These habits complement the task guides in this section.

They reflect how teams use Applications, Environments, and Services without surprising production users.

Use environments deliberately

• Keep production isolated from experiments. Ship risky changes to a staging (or similarly named) Environment first.
• Mirror Variables keys across Environments, but never reuse production secret values in non-production.

Change one thing at a time

• When debugging, separate code changes (Git Redeploy) from variable updates and Scaling edits so you know which action fixed or caused an issue.
• Prefer Redeploy with Use existing cache only when you are confident the image already matches the commit you intend to run.

Protect secrets

• Store sensitive values with secret() or your vault workflow; avoid logging them in Logs.
• Rotate credentials if they leak, then Redeploy every Service that consumed the old values.

Observe before and after

• Watch Deployments logs during rollout, then Logs for steady-state errors.
• After Domains or Proxy edits, verify TLS in a browser and with curl from outside your laptop network if possible.

Back up data, not just code

• Git already versions application source. Database Services still need Backups and tested restores.
• Document who may trigger Restore; it is a high-risk operation.

Plan for rollback paths

• Keep known-good Git tags or Docker tags you can redeploy quickly.
• For data-affecting migrations, ship backward-compatible steps or pair schema changes with restore drills.

Related

• Deployments overview
• Troubleshooting overview
• Service settings