Support initialization after creating a Helm deployment

[KodrAus]

See: https://docs.datalust.co/discuss/5ef1f87fc1d6490230b10d96

Trying to script the post-deployment of Seq in Kubernetes can be a bit of a hassle. It might be possible to put something together using init containers that can either run a set of scripts from a directory, or just a shell script/binary that interacts with the Seq API.

[KodrAus]

We could lean on seqcli for this, and use a config map to store a set of scripts that call the CLI in a container that has a connection profile configured for the upcoming Seq instance:

setup: 
  retention.ps1: |
    $signal = seqcli signal create --title 'Exceptions' --filter "@Exception is not null"
    seqcli retention create --signal-expression $signal

[andymac4182]

I had a play at this. It is designed to be just used by providing a file that will be ran once the Seq server is accepting requests.

https://github.com/andymac4182/seqinitscript

run.sh is a modified version of the file from the Seq container