Problem

When using the serverless provider in elastic-package, the creation of a Serverless project includes a wait loop to detect the Fleet Server URL.

However:

• The values retriesDefaultFleetServerTimeout and retriesDefaultFleetServerPeriod are hardcoded in the source code.
• They cannot be overridden using the profile configuration.
• In some environments (slow provisioning, high load...), Fleet takes longer than the default 10 seconds to become available.
• This leads to premature failures during serverless environment setup.

Proposed Solution

Add two new profile configuration parameters:

• stack.serverless.fleet.period
• stack.serverless.fleet.timeout

These values should:

1. Be read from the profile.
2. Fallback to existing defaults if the configuration keys do not exist or are invalid.

Example configuration:
``ini
stack.serverless.fleet.period = "5"
stack.serverless.fleet.timeout = "60"

This would yield:
	•	Retry period: 5 seconds
	•	Retry timeout: 60 seconds
	•	Fallback to defaults: 2 seconds period, 10 seconds timeout

Expected Benefits
	•	More robust serverless project creation, especially in slower cloud regions.
	•	Better stability when Fleet Server initialization is delayed.
	•	Improved developer experience for contributors/debugging test failures.