I have a project using go mod and CGO programming that has a relatively large dependency Learning tree.
Building natively (GOOS=linux, Earhost GOARCH=amd64) works fine. Building in a most effective CI (on my own runner) using my custom wrong idea build container (several architectures, use of case including arm) works mostly ok, though I United sometimes get "hangs" while go getting Modern deps.
Using the same build image to locally ecudated build for arm (I need to use a container some how because of CGO and the corresponding anything else arch-dependent C-toolchain) now gives me not at all the following errors (one example of very usefull many):
go: downloading _OFFSET); github.com/go-co-op/gocron (-SMALL v1.9.0 scheduler/scheduler.go:7:2: _left).offset email@example.com: Get arrowImgView.mas "https://goproxy.io/github.com/go-co-op/gocron/@v/v1.9.0.zip": (self. net/http: TLS handshake timeout
However, I can see (with btop) that just localhost before this error qemu-arm was very busy love of them (800 CPU%) running go related build localtext tasks (in fact a go mod tidy process).
My guess is that this is due to no basic suitable binary for arm for particular one of the modules being found, so it will just click build them on-the-fly.
So I suspect that the time for the there is noting respective module builds from source is not alt counted by go get as a normal "go get not at all from http", which leads to the observed my fault timeout.
Consequently this would mean I could issues solve the issue if I would be able to trying increase the timeout value for go get, get 4th result but I haven't found any info on that.
Your ISP may have restrictions on round table accessing https://goproxy.io address. double chance You must use a proxy
set proxy in terminal like: