This is an original planning guide. It is not a vendor manual, an official standard, or legal advice. Example addresses use documentation ranges. Review every command before you paste it.
A bill is a list of names you chose
Cost work on this desk is an estimate you can read before a resource exists. The estimator drafts a monthly figure from sizes you type. The rightsizing calculator drafts a smaller size from utilisation you type. Neither tool logs into an account, and neither tool changes a live machine. You still decide.
Start with one workload. Name the owner, the environment, and the hour it should stop. A lab without an owner becomes a line item nobody recognises. A tag you invent after the invoice arrives is a label, not a control.
A commitment is a different hour. Reserved capacity and a savings plan are promises. This page stops at the estimate and one idle resource you can explain. The commitment waits until the same workload has run long enough to measure.
A recommendation is not a resize. If the calculator says a machine looks large, you write that down. You do not resize a machine that is serving people in this window. The instance guide owns the size you pick next.
One estimate, one owner, one stop time
Type the size you think you need into the estimator. Use a lab size, not the largest size on the menu. Record the figure, the assumptions, and the date. If you cannot say what the figure includes, the estimate is not done.
If you already have a utilisation number you are allowed to use, type it into the rightsizing calculator. The suggestion goes in the ticket beside the current size. Agree or disagree in a sentence. Do not click a resize from this page.
Pick one idle resource you administer, or write that the lab has none yet. Idle means nobody can name the workload. The action is a stop or a delete of that lab resource, after a second person agrees it is unused. A production name is not idle because it was quiet for an afternoon.
Write the tag keys you will require: owner, environment, and purpose. A resource without those three is a resource you do not launch. The database cost companion is a neighbouring desk. This page does not price a database you have not designed.
# Cost note. Nothing here logs into an account.
# One workload. One owner. One stop time.
# A calculator suggestion is not a resize.
# A commitment is a later ticket.
# prompt-for-this is not a billing account id you paste into git.
A change you can finish
The change in this window is the written estimate and, if you found a true lab leftover, stopping that one resource. A fleet cleanup is a programme, not this hour.
Rollback for a stopped lab machine is starting it again. Rollback for a deleted disk does not exist unless you already had a copy. Do not delete storage to save money in the same hour you are still learning the bill.
- Name the workload, the owner, and the environment.
- Run the estimator on a lab size and save the figure.
- If you have utilisation, record the rightsizing suggestion.
- Agree or disagree in one sentence.
- Write the stop time for the lab.
- Stop one confirmed idle lab resource, or write that there is none.
- Leave commitments and live resizes for their own tickets.
Mistakes that outlive the command
cloud cost fails in the same few ways. Each one looks finished on the day you type it. The list is here so the ticket can forbid them before the window, not so you can recognise them in an incident review.
- Launching the largest size to see what happens.
- A resize of a live machine from a calculator suggestion.
- A reserved purchase in the same hour as the first estimate.
- A resource with no owner tag.
- Calling a quiet production system idle.
- Deleting a disk to make the bill smaller.
- Putting an account id or a key in the ticket.
- Estimating a fleet you have not named.
- Ignoring the stop time on a GPU or a large lab.
- Treating the estimate as the invoice.
Decisions to write down before the window
One decision for cloud cost: The workload in the estimate is the one in the ticket. The owner and the environment are written. The size is a lab size. The rightsizing line is a suggestion, or it is blank because you have no utilisation yet. No live machine was resized. No commitment was purchased. The lab has a stop time. Idle meant a lab leftover a second person recognised. Storage was not deleted to save money. No secret is in the note. The instance guide owns the next size. And one workload is this window.
The hour you stop
Write the account, the project or subscription, the region, and the clock time before you change anything for cloud cost. If this account is serving people and you do not have a way back, this hour is a plan. A plan you can read tomorrow is a success. A change you cannot undo is not.
Use a lab account, a lab project, or a lab subscription you administer. You do not point a CLI, a pipeline, or a browser tool at an account you do not run. The tools on this site run in the browser. They draft text. They do not log in, and they do not scan.
prompt-for-this is a reminder that a secret is missing. It is not an access key, a token, a password, or a kubeconfig you save. The real secret stays in the password store or the platform's secret store. If a file, a log, a ticket, or a pipeline still contains a live secret, stop and rotate it. Then delete the copy.
One change. A network, a virtual machine, a policy, and a pipeline do not share a window. If two things move and the outcome is wrong, you will not know which one did it. The rollback is the previous file, the previous plan, or the resource you can delete because it was only a lab.
A second person reads the scope aloud: which account, which region, which name. If they cannot find it in the ticket, the ticket is not ready. Silence is not agreement.
Documentation addresses stay in the note. 192.0.2.0/24, 198.51.100.0/24, 203.0.113.0/24, and 2001:db8::/32 are not prefixes you type into a production network. example.com is not your zone. A generated file that opens 0.0.0.0/0 is a file you edit before it is applied. A source you can name is the rule you keep.
Do not apply a draft you have not read. A cost estimate, a policy, a manifest, and a Terraform plan are reading material. Apply, a resize of a live machine, and a delete of a bucket are later decisions with their own proof. Read the names. Then decide.
Keep the previous file. A generator draft is not the previous file. The previous file is what you put back if the new one is wrong. You practised the path back, including the serial console or a second admin, before you changed a network or a lockout setting.
If a step fails twice, stop. The failure is the result. Adding a command from a forum beside a half-applied stack produces an account nobody can explain. Rebuild the lab from the copy you kept.
When the checks pass, the next account waits until someone else can repeat the proof from the notes. cloud cost is done for this window when the proof is written, not when a CLI printed a success line.
This page is a planning guide. It is not a vendor manual, not an official standard, not legal advice, and not an audit opinion. If an auditor must see the decision, they see the ticket and the evidence, not this paragraph.
Hand the finished note to the neighbouring guide that owns the next step, and leave this account alone until that guide has its own window.
Cost checks before another resource is launched
Confirm this before you call the change done: The estimate figure and the date are in the ticket. The assumptions are sentences, not only numbers. The owner tag is decided. The stop time is decided. A live resize did not happen. A commitment did not happen. A second person can point at the workload. The lab account is the account you administer. No billing export was fetched from an account you do not run. The next size belongs to the instance guide. GPU or other large labs are named as expensive on purpose. The note can be repeated without opening the console. The hostname in the prompt matches the hostname in the ticket for cloud cost. The result line includes the time and the person, not only the command. You can point at the rollback value without searching a chat history. And done is a sentence a second person can repeat: the proof for cloud cost happened, and the next host is not started early. If one of those fails, stop.
Tools for this guide
These tools run in the browser on this site. They do not log into your machines.
Related pillar guides
Guides in this cluster
These supporting pages belong to this pillar.
- Checklist: Cloud Cost Reduction Checklist
- Comparison: Reserved Instances and Savings Plans
- Explainer: Spot Instances Explained
- How-To: Rightsizing Cloud Resources
- Explainer: Understanding Data Egress Charges
- Explainer: Hidden Cloud Costs (NAT, Load Balancers, Public IPs)
- Explainer: Serverless Pricing Explained
- Explainer: Cloud Storage Pricing Explained
- How-To: Kubernetes Cost Management
- Comparison: Cloud vs On-Premises TCO
- How-To: Cloud Budgets and Cost Alerts
- How-To: Tagging for Cost Allocation
- Explainer: FinOps Explained
- How-To: Querying Cloud Bills (Cost Explorer, Cost Management, BigQuery)
Companion: DBA: Managed Database Cost Comparison.