Question

I have many vehicles with a RUT955 (2nd gen model) in them. My problem is that some of them go offline and I dont know how to make the device recover from it.

To detect what was going on, my team and I brought those vehicles in and carefully inspected the devices.  Here are the problems:

1. device did not display the Power LED. After testing the wiring, we saw that there was 12v on the cable.
2. device display Power LED & green LED for mobile, but no LED for signal strength.
3. device display Power LED & flashing (green?/red?) LED for mobile, but no LED for signal strength.

The only way to get them back online is to remove the power plug & re-insert it. After 2 minutes, they are operating as if nothing is wrong. They will operate normally for weeks & may or may not repeat the fault. I cannot say for sure that there is particular device that is always a problem. Some devices only had this fault once & never repeated it.

I have looked at the log.db file and it does not tell me if there was a problem or not. It just stops recording & then records appear when they are disconnected/reconnected. There is no syslogd. dmesg doesnt tell me anything useful either.

There is a record on crontab to reboot the router at 4am everyday and that has improved reliability. But I still get problem devices.

My questions are:

• How do I prevent them from going offline and staying offline?
• Do you think something happened to the OS that made crontab fail?
• Is there a log or another way to detect that there is a fault & reboot?
• Is there a known problem with the hardware power that I need to fix?

Thanks.

Comments

I have the same exact problem with 15x RUT955s in vehicles, all running on the latest test firmware (6.05.20) because any other version breaks gpsd even more.

Eventually, after repeated gpsd restarts and auto-ping reboots, the devices will no longer come back to life even after re-plugging the power cable.

I'm glad I'm not the only one experiencing these unfortunate issues.

Answer 1

Hello,

How do I prevent them from going offline and staying offline?

Depends on what "offline" means in this context. Based on your description, looks like cases (2) and (3) may have simply lost their mobile connection. This can happen, especially when the device is not stationary. A way to prevent this is using Ping Reboot. It can be set up in the Services → Auto Reboot → Ping Reboot page.

But it wouldn't help to solve case (1), as it seems like different issue. 

Do you think something happened to the OS that made crontab fail?

If we're talking about case (1), then probably not. Looks like it was a powering issue and nothing to do with the OS.

Is there a log or another way to detect that there is a fault & reboot?

Yes, there are ways. I would recommend setting up Events Reporting (Status → Events Log → Events Reporting) and Reporting Configuration (Status → Events Log → Reporting Configuration).

The first one can be used to inform you via Email or SMS of any failures. The second one can be set up to periodically sends router logs o FTP, Email or syslog server. The logs are deleted after a reboot, that's why you couldn't find anything useful. This might ensure that some helpful information will be preserved.

Also, you can make the router save logs in it's flash memory instead of RAM, which would preserve them even after a reboot. This can be set up in the System → Administration → Troubleshoot page. However, the flash space is very limited.

Is there a known problem with the hardware power that I need to fix?

There is no known problem. There could some issue with the specific device or devices. Unfortunately, this is a bit harder to troubleshoot. Is it the one device that lost power (LED was not lit up) or are there more like this? How often does it happen? 

Comments

I accept that the device will lose signal from time to time during the day. Coverage isnt great. So the pingreboot might be more reboots that necessary.

Will try it on 1 device and if its better, will roll out to all devices.

Same goes for the logs.

Is it the one device that lost power (LED was not lit up) or are there more like this?

Yes. So far just one & once. More time needed to collect data.

Also, we've swapped units on problem 1 & 2. So in future, if there is a problem we can rule out bad wiring / faulty device. So far they are fine.