ChiKim
Anyone Getting Bad Pods with Internet Issues?
I'm in US, and I get a lot more bad pods with internet issues than working pods like 7 out of 10. I'm trying to spot a community pod with rtx 4090 and the default template pytorch:2.2.0-py3.10-cuda12.1.1-devel-ubuntu22.04.
When I get a bad pod, I get error pulling image: Error response from daemon: Get "https://registry-1.docker.io/v2/": net/http: request canceled while waiting for connection (Client.Timeout exceeded while awaiting headers)
If the pod runs and if I connect via ssh and try to setup, I often run into problem with apt on ubuntu and python pip.
Sometimes I get certificate error, extremely slow speed less than 10 bytes per second, etc.
I have to keep launching different pod until I get a working one. Anyone has the same problem?
2 replies