WS API client does not emit events
sameersubudhi opened this issue · comments
Expected behavior
The WS API client when connected should always emit the chain_* events.
Actual behavior
When running Lisk Service against Lisk Core (or any other application) for a prolonged duration (at least a couple of days) using the WS API client stops emitting any of the chain_* events resulting in Lisk Service being stuck in a state of limbo even while the chain continues to progress.
Steps to reproduce
- Run Lisk Service against Lisk Core v4.0.0 on testnet for a few days
Which version(s) does this affect? (Environment, OS, etc...)
v6.0.0
Sometimes we get following timeout error when trying to automatically reinstantiate the WS API client.
2023-11-13T21:35:02.838 INFO [nodeStatus] Node synchronization in progress. Will wait for node to sync with the network before scheduling indexing.
2023-11-13T21:35:19.844 ERROR [client] Error instantiating WS client to ws://host.docker.internal:7887.
2023-11-13T21:35:19.844 ERROR [client] Could not connect in 2000ms
The WS API client has been running against my testnet nodes for several days and it has been showing chain_* events correctly.
According to @sameersubudhi, the reported behavior could have been solved with correct re-instantiation of API client by Lisk Service in LiskArchive/lisk-service#1920.
I will close this issue now. If the problem re-appears, please feel free to re-open it.