Error while creating tunnel (more than 20 lines in response)

[mbastias]

I don't know if Cloudflare made a change to their api or is only temporal, but today I tried starting the backend and it failed to start the tunnel. I checked the code and there was no problem except that the script only checks 20 lines and the metrics url was further down. Just bringing it to your attention in case a change has to be made.

[Jelosus2]

Do you have any log that can show what you just explained?

[mbastias]

Not really, I closed the remote and the logging I had wasn't from here, I had to add prints myself to realize the problem. I can try to reproduce, but as I said, I don't know if it was a coincidence. I changed the 20 to 30 and it worked. https://github.com/derrian-distro/LoRA_Easy_Training_scripts_Backend/blob/162e56de18e4d4ab96b066380606d657a68c66fa/utils/cloudflare_tunnel.py#L81

[mbastias]

I could reproduce on runpod.

What I get just from the logs:
/workspace/LoRA_Easy_Training_scripts_Backend/sd_scripts/venv/lib/python3.10/site-packages/transformers/utils/generic.py:441: UserWarning: torch.utils._pytree._register_pytree_node is deprecated. Please use torch.utils._pytree.register_pytree_node instead. _torch_pytree._register_pytree_node( /workspace/LoRA_Easy_Training_scripts_Backend/sd_scripts/venv/lib/python3.10/site-packages/transformers/utils/generic.py:309: UserWarning: torch.utils._pytree._register_pytree_node is deprecated. Please use torch.utils._pytree.register_pytree_node instead. _torch_pytree._register_pytree_node( /workspace/LoRA_Easy_Training_scripts_Backend/sd_scripts/venv/lib/python3.10/site-packages/diffusers/utils/outputs.py:63: UserWarning: torch.utils._pytree._register_pytree_node is deprecated. Please use torch.utils._pytree.register_pytree_node instead. torch.utils._pytree._register_pytree_node( tokenizer_config.json: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 592/592 [00:00<00:00, 3.19MB/s] vocab.json: 862kB [00:00, 15.5MB/s] merges.txt: 525kB [00:00, 30.0MB/s] special_tokens_map.json: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 389/389 [00:00<00:00, 1.63MB/s] tokenizer.json: 2.22MB [00:00, 61.2MB/s] config.json: 4.19kB [00:00, 11.2MB/s] Download cloudflared...: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 39.3M/39.3M [00:00<00:00, 75.9MB/s] Traceback (most recent call last): File "/workspace/LoRA_Easy_Training_scripts_Backend/main.py", line 264, in <module> app.state.TUNNEL.run_tunnel(port=config_data.get("port", 8000), config=Path(config_path) if config_path else None) File "/workspace/LoRA_Easy_Training_scripts_Backend/utils/tunnel_service.py", line 16, in run_tunnel self.running_tunnel = self.tunnel(port=port) File "/workspace/LoRA_Easy_Training_scripts_Backend/utils/cloudflare_tunnel.py", line 96, in __call__ raise RuntimeError("Cloudflared failed to start") RuntimeError: Cloudflared failed to start

I'll do the other ones on a file.

20log.txt
30log.txt

[Jelosus2]

They might have changed stuff and now it takes more than the 20 hard coded lines. @derrian-distro could you take a look at this?

[derrian-distro]

Over the weekend, probably, maybe, time isn't something I particularly have nowadays

[nawka12]

Hi, is there any ETA for this fix? Or maybe anyone has a temporary fix for this that I can apply in the meantime?

[mbastias]

Hi, is there any ETA for this fix? Or maybe anyone has a temporary fix for this that I can apply in the meantime?

Hi, I gave the temporary fix a couple of comments back here. Open the cloudflare_tunnel.py file, change the 20 in the if to 30 and it works. Maybe using postman I can get a more robust solution, that's likely to break from an API change too.