-
Notifications
You must be signed in to change notification settings - Fork 19
Open
Labels
Description
Description
Automatically retry failed operations with intelligent exponential backoff strategy.
Requirements
- Detect transient failures (network, temporary unavailability)
- Implement exponential backoff (1s, 2s, 4s, 8s, 16s)
- Maximum retry attempts configurable
- Log each retry attempt
- Different strategies for different error types
- User notification of retry attempts
Acceptance Criteria
- Detects transient vs permanent failures
- Exponential backoff implemented
- Configurable max retries (default: 5)
- Logs all retry attempts
- Different handling for network vs other errors
- User sees progress during retries
- Unit tests included (>80% coverage)
- Documentation with configuration examples
Example Usage
$ cortex install cuda
⚠️ Download failed (network timeout)
🔄 Retrying in 1 second... (attempt 1/5)
⚠️ Download failed (network timeout)
🔄 Retrying in 2 seconds... (attempt 2/5)
✓ Download successfulSkills: Python, error handling, async operations
Bounty: $25 upon merge
Priority: Medium