Problems with Reboot Scheduler - Router Locks Up

[tracker.ca]

We have been experiencing some issues with routers not coming back from a weekly scheduled reboot.

It doesn't happen every time, but consistently every couple of weeks.

So far, this is happening on 7 different routers (RT-AC56U) on 5 different locations (and configurations) with .57 and .58 Firmware.

All of these were either done with a clean, manual configuration or John's save/restore from previous firmware.

We can't troubleshoot because SysLog gets erased when the router locks up.

Anybody else experiencing this issue?

Suggestions on troubleshooting?

 

[L&LD]

Is the scheduled reboot really needed?

Upgrading to a more current firmware is what I would do, first. Along with a minimal and manual configuration (use john9527's excellent utility for the text based (human readable) output this time around).

http://www.snbforums.com/threads/no...l-and-manual-configuration.27115/#post-205573

I have a few customers with an RT-AC56U running their networks and they do not lock up or need regular reboots to perform (latest RMerlin 380.59).

 

[tracker.ca]

Well, the reason for the scheduled reboot was to avoid the random lockups. We have since disabled the scheduled reboot and have most of the routers working well for the past month.

We haven't updated t0 .59 because of the reports of problems with OpenVPN. Besides, there is no indication on the changelog that an issue like this would be fixed.

I really miss the stability of the old versions. I had a couple of N66U running .43 that worked for more than a year without a reboot...

 

[L&LD]

Well, the reason for the scheduled reboot was to avoid the random lockups. We have since disabled the scheduled reboot and have most of the routers working well for the past month.

We haven't updated t0 .59 because of the reports of problems with OpenVPN. Besides, there is no indication on the changelog that an issue like this would be fixed.

I really miss the stability of the old versions. I had a couple of N66U running .43 that worked for more than a year without a reboot...

Using a new function (scheduled reboots) is what seems to have changed the 'stability' for most of the routers? Glad that eliminating that change made most of them stable again.

Changelogs can't (and do not) predict what they'll fix or not, except for a very few specific examples. A new firmware version has many interactions that can only be known to be 'good' or 'bad' by testing them directly and within our own networks and our specific devices.

 

[RMerlin]

Sometimes, I wonder what's going on with some user setups. Here is one of my business customers. They connect almost on a daily basis to its OpenVPN server. It now has 82 days of uptime (which was when I installed it, to replace a dead Cisco ASA), and it also still has plenty of free RAM available (some users report requiring to reboot their router every two weeks due to memory leaks):

 

[RMerlin]

Suggestions on troubleshooting?

See if it also hangs when you do a manual reboot over SSH. I've seen some particular setups that would have one of the services crash during a reboot. I could never reproduce it, and debugging this requires having a serial cable attached while it crashes, so I've never been able to track it down.

 

[tracker.ca]

See if it also hangs when you do a manual reboot over SSH. I've seen some particular setups that would have one of the services crash during a reboot. I could never reproduce it, and debugging this requires having a serial cable attached while it crashes, so I've never been able to track it down.

Thanks RMerlin

I have never experienced a lockup during a manual reboot (GUI or SSH), so I am inclined to think that something is up with the Reboot Scheduler.
I will redirect the Syslog to another server and capture that next time it locks up to see if it is during shutdown.