all workers exit by websocket routine exiting

[ffteng46]

Describe the bug
A clear and concise description of what the bug is.
all workers exist with same error.First in lotus daemon,get this unknown actor code bafkqadlgnfwc6mrpmfrwg33vnz2a,then in miner daemon receives message that all workers connection timeout and disconnect from miner.All the miners have the same log RPC client error: sendRequest failed: websocket routine exiting,all workers exit but lotus and miner still work.
To Reproduce
Steps to reproduce the behavior:

1. Run '...'
2. See error
   restart all workers with lotus-worker run --*** --**
   Expected behavior
   A clear and concise description of what you expected to happen.
   all workers works fine for sealing.
   Screenshots
   If applicable, add screenshots to help explain your problem.
   
   
   
   

Version (run lotus version):
Daemon: 0.10.0+git.a6b218075.dirty+api0.16.0
Local: lotus version 0.10.0+git.a6b218075.dirty
Additional context
Add any other context about the problem here.

[TerenceClark]

I also encountered websocket routine exiting because it took me too long to restart Miner, but the API connection timeout was 30s, so it would not be reconnected after the timeout, resulting in websocket routine exiting being reported all the time. By lengthening the timeout, it will keep retrying to connect to Miner until Miner is started.

nodeApi, closer, err = lcli.GetStorageMinerAPI(cctx,
	jsonrpc.WithTimeout(30*time.Minute))