Broken my fly ssh console command

1

Not quite sure how I’ve done this but I appear to have broken my fly ssh console command:

$ fly ssh console
Connecting to tunnel ⣷ Error tunnel unavailable: failed probing "vodon": context deadline exceeded

I’ve tried starting the agent and adding the key again:

$ eval `ssh-agent`
Agent pid 6113
$ fly ssh issue vodon sam@richardson.co.nz --agent
Populated agent with cert:
ssh-ed25519-cert-v01@openssh.com REDACTED
$ fly ssh console
Connecting to tunnel ⣷ Error tunnel unavailable: failed probing "vodon": context deadline exceeded

The main change is that I’ve enabled IPv6 on my home router now. Could that have something to do with it?

2

Hmm.

Assuming you are on the latest version of the CLI, you could try
fly doctor
… and see what that says.

Given that ... failed probing "name": context deadline exceeded error, that sounds like an error other people have had (in their case when deploying, but I assume it affects all commands).

Maybe take a look at this (it’s a long thread). Maybe the IPv6 affects the peer or how the innards work:

3

Fly doctor reports much the same:

$ fly doctor

TEST            STATUS
Agent           PASS
Docker (local)  PASS
Probe (app)     failed probing vodon: failed probing "vodon": context deadline exceeded
Token           PASS
Unix socket     PASS

I’ll have a poke around that topic

4

Hmm, I am using fly under the WSL2 subsystem on Windows.

5

Try running fly wg status, and then select the organization and peer you’re using (you can see the peer in ~/fly/config.yml). You should see something like this:

➜  ~ fly wg status
? Select organization: Kurt (personal)
? Select peer: interactive-Kurts-MacBook-Pro-kurt-fly-io-458
Alive: true
Last Source Address: 99.104.68.96:49751
Installed On Gateway At: 2022-01-22T17:42:21Z
Traffic: rx:2064391728 tx:86009584

If that looks weird, run:

fly agent stop

Then edit ~/fly/config.yml to remove the peer block entirely. It’ll create a new one.

6

Hi, try scaling app to higher memory