How to end a Linux Container (LXC) from within?
Linux Containers has a tool "lxc-stop" to stop a container. Is there a way to use that within the container itself? Or is it not necessary if all the processes within exit on their own?
I have some scripts that currently run just in a legacy chroot. But there are issues I believe Linux Containers will solve, such as setting up distinctive network access. I don't really want to run these like a full system, but just as a quick work task in a whole different context. There will be enough file tree to run what I need (already basically works using chroot). But these things run quick (under a minute in many cases), and I don't want to set up something on the outside to monitor when the task is done just to run "lxc-stop". I want to be able to have it start, and have a process exit back when it is done, just like chroot does. But I want to do this with the more advanced isolation of LXC.
So far all the documentation on this is more like treating it as a virtual machine. If there is any documentation that tells how to use LXC WITHOUT those tools (e.g. what setup do the tools themselves need to do), that might work if it's not too complicated (e.g. /proc or /sys settings would probably not be too complicated). Otherwise, how to do these things with the tools would be what I am looking for. They would not be running an init/rc scheme, but just the script for the task (pre-inserted into the tree, of course).
|All times are GMT -5. The time now is 11:35 AM.|