FOR TIPS, gUIDES & TUTORIALS

subscribe to our Youtube

GO TO YOUTUBE

14455 questions

17168 answers

28195 comments

0 members

We are migrating to our new platform at https://community.teltonika.lt. Moving forward, you can continue discussions on this new platform. This current platform will be temporarily maintained for reference purposes.
+1 vote
236 views 2 comments
by anonymous

Hello,

today I had a global connectivity outage on more than 25 RUT240 devices. Most of them recovered in 15-30 min after this outage except for 3.

Currently I'm investigating the cause of this issue with the networkprovider(M2M,data-only,freeeway). My guess is that there was some issue on the mobile network that caused that outage, what bothers me is that 3 devices didn't recover.

One device recovered after a power cycle. The other 2 weren't power cycled and are still offline. I downloaded a troubleshoot log from that device to inspect the reason for this behaviour.

I'm wondering because:

1) I enabled the ping reboot to 8.8.8.8 on a 5 min interval.

From my understanding a power cycle and a ping triggered reboot do the same to the router: reboot. Why did the power cycle help and not that ping reboot? Is there any possibilty that 8.8.8.8 is pingable but the router can't connect to "the rest of the internet" ?

2) This setup is replicated multiple times

In sum I have more than 30 RUT240 in the field. All updated to at least RUT2_R_00.07.02.4 firmware, all with the same configuration as the attached troubleshoot log, some with EC25 Quetel modem, some with SLM 750 MEIG modem. The configuration is not heavily modified from the factory configuration, just added my wifi, an application script, ping reboot and 24 hour reboot.

There are devices with exact the same setup(Firmware version + modem) that recovered. What could be the reason for these differences?

3) Provider support advice

The support on the providers side suggested "it might be necessary to restart the device to force a full new network session" In my understanding a reboot will do that. Even a software triggered reboot from the ping monitoring. Is there any way to make this more reliable? Or open a new network session automatically if connectivity is bad?
Its always unpleasant to call customers asking them to pull the plug and restart their device to fix connectivity issues - also this isn't a solution on scale.

I really need to fix those issues and harden the routers against any mobile-connectivity hickups. I have a really hard time tracking down these issues, since they occur really rarely and I can't find a good way to debug them. So I also would appreciate any advice in this direction. For now looking at troubleshoot logs after those events is the only way for me to get an idea of whats going wrong. 

Kind regards

Dennis

1 Answer

0 votes
by anonymous
A software "reboot" and a power cycle are not identical. Because the power cycle also "reboots" the modem, which by itself has quite some intelligence. Having an even larger fleet of remote RUT955s, than your RU240s, but running "official" openwrt, therefore I explicitly do a modem reset  on every boot of the RUT. Would be interesting to know, whether Teltonikas RUTOS does something similar.
by anonymous

Hello,

A software "reboot" and a power cycle are not identical. Because the power cycle also "reboots" the modem, which by itself has quite some intelligence.

Can someone from Teltonika please verify that first sentence?
I'm aware that the modem also has firmware and intelligence, so my approach is to restart the whole device to have a fresh and deterministic state after each network outage / 24 hours.

The autoreboot options are quite diverse. The wiki page doesnt give any information what happens on each action in detail.

Especially the differences between "Modem reboot", "(Re)register","Restart mobile connection" and "Device reboot" are interesting in my case, keeping in mind that one troubleshoot solution on the network providers side was "to force a full new network session".

Best regards

Dennis

by anonymous
>  Is there any possibilty that 8.8.8.8 is pingable but the router can't connect to "the rest of the internet" ?

Yes, I have been burned before by a similar issue. dnsmasq had become unresponsive / only able to return REFUSED to any query.

Now I ping dns.google in IPv4 mode not 8.8.8.8.