T-Rex terminates immediately on WSL2
richardwu opened this issue · comments
Thanks for releasing the source files.
I've been trying to get this thing to work on WSL2: I've managed to get the container to run but it seems terminate immediately after building the DAG.
docker run --dns 8.8.8.8 --name='trex-miner' -e WALLET=0x997d01bA51fF2f13436BF97bb9545da8Bb625465 -e SERVER=stratum+ssl://us2.ethermine.org:5555 -e WORKER=narwhal -e ALGO=ethash -e NVIDIA_DISABLE_REQUIRE=1 -p '4067:4067/tcp' -p '4068:4068/tcp' --runtime=nvidia rwu1997/nv-docker-trex
Starting T-rex miner...
============================================================
Server: stratum+ssl://us2.ethermine.org:5555
Algorithm: ethash
Wallet: 0x997d01bA51fF2f13436BF97bb9545da8Bb625465
Worker: narwhal
============================================================
20210316 02:20:01 WATCHDOG: initialize error, memory dump: eyJhIjoiM0dwc2pQdUZWYlhEaHllY2ZVb1AxNTBpcjc1ZFlaaG03ckRMYnYxR3FPTmFRSUNyaVFWRVdpMm1oaXRLYVh1aFJNV1g0UW9YYXdDSS9PeG1SOCtBM0xaVkl5STZrUlZXNDdQUGtuNzNrV2w1aTVDWUJkNER0Si9WU3VrK1dHaExkNVJGZTZ0U0c5TEVxWm9wcEFMczh3PT0iLCJiIjoicnRzbWZSVDJLa0Vva0ZXcTBXRTVXZz09In0K
20210316 02:20:02 T-Rex NVIDIA GPU miner v0.19.11 - [CUDA v11.10 | Linux]
20210316 02:20:02 r.9c70a63f2290
20210316 02:20:02
20210316 02:20:02 NVIDIA Driver v470.05
20210316 02:20:02 CUDA devices available: 1
20210316 02:20:02
20210316 02:20:02 WARN: DevFee 1% (ethash)
20210316 02:20:02
20210316 02:20:02 URL : stratum+ssl://us2.ethermine.org:5555
20210316 02:20:02 USER: 0x997d01bA51fF2f13436BF97bb9545da8Bb625465
20210316 02:20:02 PASS:
20210316 02:20:02 WRK : narwhal
20210316 02:20:02
20210316 02:20:02 Starting on: us2.ethermine.org:5555
20210316 02:20:02 Using protocol: stratum1.
20210316 02:20:02 Authorizing...
20210316 02:20:02 ApiServer: HTTP server started on 0.0.0.0:4067
20210316 02:20:02 ----------------------------------------------------
20210316 02:20:02 For control navigate to: http://172.17.0.2:4067/trex
20210316 02:20:02 ----------------------------------------------------
20210316 02:20:02 ApiServer: Telnet server started on 127.0.0.1:4068
20210316 02:20:02 Main loop finished. Cleaning up resources...
20210316 02:20:02 ApiServer: stopped listening on 0.0.0.0:4067
20210316 02:20:02 ApiServer: stopped listening on 127.0.0.1:4068
20210316 02:20:02 Authorized successfully.
20210316 02:20:02 ethash epoch: 401, block: 12046948, diff: 4.00 G
20210316 02:20:02 ethash epoch: 401, block: 12046948, diff: 4.00 G
20210316 02:20:02 WARN: GPU #0(000b00): Zotac NVIDIA GeForce RTX 3080, intensity set to 22
20210316 02:20:03 GPU #0: generating DAG 4.13 GB for epoch 401 ...
20210316 02:20:05 T-Rex finished.
WATCHDOG seems to fail, but I don't think that's what's killing the container.
Did you take a look at this issue from the Trex repo?
Thanks for the x-ref. Looks like I'm on Build 21322 so I shouldn't be affect by that issue.
It may be because my CUDA for WSL is outdated (11.0 vs 11.2). Going to circle back once NVIDIA re-enables WSL CUDA drivers again on their website...
Going to close this for now. Feel free to re-open once you get WSL CUDA running again