21:37:59|gpuminer1 set work to: #3fd535cc, target #000000022e85e0cc i 21:37:59|gpuminer2 set work to: #3fd535cc, target #000000022e85e0cc i 21:37:59|gpuminer3 set work to: #3fd535cc, target #000000022e85e0cc i 21:37:59|stratum Received new job #6b69 i 21:37:59|gpuminer0 set work to: #e82724b6, target #000000022e85e0cc -- miner was frozen --
Ah, GPU's on my test rig running 110 stopped ten minutes ago. No messages. Ctrl-c managed to exit. Re-started and mining again. See how it goes next 3 hours.
Never worked with Nvidia. What would i need to download for upcoming GTX 1080/1070? Will it work on Ubuntu 14.04 or should I use Win 7 instead? want to know before deciding whether to get in line on May 27 or not,
PS: nevermind, found instructions for Ubuntu on the prior page, thanks. Not sure what it means that I can find executable in ethminer's folder. Does it mean to cd into that folder, then do ./ethminer....?
same here, halt after a few hours. i think @bitcanuck is right, another deadlock. there was a mutex that was disabled in 1.0.8 that i carelessly enabled again. will test a bit more before releasing new binaries.
@ethfan I disabled the mutex just now, so if you update the source and rebuild, the issue may be resolved. Not 100% sure, have only tested it for 150 minutes now. But I can't think of anything else if 1.0.8 was working well for you.
also updated the binary. the more testers, the faster the bug will surface if it;s still there.
I also had the bad file descriptor error several times in the last two days shortly after installing 1.0.8 on both my rigs.
I followed the fix //m_socket.close() this morning. I'm going to let them run with it to see if the problem pops up again. That could take sometime considering the last crash took almost 24 hours to reappear.
Hello guys and girls, I'm hoping someone can help me with an otherwise hopeless-seeming situation. I have Win7 pro running on an ASrock H81 pro-btc with 6x gtx 750-ti (gigabyte 2MB series). Corsair 760 powers this rig. I cannot for the life of me get my rig to mine ethereum, even when using only 1 gpu. I've tried every version of ethminer out there, over the past few months of releases. I've tried installing the older version of Nvidia drivers that everyone raves about (364.51?) - Same results as with new drivers. I've tried installing the Nvidia Dev toolkit - same result. I tried installing the C++ redistributable libraries - Same result. Tried under clocking in Afterburner - same result. Tried reinstalling windows 7 after format - same result. Tried running with no afterburner installed and no overclock - same result
I have not tried purchasing HDMI cables to connect to each GPU, could this be the issue? Or the DVI terminators - are those possibly necessary?
Just as an aside I'm able to stably run X11, QUARK, and others on this rig with ccminer, very nice and stable.
I keep getting ClEnqueueBuffer(-5) or clEnqueueBuffer(-4) and then a driver crash. This is AFTER successful DAG creation and it seems as if the miner is just about to start mining and poof. This is when I use -U in the bat file. When I use -G I get a BSOD with a nvm**.sys error dump. I'm trying to see if there is a solution for Win 7 before I dual boot into UBU 16.04 and try to start from scratch there.
I know win7 doesnt like 6 GPU's but i even tried running the batch file with --cuda-devices 0 - so only 1 GPU.
I ask, please, as this becomes a very sad situation for me considering the exponential growth ETH is experiencing.
Has anyone else had/solved this issue?
