fix(ci): Delete mainnet and testnet disks separately

[teor2345]

Motivation

In PR #6581, I started creating testnet cached state disks.

But the current disk cleanup job will treat them like mainnet disks, causing CI bugs like:

• CI will take 3 days to run if all the recent disks are testnet, and all the mainnet disks are deleted
• CI will fail if all the mainnet disks are deleted in the middle of a job

This is part of #6367.

To avoid these bugs, I temporarily disabled the disk deletion workflow in GitHub Actions.

Solution

Delete the mainnet and testnet disks separately, so each network always keeps a few disks.

Review

This is an urgent fix because it could stop all PRs merging for 3 days.

Here's a manual run of this workflow from this branch, after PR #6581 generated some testnet disks:
https://github.com/ZcashFoundation/zebra/actions/runs/4824829115

Reviewer Checklist

• Will the PR name make sense to users?
  • Does it need extra CHANGELOG info? (new features, breaking changes, large changes)
• Are the PR labels correct?
• Does the code do what the ticket and PR says?
  • Does it change concurrent code, unsafe code, or consensus rules?
• How do you know it works? Does it have tests?

Follow Up Work

Re-enable disk deletion by turning this action back on:
https://github.com/ZcashFoundation/zebra/actions/workflows/delete-gcp-resources.yml

[teor2345]

Anyone can re-approve this PR, Deirdre has already reviewed it.

[gustavovalverde]

LGTM!