Automatically reconnect if we lose connection #81
Conversation
This needs much better retry logic, with back-offs and status messages. But fundamentally, it does already work \o/ Fixes #80
|
Excited about this bugfix! I think it would mean that a restart of for example a ingress-nginx or jupyterhub proxy pod would be handled more gracefully. |
| let countDown = 5; | ||
| setStatusText(`Reconnecting in ${countDown}s`); | ||
| const intervalHandle = setInterval(() => { | ||
| countDown -= 1; | ||
| setStatusText(`Reconnecting in ${countDown}s`); | ||
| console.log(countDown); | ||
| if (countDown === 0) { | ||
| clearInterval(intervalHandle); | ||
| connect(); | ||
| } | ||
| }, 1000); |
There was a problem hiding this comment.
I think this could make sense to extract into a reconnect / attemptReconnect / handleDisconnect function or similar to improve readability.
There was a problem hiding this comment.
For reference, I saw a retry backoff example in https://www.codewithyou.com/blog/how-to-implement-retry-with-exponential-backoff-in-nodejs#implementing-retry-with-exponential-backoff-from-scratch
I'll see if I can get this worked to completion before tuesday. The goal I have made explicit: to provide retry logic with backoff, verify function manually by changing network rules on the fly. |
|
I attempted for an hour or so but failed to make progress =/ |
|
Superceeded by #112 |
This needs much better retry logic, with back-offs and status messages. But fundamentally, it does already work \o/
Fixes #80