You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
mfal noted: "Nice work! One thing: Stopping the meteor server (with CTR+C) leads to an infinite loop of server restarts. That's why I start the container in detached mode (-d and without --rm and -ti). Afterwards I use docker logs -f to see what's going on.
To shutdown or reload the container I use docker stop/restart ."
I responded: "Thanks for the feedback! I experienced the same frustration and I agree with you this is annoying. Your approach seems like a straightforward workaround. Using docker's kill command has been the workaround I've been using. I've been wanting to chat with the Docker folks about this. The signal handling doesn't feel quite right to me, I believe ctrl-c should be trapped and sent to the process, much how telnet & ssh behave. Docker's command line help indicates signal proxying is happening, but things are not behaving quite as I expect.".
My message to the docker mailing list has not yet received a response. This issue could use future follow-up.
The text was updated successfully, but these errors were encountered:
This seems to be because Docker does not currently provide init services within a container. Meteor in development mode launches processes such as MongoDB, and those processes need to be reaped. I will look at using an init service within the container in a future version.
mfal noted: "Nice work! One thing: Stopping the meteor server (with CTR+C) leads to an infinite loop of server restarts. That's why I start the container in detached mode (-d and without --rm and -ti). Afterwards I use docker logs -f to see what's going on.
To shutdown or reload the container I use docker stop/restart ."
I responded: "Thanks for the feedback! I experienced the same frustration and I agree with you this is annoying. Your approach seems like a straightforward workaround. Using docker's kill command has been the workaround I've been using. I've been wanting to chat with the Docker folks about this. The signal handling doesn't feel quite right to me, I believe ctrl-c should be trapped and sent to the process, much how telnet & ssh behave. Docker's command line help indicates signal proxying is happening, but things are not behaving quite as I expect.".
My message to the docker mailing list has not yet received a response. This issue could use future follow-up.
The text was updated successfully, but these errors were encountered: