[OLD] LogicMachine firmware 2025.08

[alcalini]

Hi everyone, and thanks in advance for any time and guidance you can share to help us resolve the unexpected reboots and intermittent trend-persistence issues we’re seeing on one of our LogicMachines. We’ll attach screenshots (System start alerts, trend charts, and system load) to illustrate the behavior.

• “System start” entries (reboot log list)
  
• Daily trend charts showing flat segments after a reboot
  
• System load (CPU/RAM) view
  
• Alertas.
  

Setup / environment

• Device: LogicMachine Lite IMx6 (2023) with SD storage mounted 
  
• Power: controller fed by a UPS (no brownouts)
  
• Firmware tested:
  • 2025.08 → many more reboots per day
    
  • 2024-04-26 → noticeably fewer reboots with the same project
    
• Workload: trends enabled (we’ve heavily reduced the number and retention)
  

What’s happening

• After a reboot, trends often show a flat line (last held value) until the next internal save.
  
• On 2025.08 we see frequent “System start” entries, whereas 2024-04-26 is more stable.
  
• CPU and RAM look normal in the system load page (no clear overload).
  
• SD card looks healthy: mounted, writable, plenty of free space; simple write/read checks pass.
  
• Power is stable (UPS).
  

What we already tried

• Reduced the number of trends and long retentions to rule out storage/retention pressure.
  
• Compared behavior across two firmware versions (2025.08 vs 2024-04-26).
  
• Basic SD diagnostics (mount state, free space, simple writes)—no obvious errors.
  
• Checked system load—no spikes that correlate with reboots.
  

Observations

• Compared with the older behavior (save once per 24 h at midnight), the system now seems to write trend data more frequently, which is good—but unexpected reboots still create gaps/flat segments until the next save.
  
• Reboot cadence is significantly higher on 2025.08 for us.
  

Version stability timeline (attached:Alerts.gif)

• Nov 3–4: Tested on LM 2025.08 with the full set of trends enabled. Result: numerous spontaneous reboots across both days (see Alerts.gif
  → “System start” entries).
  
• Nov 13: Upgraded again to LM 2025.08, this time with almost all trends removed to rule out load/retention. Result: the reboot pattern persisted—multiple restarts per day despite the minimal trend load.
  
• Nov 15 ~22:00: Rolled back to LM 2024.04.26. Under the same conditions (UPS power, SD verified OK, same project minus cosmetic changes), reboot frequency dropped noticeably and the unit behaved much more stably.
  

These A/B tests suggest the issue is not directly tied to trend volume or SD health, as the reboot rate is significantly higher on 2025.08 even with trends minimized.

[*]We’ve already reviewed forum threads on trend-save issues and spontaneous reboots. Most guidance focuses on power quality, SD integrity, and reducing the number/retention of trends. We’ve addressed all three (UPS supply, SD verified/writable with ample free space, trends drastically reduced), so our case seems unrelated to those common causes.

[*]We would greatly appreciate any advice or collaboration to improve the system’s stability. We can share the full project and/or a complete device backup and run any tests you recommend to help diagnose and resolve the reboot and trend-persistence problems.