During the meeting, Kevin discussed a solution using Redis to create manual games in case of Rapid API failure. He explained that a background task is set to run every 48 hours, which extends the game expiration date by another year. However, he expressed concern about accidentally clearing the cache and losing the data. Suggestions were made to implement safeguards such as creating a backup or using a third-party logging service to store the JSON data. It was emphasized that in a redundant scenario, backup systems are crucial, but it's also important to have enough room to find a new solution if necessary. The discussion revolved around balancing redundancy and tap dancing in a dire situation. It was suggested to aim for 80% readiness with another vendor, allowing for a smoother transition in case of any issues. The importance of global optimization and reasonable accommodations was highlighted, along with the expectation of a brief panic mode. Overall, the meeting touched upon the need for a backup plan and ensuring the system can continue running smoothly after the initial 48-hour period.
(Source: Office Hours 7/24/23 )
