Cloud Function Naming FAQ
Short answers to the questions developers ask most about naming serverless functions.
What characters are allowed in AWS Lambda function names?
Letters (a-z, A-Z), digits (0-9), hyphens (-), and underscores (_), up to 64 characters. Spaces, dots, and slashes are not allowed. The NimbusName generator only produces names within these rules.
How long can a cloud function name be?
AWS Lambda allows up to 64 characters, Azure Functions up to 60, and Google Cloud Functions up to 63. Staying under 50 characters keeps a name portable across all three providers.
Should I include the environment in the function name?
Yes, if one account or project hosts multiple environments. Use short, fixed spellings — dev, staging, prod, test — in the same position in every name. If each environment has its own account, the tag is redundant and can be dropped.
Is kebab-case or snake_case better for function names?
Kebab-case (billing-prod-process-payment) is the safest default: it is valid on AWS, Azure, and GCP and reads well in consoles and logs. Snake_case is fine on AWS and Azure but not valid on Google Cloud Functions.
Can I rename a Lambda function after creating it?
There is no rename operation — the name is part of the function's ARN and determines the default CloudWatch log group. Renaming means creating a new function and migrating triggers, permissions, and event mappings. Choose carefully up front.
Does the function name affect cost or performance?
No. The name has no effect on pricing, cold starts, or execution speed. It affects operability: debugging, IAM auditing, log searching, and incident response.
Should the function name match the file or handler name?
It helps but is not required. Many teams keep the handler file generic (index.handler) and put the descriptive identity in the deployed function name, which is what operators actually see.
What is a good pattern for scheduled functions?
Lead with the trigger: cron-daily-cleanup, cron-nightly-reconcile. The schedule is the function's identity, so making it the first token keeps related jobs grouped and obvious.