new example mqtt_airlift_subscribe_time_keepalive.ino

[mikeysklar]

This approach stabilizes long-running MQTT connections on AirLift boards and prevents the “No socket available” disconnect condition that users are encountering.

• Manual KeepAlive interval control
• A scheduled MQTT PING loop
• Connection retry logic

related links:

• Adafruit forum thread
• nina-fw GitHub issue

tested on:

┌──────────────────────────────┬────────────────────────────────────┐
│Component/Library             │Version/Path                        │
├──────────────────────────────┼────────────────────────────────────┤
│Board/Platform                │Adafruit SAMD (Feather M4)          │
│Arduino IDE                   │2.3.6                               │
│Board Core                    │adafruit:samd 1.7.16                │
│Adafruit Zero DMA Library     │1.1.3                               │
│Adafruit-WiFiNINA             │1.6.0 (Airlift fork)                │
│Adafruit MQTT Library         │2.6.1                               │
│nina-fw Firmware              │KeepAlive issue present             │
└──────────────────────────────┴────────────────────────────────────┘

credit to Bruce Conrad from above forum thread for code that demonstrates keepalive issue and workaround being used here.

[mikeysklar]

@tyeth - I've tested this example code and it works. I'll work on updating the nina-fw to avoid the issue entirely, but many users will not be updating the Airlift Firmware so this work around will be helpful.

[brentru]

@mikeysklar Overall, this is great. Proper SPI configuration for AirLift and a lot of print() statements for diagnostics will help people using AMQTT with airlift.

I have some feedback, also could you update the library version within library.properties so we can release this after approved?

[brentru]

This works, but more commonly these functions would be implemented before setup() and loop(), not prototypes.

[brentru]

@mikeysklar Approved and merging in