-
Notifications
You must be signed in to change notification settings - Fork 211
Open
Description
Consider a smaller exponential backoff that will allow for up to 90% improvement in most cases where retries occur.
Possible implementation:
- Maybe update from 1s → 100ms
- Maybe add variance so that we do not constantly hit rate limits since all are at the exact same time.
Considerations
- Will a faster backoff actually work against us and cause more rate limits?
Reference
ethers.js
How does exponential backoff work? ethers-io/ethers.js#3252, who is a big proponent of randomized exponential backoff.
Metadata
Metadata
Assignees
Labels
No labels
Type
Projects
Status
Backlog