-
Notifications
You must be signed in to change notification settings - Fork 657
Description
Hardware
RAK4631 WisBlock Core (nRF52840 + SX1262)
RAK19007 Base Board
868 MHz
+8 dBi omnidirectional antenna
powered via Solar and Lion Battery
Firmware versions tested
MeshCore 1.12
MeshCore 1.13
MeshCore 1.14
All versions show the same behaviour.
Description
My node runs fine initially but freezes after about 1–2 days.
After the freeze the node does not respond anymore and only a reboot fixes it.
The location is in a RF shadow for transmitting, but reception is extremely good.
With the +8 dBi antenna I receive a very large number of nodes (sometimes even traffic from England). Because of this the node list grows very large.
When the node list approaches ~200 entries the device eventually freezes.
Observations
RX is very strong (large number of nodes received)
TX range is limited due to terrain shadow
Node list grows continuously
freeze happens after ~24–48 hours
reboot restores normal operation
behaviour is identical across firmware 1.12–1.14
Configuration
Frequency: 868 MHz
Hop limit reduced to 12
No external sensors connected
Only LoRa + BLE used
Additional notes
The same hardware runs Meshtastic for months without any stability issues, even during summer temperatures.
So this appears to be specific to MeshCore.
My suspicion is that the node list grows too large and eventually causes a memory issue on the nRF52840 (256 KB RAM).
Questions
Is there currently a limit for the node table size?
Is there a way to automatically prune old nodes?
Would it make sense to add a maximum node limit (e.g. 100–150 entries)?
Is there a watchdog or auto-reboot option available?
If needed I can provide logs or additional information.