Rancher strategy instructions

[cpursley]

Hi, thank you for the awesome library.

I'm trying to use the Rancher strategy but am confused about the configuration.

In the following section:

Create a wrapper script which will interpolate the current ip of the container.

#!/bin/sh

export CONTAINER_IP="$(hostname -I | cut -f1 -d' ')"
export REPLACE_OS_VARS=true

/app/bin/app "$@"

^ where does this wrapper script go? Into the vm.args.eex file? Into the apps root directory or under apps bin/ folder? Or directly in the K8s environment?

What if anything should go into env.sh.eex?

Also, is it exactly /app/bin/app "$@" or /app/bin/{myappname}? Ditto for -name {myappname}@${CONTAINER_IP}`?

Thanks!

[cpursley]

Ah, this looks like a pre-K8s (pre 2.0) strategy?

[bitwalker]

The wrapper script can go anywhere, it simply handles starting your application, whether that is calling the release script (i.e. bin/yourapp) or just invoking Mix, i.e. mix run --no-halt. The purpose of it is to export those values into the environment so they can be used by the strategy.

To be clear, I don't use the Rancher strategy, so I have no idea how up to date it is - if it should be deprecated and people aren't using it, I'm happy with removing it, since I don't like having things in core that I can't maintain myself. It has been around for awhile though.