Ashepp Posted October 27, 2021 Share Posted October 27, 2021 @PTRFRLL thanks for the quick fix. Unfortunately I'm seeing an issue where the miner starts 24.2 and is working fine but then an update is initiated to 24.5 and the miner is broken again. I think I have the update flag set in my config. Is there a way to stop it doing the update? Quote Link to comment
PTRFRLL Posted October 27, 2021 Author Share Posted October 27, 2021 44 minutes ago, Ashepp said: @PTRFRLL thanks for the quick fix. Unfortunately I'm seeing an issue where the miner starts 24.2 and is working fine but then an update is initiated to 24.5 and the miner is broken again. I think I have the update flag set in my config. Is there a way to stop it doing the update? As far as I'm aware, setting autoupdate: false in the config file should do the trick. { ... "autoupdate": false } Quote Link to comment
S1ph3r Posted November 5, 2021 Share Posted November 5, 2021 On 3/13/2021 at 4:53 AM, darkangelus said: Thanks, i wasted two hours last night trying to get a vm passthrough to work for mining and i wish i'd thought of checking docker. Running fine on my 1650 super 4gb, hashrate of 40 at 80w. I love the idea of mining on a slim docker instead of resource hog VM. i'm about to try this out my self and if it goes really good get some risers on my unraid server. this is awesome. Quote Link to comment
Elembemedia Posted November 6, 2021 Share Posted November 6, 2021 I’m loving trex, but would like to be able to use the config in the webui. Every time I set things it just reverts back after a restart, how can I get that to work? Quote Link to comment
PTRFRLL Posted November 6, 2021 Author Share Posted November 6, 2021 3 hours ago, Elembemedia said: I’m loving trex, but would like to be able to use the config in the webui. Every time I set things it just reverts back after a restart, how can I get that to work? When you make a change in the WebUI, it should modify the config.json file so that it persists. If you make a change on the WebUI, do you see that change in the config.json file? Quote Link to comment
PTRFRLL Posted November 6, 2021 Author Share Posted November 6, 2021 I've updated some of the docker image tags. If you still use CUDA-10. Be sure to update your image tag, otherwise you won't get updates going forward: ptrfrll/nv-docker-trex:cuda10-latest Quote Link to comment
Elembemedia Posted November 6, 2021 Share Posted November 6, 2021 When I make changes in the webui and then save them they do save to the config file. When I then restart the container the settings revert in the config file. Does have all the wallet address, worker name, etc in the container attributes overwrite things? Quote Link to comment
PTRFRLL Posted November 6, 2021 Author Share Posted November 6, 2021 28 minutes ago, Elembemedia said: When I make changes in the webui and then save them they do save to the config file. When I then restart the container the settings revert in the config file. Does have all the wallet address, worker name, etc in the container attributes overwrite things? oh I see. Yes, all the variables (wallet, password, also, etc) overwrite the config file. I’ll update the image so that doesn’t happen. Quote Link to comment
Elembemedia Posted November 7, 2021 Share Posted November 7, 2021 Can I remove them in the container now? Quote Link to comment
badbadntgood Posted November 23, 2021 Share Posted November 23, 2021 Is there any way to set the GPU power usage without the GUI? I set the value to 60 and since restarting the container I'm getting an error message within the logs stating that the value must be higher than 75 but I'm unable to access the GUI to change it. I have attempted --pl 75 through the docker container CLI and the command is not recognized. Quote Link to comment
PTRFRLL Posted November 23, 2021 Author Share Posted November 23, 2021 20 minutes ago, badbadntgood said: Is there any way to set the GPU power usage without the GUI? I set the value to 60 and since restarting the container I'm getting an error message within the logs stating that the value must be higher than 75 but I'm unable to access the GUI to change it. I have attempted --pl 75 through the docker container CLI and the command is not recognized. You can run nvidia-smi -pl 75 on Unraid itself (not within the container) Quote Link to comment
badbadntgood Posted November 23, 2021 Share Posted November 23, 2021 8 minutes ago, PTRFRLL said: You can run nvidia-smi -pl 75 on Unraid itself (not within the container) I figured out another way just so it would be limited within the docker container shortly after your reply (of course). I edited the config.json file via unraid CLI to 75 instead of 60. Thank you for the assistance, nonetheless. Quote Link to comment
repomanz Posted November 27, 2021 Share Posted November 27, 2021 @PTRFRLL Hi - thanks for the trex miner container. Running into an issue hoping forum can assist. When starting container I get this in the log: have the following 2 entries assigned in the container: nvidia-smi output: Quote Link to comment
repomanz Posted November 27, 2021 Share Posted November 27, 2021 24 minutes ago, repomanz said: @PTRFRLL Hi - thanks for the trex miner container. Running into an issue hoping forum can assist. When starting container I get this in the log: have the following 2 entries assigned in the container: nvidia-smi output: fixed this after more reading: change the container repo to this: ptrfrll/nv-docker-trex:cuda11 Quote Link to comment
Rhomax Posted November 29, 2021 Share Posted November 29, 2021 Great work on this, have been using this for months with no major issues. I did an update and now I consistently get the following error... followed by Miner shutdown. 20211129 20:59:17 TREX: Can't initialize device [ID=0, GPU #0], NVML wasn't initialized I tried disabling NVML in the config file and that hasn't helped. Any ideas? Quote Link to comment
PTRFRLL Posted November 29, 2021 Author Share Posted November 29, 2021 19 minutes ago, Rhomax said: Great work on this, have been using this for months with no major issues. I did an update and now I consistently get the following error... followed by Miner shutdown. 20211129 20:59:17 TREX: Can't initialize device [ID=0, GPU #0], NVML wasn't initialized I tried disabling NVML in the config file and that hasn't helped. Any ideas? What CUDA version are you using? This error came up with a newer version of T-Rex and I had to add a manual install NVML into the CUDA 11 container. The problem is, that package has not been back-ported to a version compatible with the CUDA 10 version Quote Link to comment
Rhomax Posted November 30, 2021 Share Posted November 30, 2021 2 hours ago, PTRFRLL said: What CUDA version are you using? This error came up with a newer version of T-Rex and I had to add a manual install NVML into the CUDA 11 container. The problem is, that package has not been back-ported to a version compatible with the CUDA 10 version I'm using CUDA11 with ptrfrll/nv-docker-trex:cuda11 Quote Link to comment
PTRFRLL Posted November 30, 2021 Author Share Posted November 30, 2021 2 hours ago, Rhomax said: I'm using CUDA11 with ptrfrll/nv-docker-trex:cuda11 Try using the ptrfrll/nv-docker-trex:latest tag (or if you want to keep a consistent version use: ptrfrll/nv-docker-trex:3.5.2). I tweaked the tags a bit with the newer version. its possible you're on the old version without the NVML package Quote Link to comment
Rhomax Posted November 30, 2021 Share Posted November 30, 2021 19 minutes ago, PTRFRLL said: Try using the ptrfrll/nv-docker-trex:latest tag (or if you want to keep a consistent version use: ptrfrll/nv-docker-trex:3.5.2). I tweaked the tags a bit with the newer version. its possible you're on the old version without the NVML package When I use the latest tag I get the following error... It's interesting because previously I didn't have any tag at all and it was working fine. I assumed that this meant it was defaulting to CUDA10 which is why I added the CUDA11 tag in the first place. Can't start T-Rex, can't initialize CUDA engine, cuda exception: CUDA_ERROR_COMPAT_NOT_SUPPORTED_ON_DEVICE. Is NVIDIA driver installed? Quote Link to comment
Rhomax Posted November 30, 2021 Share Posted November 30, 2021 21 hours ago, PTRFRLL said: Try using the ptrfrll/nv-docker-trex:latest tag (or if you want to keep a consistent version use: ptrfrll/nv-docker-trex:3.5.2). I tweaked the tags a bit with the newer version. its possible you're on the old version without the NVML package Tried updating the Nvidia drivers to the 495.44 and recreated the docker several time and still have the same result. The summary is: Using no tags or tags CUDA11 or latest or 3.5.2 I get the following error: TREX: Can't initialize device [ID=0, GPU #0], NVML wasn't initialized I'm wondering if the NVML error message is not the actual root cause and it's down to something like device initialisation more generally but this is what I see on nvidia-smi which doesn't have anything obvious stand out to me. +-----------------------------------------------------------------------------+ | NVIDIA-SMI 495.44 Driver Version: 495.44 CUDA Version: 11.5 | |-------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |===============================+======================+======================| | 0 NVIDIA GeForce ... Off | 00000000:01:00.0 Off | N/A | | 53% 54C P2 45W / 200W | 131MiB / 7982MiB | 0% Default | | | | N/A | +-------------------------------+----------------------+----------------------+ +-----------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=============================================================================| | 0 N/A N/A 1646 C /trex/t-rex 129MiB | +-----------------------------------------------------------------------------+ Quote Link to comment
