Works on benchmark mode but fails in normal cases

[greentree2010]

I have several questions. I really appreciate if you could give me any hint.

1. What is the benchmark mode? What is the different with the normal mode?

2. The miner works when the benchmark flag is set but fails to run when the flag is removed.

Command: $eth_proxy=stratum+http://x.x.x.x:8080 $./t-rex -d 7 -o $eth_proxy -a ethash --log-path log3.txt -P --benchmark

20210808 14:12:25 T-Rex NVIDIA GPU miner v0.21.5 - [Linux] 20210808 14:12:25 r.738648414f44 20210808 14:12:25 20210808 14:12:25 NVIDIA Driver v450.102.04 20210808 14:12:25 CUDA devices available: 8 20210808 14:12:25 20210808 14:12:25 WARN: DevFee 1% (ethash) 20210808 14:12:25 20210808 14:12:25 URL : stratum+http://x.x.x.x:8080 20210808 14:12:25 USER: 20210808 14:12:25 PASS: 20210808 14:12:25 WRK : 20210808 14:12:25 20210808 14:12:25 Starting on: x.x.x.x:8080 20210808 14:12:25 Using protocol: stratum1. 20210808 14:12:25 Start working in get-work mode... 20210808 14:12:25 ApiServer: HTTP server started on 127.0.0.1:4067 20210808 14:12:25 --------------------------------------------------- 20210808 14:12:25 For control navigate to: http://127.0.0.1:4067/trex 20210808 14:12:25 --------------------------------------------------- 20210808 14:12:25 ApiServer: Telnet server started on 127.0.0.1:4068 20210808 14:12:26 GPU #7(008800): TITAN Xp, intensity 22 20210808 14:12:56 WARN: GPU #7: TITAN Xp is idle, last activity was 30 secs ago 20210808 14:13:01 WARN: GPU #7: TITAN Xp is idle, last activity was 35 secs ago 20210808 14:13:06 WARN: GPU #7: TITAN Xp is idle, last activity was 40 secs ago 20210808 14:13:11 WARN: GPU #7: TITAN Xp is idle, last activity was 45 secs ago 20210808 14:13:16 WARN: GPU #7: TITAN Xp is idle, last activity was 50 secs ago 20210808 14:13:21 WARN: GPU #7: TITAN Xp is idle, last activity was 55 secs ago 20210808 14:13:22 WARN: WATCHDOG: T-Rex has a problem with GPU, terminating... 20210808 14:13:23 WARN: WATCHDOG: recovering T-Rex

[trexminer]

Benchmark mode is for benchmarking your GPUs without actually mining. It's failing in normal mode because it's not receiving data it expects from the ethproxy I think. Try running

./t-rex -d 7 -o $eth_proxy -a ethash --log-path log3.txt -P

and post the log file again.

[greentree2010]

it repeatedly shows

20210808 17:05:54.749[3460677.786] Starting on: x.x.x.x:8080 20210808 17:05:54.749[3460677.786] [asio_thread_func] start 20210808 17:05:54.749[3460677.786] [submit_work_thread] start 20210808 17:05:54.749[3460677.786] [solo] start 20210808 17:05:54.749[3460677.786] [get_stratum_work_thread] start 20210808 17:05:54.751[3460677.787] Connecting to: x.x.x.x:8080, ip: x.x.x.x 20210808 17:05:55.121[3460678.157] ApiServer: HTTP server started on 127.0.0.1:4067 20210808 17:05:55.121[3460678.158] --------------------------------------------------- 20210808 17:05:55.121[3460678.158] For control navigate to: http://127.0.0.1:4067/trex 20210808 17:05:55.121[3460678.158] --------------------------------------------------- 20210808 17:05:55.121[3460678.158] ApiServer: Telnet server started on 127.0.0.1:4068 20210808 17:05:55.331[3460678.367] GPU #7(008800): TITAN Xp, intensity 22 20210808 17:05:58.076[3460681.113] Using protocol: stratum1. 20210808 17:05:58.076[3460681.113] Start working in get-work mode... 20210808 17:05:58.076[3460681.113] x.x.x.x:8080 asks job for block 0 20210808 17:05:58.151[3460681.188] -OUT-> POST / HTTP/1.1^M Host: x.x.x.x:8080^M Accept: /^M Content-Type: application/json^M charsets: utf-8^M Content-Length: 48^M ^M {"id":1,"jsonrpc":"2.0","method":"eth_getWork"} 20210808 17:05:58.351[3460681.388] -OUT-> POST / HTTP/1.1^M Host: x.x.x.x:8080^M Accept: /^M Content-Type: application/json^M charsets: utf-8^M Content-Length: 48^M ^M {"id":1,"jsonrpc":"2.0","method":"eth_getWork"} 20210808 17:05:58.552[3460681.588] -OUT-> POST / HTTP/1.1^M Host: x.x.x.x:8080^M Accept: /^M Content-Type: application/json^M charsets: utf-8^M Content-Length: 48^M ^M {"id":1,"jsonrpc":"2.0","method":"eth_getWork"} 20210808 17:06:26.151[3460709.187] WARN: GPU #7: TITAN Xp is idle, last activity was 30 secs ago 20210808 17:06:26.186[3460709.222] -OUT-> POST / HTTP/1.1^M Host: x.x.x.x:8080^M Accept: /^M Content-Type: application/json^M charsets: utf-8^M Content-Length: 48^M ^M

[greentree2010]

It seems that there is no 'IN' request from ethproxy. I guess OUT is the request for work from miner to ethproxy, IN is the work request from ethproxy. Is it right? If yes then I may guess the root cause.

[trexminer]

That's correct. Ethproxy is not providing any jobs for the miner to solve.