chore(session-recordings): separate session recordings from events

[hazzadous]

We currently have session recordings events going to the same endpoint
as analytics events and decide endpiont. The characteristics of session
recordings is quite different, and our expectations of priority. It's
much more important that we capture events and service decide endpoints
than we capture all session recordings.

This separates the deployment to offer more flexibility in providing
different service availability.

Description

Type of change

• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)

How has this been tested?

Checklist

• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• I have added tests that prove my fix is effective or that my feature works

[pauldambra]

Suggested change

	env: []
	env: [
	DEPLOYMENT_ROLE: "SESSION_RECORDINGS_CAPTURE"
	]

[pauldambra]

I came poking about in this repo to figure out how to make exactly this and this PR is sat here like it's already xmas 🎉

I'm guessing it's safe to deploy this because we won't route traffic to it automatically?

I'd be interested in getting it reading production traffic but not pushing on to actually ingest so we can run some experiments on how we're chunking and compressing recordings...

[hazzadous]

I came poking about in this repo to figure out how to make exactly this and this PR is sat here like it's already xmas 🎉

I'm guessing it's safe to deploy this because we won't route traffic to it automatically?

I'd be interested in getting it reading production traffic but not pushing on to actually ingest so we can run some experiments on how we're chunking and compressing recordings...

I can't remember what the rollout strategy is but yeah I think something like get the deployment up first with no traffic routed (so we can be sure the ingress isn't pointing to a non-existent deployment on initial deploy), then do another to have things routed.

Please accept this present 🎁 and feel free to get it deployed.