PTRFRLL Posted November 30, 2021 Author Share Posted November 30, 2021 8 minutes ago, Rhomax said: Tried updating the Nvidia drivers to the 495.44 and recreated the docker several time and still have the same result. The summary is: Using no tags or tags CUDA11 or latest or 3.5.2 I get the following error: TREX: Can't initialize device [ID=0, GPU #0], NVML wasn't initialized I'm wondering if the NVML error message is not the actual root cause and it's down to something like device initialisation more generally but this is what I see on nvidia-smi which doesn't have anything obvious stand out to me. +-----------------------------------------------------------------------------+ | NVIDIA-SMI 495.44 Driver Version: 495.44 CUDA Version: 11.5 | |-------------------------------+----------------------+----------------------+ | GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |===============================+======================+======================| | 0 NVIDIA GeForce ... Off | 00000000:01:00.0 Off | N/A | | 53% 54C P2 45W / 200W | 131MiB / 7982MiB | 0% Default | | | | N/A | +-------------------------------+----------------------+----------------------+ +-----------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=============================================================================| | 0 N/A N/A 1646 C /trex/t-rex 129MiB | +-----------------------------------------------------------------------------+ Can you try this tag and see if it works. I updated the base image and bumped the T-Rex version (note: I moved it to Githubs Container Registry, so you'll need to include the ghcr.io part): ghcr.io/ptrfrll/nv-docker-trex:test Quote Link to comment
Rhomax Posted December 1, 2021 Share Posted December 1, 2021 1 hour ago, PTRFRLL said: Can you try this tag and see if it works. I updated the base image and bumped the T-Rex version (note: I moved it to Githubs Container Registry, so you'll need to include the ghcr.io part): ghcr.io/ptrfrll/nv-docker-trex:test I get an unauthorised message Error response from daemon: Head https://ghcr.io/v2/ptrfrll/nv-docker-trex/manifests/test: unauthorized Quote Link to comment
PTRFRLL Posted December 1, 2021 Author Share Posted December 1, 2021 29 minutes ago, Rhomax said: I get an unauthorised message Error response from daemon: Head https://ghcr.io/v2/ptrfrll/nv-docker-trex/manifests/test: unauthorized Oops! I forgot to set the container as public. Give it a try now. Quote Link to comment
Rhomax Posted December 1, 2021 Share Posted December 1, 2021 Thanks for your help! Same NVML issue unfortunately... this is what the logs look like. Much the same as before 20211201 03:15:53 T-Rex NVIDIA GPU miner v0.24.7 - [Linux] 20211201 03:15:53 r.3aed4eddcdb3 20211201 03:15:53 20211201 03:15:53 20211201 03:15:53 NVIDIA Driver version N/A 20211201 03:15:53 20211201 03:15:53 + GPU #0: [00:01.0|2489] GeForce RTX 3060 Ti, 7982 MB 20211201 03:15:53 20211201 03:15:53 WARN: DevFee 1% (ethash) 20211201 03:15:53 20211201 03:15:53 URL : stratum+tcp://daggerhashimoto.usa.nicehash.com:3353 20211201 03:15:53 PASS: x 20211201 03:15:53 WRK : Unraid-TRex 20211201 03:15:53 20211201 03:15:53 WARN: NVML is disabled. You won't see GPUs stats. 20211201 03:15:53 Starting on: daggerhashimoto.usa.nicehash.com:3353 20211201 03:15:53 ApiServer: HTTP server started on 0.0.0.0:4067 20211201 03:15:53 ---------------------------------------------------- 20211201 03:15:53 For control navigate to: http://172.17.0.7:4067/trex 20211201 03:15:53 ---------------------------------------------------- 20211201 03:15:53 Using protocol: stratum2. 20211201 03:15:53 TREX: Can't initialize device [ID=0, GPU #0], NVML wasn't initialized 20211201 03:15:53 WARN: Miner is going to shutdown... 20211201 03:15:53 Extranonce is set to: 4b5c0e 20211201 03:15:53 Authorizing... 20211201 03:15:53 Main loop finished. Cleaning up resources... 20211201 03:15:53 ApiServer: stopped listening on 0.0.0.0:4067 20211201 03:15:53 Authorized successfully. 20211201 03:15:53 ethash epoch: 457, diff: 1.04 G 20211201 03:15:56 T-Rex finished. Quote Link to comment
Rhomax Posted December 1, 2021 Share Posted December 1, 2021 Actually, looks like this latest image worked once I removed the 'disable NVML' config option. I'd set it to true whilst trying to work out what was going on. Thanks for your help @PTRFRLL ! Really appreciated it Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.