SSH Failing Unexpectedly #2035
andrewvaughan
started this conversation in
General
Replies: 1 comment
-
@catthehacker any chance you might have some insight here? |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hello all -
I've been using https://github.com/nektos/act for a few months now with great results. Suddenly, a bunch of my repositories are failing for reasons seeming to do with SSH, but I'm struggling to get to the bottom of the problem.
Also, in the same time period, I have never been able to get
act
to complete a script. It always just hangs, regardless of what is being done. No errors or messages in the verbose logs, just hanging whether successful or failing. I even went back a few versions and still have the issue. I don't know if these are two issues from the same problem, but I'm having a very difficult time working at the moment.My gut tells me that
act
is making an SSH call to ... something (likely GitHub) ... behind the scenes and is getting a prompt that is not handled well in the container and is just causing a hang, but with limited insight into what's going on, it's hard to tell.Any ideas or help would be extremely appreciated.
act
docker
Suddenly, all of my https://github.com/actions/checkout calls started hanging with no errors or details, even with verbose logging.
I finally got to the point where I think that there is a prompt somewhere that the system is waiting for - either confirming a host or entering a password, neither of which make sense to me. I was using 1Password as an
IdentityHost
for SSH on my local development machine, which required fingerprint access to clone repositories, so I disabled that, but I still have the issue.I was able to confirm that the token I was sending with
-s "$(gh auth token)
was valid by calling the GitHub API withcurl
, so I was really hitting my head on the desk as to what the issue was.In another repository, I tried running
act
against a known-working oxsecurity/megalinter workflow. This also failed during a checkout, but I was able to get more information:I had never seen this error before - it seemed really unusual. The megalinter workflow has configuration options for
PRE_COMMANDS
so I added the following to be run first:The last command scanning the
github.com
keys and adding them all to known hosts. Running again, the error changed:I have never had an issue pulling a private repository with this token before... so this is really surprising.
And that's where I am at. I've had this happen with
act
versions 0.2.49, 0.2.51, and 0.2.52 (latest release).This is the more or less the command I'm running (you can see the exact command at the top of each of the logs below):
act --rm --container-architecture linux/amd64 -s "$(gh auth token)" pull_request
And here are the workflows and their (sanitized) verbose logs:
Verbose Logs for Above
The system hangs indefinitely at the end of that log.
Verbose Logs for Above
The logs above include the
PRE_COMMANDS
listed adding theknown_hosts
to remove all doubt, but I have a feeling this is all intertwined into whatever the issue is. I've never had to manually create known hosts like that before in theact
Docker container.Thanks again... any help would be extremely appreciated as all of my projects are currently stuck with this ongoing issue.
Beta Was this translation helpful? Give feedback.
All reactions