Many kind thanks in advance, Michael US Mining Rentals
So what's the consensus? Are 750Ti with 2gb RAM capable anywhere, on any OS at all? Also, thanks for your quick responses, Genoil - I've tried -U and -G to no avail. Will try again with HDMI cables in a few days. However I still can't tell if I'm barking up the wrong tree altogether with these GPUs
with a single 750ti and 1.1 I keep getting this set_constants error
C:\ethminer>ethminer -U -S coinotron.com:3344 -O xxxx.xxxxx -t 1 --farm-recheck 200 --cuda-block-size 128 --cuda-schedule auto Genoil's ethminer 0.9.41-genoil-1.1 ===================================================================== Forked from github.com/ethereum/cpp-ethereum CUDA kernel ported from Tim Hughes' OpenCL kernel With contributions from nerdralph, RoBiK, tpruvot and sp_
Please consider a donation to: ETH: 0xeb9310b185455f863f526dab3d245809f6854b4d
Using grid size 8192, block size 128
i 13:44:35|main Connecting to stratum server coinotron.com:3344 i 13:44:35|stratum Connected to stratum server coinotron.com : 3344 i 13:44:35|stratum Starting farm m 13:44:35|main Waiting for work package... i 13:44:35|stratum Subscribed to stratum server i 13:44:35|stratum Received new job #141920 i 13:44:35|cudaminer0 set work; seed: #cad4f4a9, target: #00000001021c6b81 i 13:44:35|cudaminer0 Initialising miner... m 13:44:35|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0] m 13:44:36|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0] i 13:44:36|stratum Authorized worker nitefire.miner002 m 13:44:36|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0] m 13:44:36|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0] m 13:44:36|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0] m 13:44:36|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0] m 13:44:37|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0] m 13:44:37|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0] Using device: GeForce GTX 750 Ti (Compute 5.0) m 13:44:37|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0] Cuda error in func 'set_constants' at line 128 : invalid device symbol.
@USMiningRentals don't waste your time on Windows for now. There's a tiny spark of hope that a new Win10 update scheduled for July may solve the issues with Nvidia cards on Windows, but it could also just be for the new 10x0 cards.
@RyanD you can use the 1.0.8 ones. When i release the final 1.1, I will include 750Ti support, but honestly that's only to avoid these 'set_constants' questions . Because.....it......will.....crawl.....or.....crash....anyway....
@USMiningRentals if you really want to see it working, you could start witha very low grid size, i.e. --cuda-grid-size 256 and gently go up (multiply by 2 each time... so 512, 1024) until you system becomes unresponsive. Then still you only get like 0.5MH/s tops, while it used to do 8.5MH/s when Ethereum was launched.
Thanks Genoil, I will definitely give the tuning a shot. espescially in ubu. (Possibly) unrelated question: Do I have to have geth running and synced during dag gen/mining?
I am typing this on my main desktop (from which I monitor all my rigs). It has a solitary 750Ti driving dual displays AND it is mining with 108 (with -G; some issues with -U). It is getting just under 4Mh/s. Barely worth the extra electricity to run it.
Comments
Will let you know if it happens again, will run it overnight. It is weird cause everything was running smoothly.
Just tested 1.1 It ran for about 10-15 minutes, after that, the GPU cooled down and fans stopped (idle temp), but the miner kept running as if it is was still mining? Also on the pool shares didn't go up and hashrate dropped to 0.
After restarting multiple times it ran for about 5/10 minutes, fans cool down and stop, and miner keeps 'mining' or freezes... When I run 1.0.8 again no problems at all...
* OS: Windows 10 x64
* GPU's: R9 390
* Driver: Catalyst 15.12
* Pool: MiningPoolHub
What would i need to download for upcoming GTX 1080/1070?Will it work on Ubuntu 14.04 or should I use Win 7 instead?want to know before deciding whether to get in line on May 27 or not,PS: nevermind, found instructions for Ubuntu on the prior page, thanks.
Not sure what it means that I can find executable in ethminer's folder.
Does it mean to cd into that folder, then do ./ethminer....?
I ran seperate instances with 1.1 for each card.
After 3 hours 1 instance was frozen and 3 where still working after 5 hours...
Back to 1.0.8 for now
also updated the binary. the more testers, the faster the bug will surface if it;s still there.
I also had the bad file descriptor error several times in the last two days shortly after installing 1.0.8 on both my rigs.
I followed the fix //m_socket.close() this morning.
I'm going to let them run with it to see if the problem pops up again. That could take sometime considering the last crash took almost 24 hours to reappear.
Thanks for the help!
I have Win7 pro running on an ASrock H81 pro-btc with 6x gtx 750-ti (gigabyte 2MB series). Corsair 760 powers this rig.
I cannot for the life of me get my rig to mine ethereum, even when using only 1 gpu. I've tried every version of ethminer out there, over the past few months of releases.
I've tried installing the older version of Nvidia drivers that everyone raves about (364.51?) - Same results as with new drivers.
I've tried installing the Nvidia Dev toolkit - same result.
I tried installing the C++ redistributable libraries - Same result.
Tried under clocking in Afterburner - same result.
Tried reinstalling windows 7 after format - same result.
Tried running with no afterburner installed and no overclock - same result
I have not tried purchasing HDMI cables to connect to each GPU, could this be the issue?
Or the DVI terminators - are those possibly necessary?
Just as an aside I'm able to stably run X11, QUARK, and others on this rig with ccminer, very nice and stable.
I keep getting ClEnqueueBuffer(-5) or clEnqueueBuffer(-4) and then a driver crash.
This is AFTER successful DAG creation and it seems as if the miner is just about to start mining and poof.
This is when I use -U in the bat file.
When I use -G I get a BSOD with a nvm**.sys error dump.
I'm trying to see if there is a solution for Win 7 before I dual boot into UBU 16.04 and try to start from scratch there.
I know win7 doesnt like 6 GPU's but i even tried running the batch file with --cuda-devices 0 - so only 1 GPU.
I ask, please, as this becomes a very sad situation for me considering the exponential growth ETH is experiencing.
Has anyone else had/solved this issue?
Many kind thanks in advance,
Michael
US Mining Rentals
AFAIK gtx750 are not capable of mining ethereum anymore
Oh and replace -G with -U when using my miner for added performance with Nvidia cards.
Also, thanks for your quick responses,
Genoil - I've tried -U and -G to no avail.
Will try again with HDMI cables in a few days.
However I still can't tell if I'm barking up the wrong tree altogether with these GPUs
C:\ethminer>ethminer -U -S coinotron.com:3344 -O xxxx.xxxxx -t 1
--farm-recheck 200 --cuda-block-size 128 --cuda-schedule auto
Genoil's ethminer 0.9.41-genoil-1.1
=====================================================================
Forked from github.com/ethereum/cpp-ethereum
CUDA kernel ported from Tim Hughes' OpenCL kernel
With contributions from nerdralph, RoBiK, tpruvot and sp_
Please consider a donation to:
ETH: 0xeb9310b185455f863f526dab3d245809f6854b4d
Using grid size 8192, block size 128
i 13:44:35|main Connecting to stratum server coinotron.com:3344
i 13:44:35|stratum Connected to stratum server coinotron.com : 3344
i 13:44:35|stratum Starting farm
m 13:44:35|main Waiting for work package...
i 13:44:35|stratum Subscribed to stratum server
i 13:44:35|stratum Received new job #141920
i 13:44:35|cudaminer0 set work; seed: #cad4f4a9, target: #00000001021c6b81
i 13:44:35|cudaminer0 Initialising miner...
m 13:44:35|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0]
m 13:44:36|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0]
i 13:44:36|stratum Authorized worker nitefire.miner002
m 13:44:36|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0]
m 13:44:36|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0]
m 13:44:36|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0]
m 13:44:36|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0]
m 13:44:37|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0]
m 13:44:37|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0]
Using device: GeForce GTX 750 Ti (Compute 5.0)
m 13:44:37|main Mining on PoWhash #9004d0cd : 0.00MH/s [A0+0:R0+0:F0]
Cuda error in func 'set_constants' at line 128 : invalid device symbol.
C:\ethminer>
thoughts?
@USMiningRentals if you really want to see it working, you could start witha very low grid size, i.e. --cuda-grid-size 256 and gently go up (multiply by 2 each time... so 512, 1024) until you system becomes unresponsive. Then still you only get like 0.5MH/s tops, while it used to do 8.5MH/s when Ethereum was launched.
(Possibly) unrelated question: Do I have to have geth running and synced during dag gen/mining?
Thanks, just playing around with it and I have 750ti in this desktop