Yarn Install result in code server timed out

[7HE-W0R1D]

I installed code server and an unofficial node 15 on my raspberry pi zero. It can start normally and I can access it with https over browser. However, several dependencies and modules were missing and when I was installing "spdlog", I accidentally installed the official build of node, which is only node 10 on my armv6. Then I reinstalled node 15 but whenever I type code-server, it responses with a error timed out I also tried deleting the code-server folders and run yarn add code-server again but the error persists. I have no idea about how to fix that since there are no extra info provided by the code server. Any help would be appreciated!

[code-asher]

Weird, that's the only thing it outputs?

[7HE-W0R1D]

Yay it worked!!!!! After like an afternoon building the package terminal now shows up and I can type things in it!!! Though it's a bit laggy (2-3 sec) but I'm sooooo glad it works! Now there's only extensions to deal with ;)
Edit: the lag is due to my NAT traversal service, when connected locally there are no noticeable lags.

[7HE-W0R1D]

Now that the server is totally usable, github's back, python's able to run through command line(no triangle run button, probably lacking some extensions? Mentioned it here), and Java files are running smoothly. My NAT traversal service is a bit unstable so connect through weblink is not so fast and sometimes disconnects, but connecting locally now really worked flawlessly.
Huge thanks to @code-asher for this great project and all the time & effort you spent these days helping me!

[4oo4]

I had this same issue, but for some reason I just had to remove old certificates from ~/.local/share/code-server to fix it. Is self-signed.cert the name that Code Server still uses for creating certificates (the docs seem to say that it's self-signed.cert)? I also have a localhost.cert in that same folder, the localhost one was still valid whereas the self-signed ones had been expired for a while.

I was able to launch Code Server right after upgrading to 3.9.0, but for some reason the issue didn't appear until about a week afterward. My hunch is that the two sets of certificates was potentially causing issues.

Cheers

[code-asher]

@4oo4 We changed the generated certificate and key file names. They are now:

1. ~/.local/share/code-server/$hostname.crt
2. ~/.local/share/code-server/$hostname.key

$hostname is taken from --cert-host and falls back to localhost if not set.

It does seem strange the old files could cause an issue though. We don't load them or anything like that. 🤔

[4oo4]

@code-asher That makes sense. I'll keep an eye on it and see if I keep getting it, since after first upgrading to 3.9.0 I didn't have any issues like that, and the old certs have been sitting out there for a while. So there's a chance it could have been something else, like me restarting my docker container that actually fixed it. If it recurs I'll open a new issue.

Thanks