Did you encounter issues with event order during asynchronous retries?

Yes, with asynchronous retries, there are often issues with the order of events, especially if retries are initiated in parallel or without proper synchronization.

For example, if requests are retried asynchronously, a response from a later request may arrive before an earlier one, leading to an inconsistent state.

To avoid such problems, the following are used:

• Queues or semaphores for sequential execution of retries.
• Canceling previous attempts when a new one is started.
• Using unique identifiers to match responses with requests.

Thus, it is important to control the order and state during asynchronous retries to prevent race conditions and incorrect data.