* fix(share sub): fix the issue that the number of subscriptions dropped to 0 during the picking subscriber and caused a crash
* Connection Busy Alarms (#3992)
feat(emqx_connection): improve port_busy alarm
* fix(emqx_connection): tune the congestion alarm params
* chore(deps): upgrade esockd to 5.7.5
* fix(appup): correct the appup file
- Introduce the 'pipeline' design pattern in emqx_protocol module
- Reame the '{dispatch, ...' to '{deliver, ...' in emqx_broker module
- Rename type 'credentials' to 'client'
1. Add do_add_route/1 do_add_route/2, do_delete_route/1,
do_delete_route/2 APIs in emqx_router module
2. Improve the code of emqx_trie module
3. Update the emqx_broker module to call the new APIs of emqx_router
* Add acking mechamism for shared dispatch
For QoS0 messages, no acking
For QoS1/2 messages, 'ACK' at any of events below:
- ACK when QoS is downgraded to 0
- Message is sent to connection process
'NACK' at any of events below:
- Message queue is full and the receiving session starts to drop old messages
- The receiving session crash
Upon 'NACK', messages are dispatched to the 'next' subscriber in the group,
depending on the shared subscription dispatch strategy.
Prior to this change, if a producer session produces to two
or more shared subscriber groups, the previously picked subscriber
for sticky strategy may not be a valid member for the next group.
emqx-ci-robot