There are several potentially problematic behaviours with streaming pull stream retries that may have been making some of the known issues worse (or causing them).

• The minimum retry time is too high
• The default maximum retry time is too high (5 minutes)
• The back-off behaviour could result in very long retry times with long-term streams
• Streams within the stream pool are retried all at once, which might increase lag

In order to better get debugging information on these in the future, we also want to add some more stream logging messages for when things disconnect and reconnect. These will go out to the 'debug' channel of the subscription.