ease of use: debug mode

[cyberhck]

Currently, if we wanted to use debug (heaptrack) mode on kubernetes, we have few issues:

• the file it writes to, doesn't have any identifier, so if there are multiple pods, they override the same file
• There's no easy way to specify that a container shouldn't be restarted inside the pod. (so that we get new pod if the current pod crashes)

Describe alternatives you've considered

• writing our own charts and forking dockerfile ourselves, but this is unreasonable

If you guys are okay to change this, I'd like to send a PR that:

• let's user specify pod lifecycle
• use ${hostname} in the heaptrack file path

Tasks

Preview Give feedback

1. add ability to specify pod lifecycle
2. use ${hostname} in the filepath when using debug mode

[BrynCooke]

This all sounds reasonable. @garypen Do you agree?

[cyberhck]

if @garypen agrees, I want to create a PR for it.

[garypen]

@cyberhck That seems like a reasonable approach.