H series upgrade to 1.39 fails when memory usage is high

Hi @GiuseppeR ,

There's a directory used to store system cache and temporary files. When the firewall downloads firmware from Nebula, the firmware file is also temporarily stored in this directory. If there isn't enough available space in this directory, the system displays a notification recommending that the user reboot the firewall to clear the temporary cache and free up space.

In your case, it appears that cache or temporary files had accumulated in this directory, leaving insufficient space for the firmware download and generating the log entry as "[Nebula] Automatic firmware update was unsuccessful. Please reboot the device first, then do the firmware update."

Hi @GiuseppeR ,

Regarding your suggestion to schedule an automatic reboot with a notification when a manual or automatic update is triggered from Nebula and the device's memory usage is above a certain threshold, we think an automatic reboot may cause confusion for users. For example, the user may assume the firmware upgrade has already started / interpret the reboot as an upgrade failure or an unexpected system restart.

Therefore, we believe the current behavior is safer and clearer. When the system detects insufficient resources or temporary storage for the firmware download, it generates an event log asking the user to reboot the device first. After the reboot, the system releases temporary resources and cleans up temporary data, and the user can then retry the firmware upgrade.

For this reason, we currently prefer to keep the manual reboot behavior.