Add retries to chainservice's RPC provider reconnect logic
bitwiseguy opened this issue · comments
Currently if the chainservice
's RPC provider produces an error on either the eventSub
or newBlockSub
channels, the chainservice
attempts to reconnect once. If that reconnect fails, the chainservice
causes the node to panic.
We should make the reconnect more robust by:
- adding additional retry attempts (with exponential backoff?)
- need to account for missed chain events while RPC provider is disconnected
- should the
messageservice
still process messages while the RPC provider is disconnected?