From SpinetiX Support Wiki
- After three reboots in less then 30 minutes, the HMP automatically disables the rendering of the user content and displays a "Safe Mode" text box along with a small blinking square and a slowly changing background color from white to black (to avoid screen burn-in).
- After one hour, the HMP will automatically reboot into normal operating mode in order to check if the problem has been solved; if the content-related problem persists, the HMP will probably end up in Safe mode again.
What to do when the HMP is in Safe mode
When the HMP is in Safe mode, you need to do the following:
- Remove any content that has been added recently on the HMP.
- If the HMP is configured to retrieve content from an external server using Pull Mode, remove the pull actions (or the ics file controlling them). This is to avoid a faulty content being retrieved again.
- Open HMP Control Center and check the logs to find any indications regarding the content at fault (for instance, the HMP reboots after the same amount of time, always after rendering a certain resource etc.).
- If you still don't manage to find the cause and solve the issue, then generate a standard report. Contact your local support (if available, otherwise SpinetiX Support) and provide the report and a copy of the content present on the HMP at the time of Safe mode.
Why the HMP goes into Safe mode
The HMP has a software watchdog that observes the internal processes and if something goes wrong, for instance, no frame has been rendered within 60 seconds (or 30s before 3.1.0 firmware release), it will trigger a device reboot.
If there are three reboots in less than 30 minutes, it's clear that something is not right and some measures must be taken (see above). If the HMP is forced to reboot again, then Pull Mode related actions are also disabled, to avoid a faulty content being retrieved this way.
Examples of content that could trigger Safe mode:
- Content out of the HMP specifications or requiring too much memory.
- Data source too large to be parsed - for instance a calendar containing a lot of expired events (in which case, the solution is to remove those old events).
- Scheduled content that is not protected with
- Synchronized content having items for which the duration cannot be computed.