I currently have a simple script which acts as a tracking pixel and executes some logic and then saves the results back to Amazon DynamoDB using a PHP API function.
The script is loaded (as a pixel) and executes the logic in real-time.
This should be rewritten so that:
- The tracking pixel simply writes the URL to a log file (no processing in real-time)
- The main script should be designed as a cronjob to pull these logs and process in bulk
Therefore the server has far-less realtime load. The cronjob script can also be more efficient due to fewer database calls.
This should hopefully be a 1 to 4 hour job. There is likely to be ongoing ad-hoc work offered if this goes well.
Please tell me:
How long you think this would take, and whether you are familiar with Amazon's DynamoDB API (if not - are you familiar with using objects etc?)