cant reboot router

[andresmorago]

hello guys

I have a ac3100 running 384.15. router is on a remote location and i currently dont have physical access to it.
i have started to notice that the router is running weird:
*log stopped showing. the last entry is from 2 days ago
*ssh doesnt work at all
*none of the custom webpages show under tools

i can connect to the router via vpn and access the web gui normally. but when i hit the reboot button, the percentage progress shows normally but the router just doesnt reboot at all (i have tried with different web browsers) and it takes me back again to the main page. vpn keeps connected

is there anything else i could do remotely?

thanks so much

 

[martinr]

You say SSH doesn’t work at all, but you didn’t elaborate.

i presume you have SSH enabled (probably LAN only) and have tried to connect via OpenVPN. So what happens when you try to connect via SSH through the VPN; is there an error message? When everything worked previously, could you connect via SSH?

 

[andresmorago]

Hi.
ssh has been always enabled on lan only. Nothing has changed in my setup.

Everything used to work before without any issues as I was able to first connect to the vpn and then connect to ssh through it. Now I get this error (see attached) when trying to connect to ssh from any vpn client.

As I mentioned before, vpn still responds as well as the rest of the router (Wan, gui, pings, Wi-Fi) but I’m just not able to connect to ssh or Reboot it from gui and the custom web pages won’t show.

I have also tried to disable and re enable ssh from GUI. The process itself works and the progress percentage goes all the way to 100% but the ssh connection maintains dead

I cannot connect straight to ssh since I’m not at the routers premises. Everything has to be done through the vpn

 

[martinr]

Out of interest, what’s at 192.168.1.1? (And 192.168.100.1?)

 

[andresmorago]

Out of interest, what’s at 192.168.1.1? (And 192.168.100.1?)

Nothing. They are for a different environment.
This router only responds at 10.0.0.1 and has always done so

 

[martinr]

Nothing. They are for a different environment.
This router only responds at 10.0.0.1 and has always done so

Thanks. Just checking.

 

[martinr]

Have you looked in syslog for any info after, for example, disabling SSH and then re-enabling SSH in the gui?

 

[martinr]

Another thought on the Putty connection error: you can set logging in Putty. That might also give you a clue what the network error is, and whether or not you’ll be able to cure it remotely.

 

[andresmorago]

Have you looked in syslog for any info after, for example, disabling SSH and then re-enabling SSH in the gui?

Nothing. Syslog is dead from 3 days ago. Nothing new will come up

Another thought on the Putty connection error: you can set logging in Putty. That might also give you a clue what the network error is, and whether or not you’ll be able to cure it remotely.

Thanks. I’ll try this but I believe there is something going on with the router that needs to be fixed on premises

 

[L&LD]

@andresmorago, try this sequence to revive the router.

• Safely remove all USB drives/devices from the GUI.
• Check if you can ssh in now (probably not...).
• Reboot via the GUI.

HTH.

 

[kck]

I see this periodically on my AC88Us (2 of them). From what I can tell a process on the router gets into the Linux uninterruptible sleep state (D in a ps listing). Once this happens you won't be able to soft reboot. You'll see what you are seeing which is the reboot countdown but afterwards you're where you were. In some cases you will lose your web access entirely. The only out is a power cycle. For a remote system this sucks (I have 2 homes 1000 miles apart so can't just go fix things.) What I have done is to build an interrupter relay with an ESP-01 that can kill power for 10 seconds and then restore it. It can be triggered from any node on the local network. I also run a small watchdog on a couple of my systems. If they lose internet access for more than a few hours they will trigger the reboot. This covers most of my failure cases since generally the router is still routing when it goes autistic. If I still have VPN access into the remote network when this happens I manually trigger things from one of my remote systems. If internet access dies one of the watchdogs notices and triggers. The only case not covered is if the router goes totally dead and drop WiFi etc. I could improve matters and catch this case if the ESP-o1 itself checked the internet and triggered autonomously but haven't needed that in any case I've seen. Overall, annoying but . . .

 

[andresmorago]

Another thought on the Putty connection error: you can set logging in Putty. That might also give you a clue what the network error is, and whether or not you’ll be able to cure it remotely.

I see this periodically on my AC88Us (2 of them). From what I can tell a process on the router gets into the Linux uninterruptible sleep state (D in a ps listing). Once this happens you won't be able to soft reboot. You'll see what you are seeing which is the reboot countdown but afterwards you're where you were. In some cases you will lose your web access entirely. The only out is a power cycle. For a remote system this sucks (I have 2 homes 1000 miles apart so can't just go fix things.) What I have done is to build an interrupter relay with an ESP-01 that can kill power for 10 seconds and then restore it. It can be triggered from any node on the local network. I also run a small watchdog on a couple of my systems. If they lose internet access for more than a few hours they will trigger the reboot. This covers most of my failure cases since generally the router is still routing when it goes autistic. If I still have VPN access into the remote network when this happens I manually trigger things from one of my remote systems. If internet access dies one of the watchdogs notices and triggers. The only case not covered is if the router goes totally dead and drop WiFi etc. I could improve matters and catch this case if the ESP-o1 itself checked the internet and triggered autonomously but haven't needed that in any case I've seen. Overall, annoying but . . .

Thanks To all for the provided information. It seems that I’ll need to implement some way out that includes a power cycle.
Is there anything I could do script-wise in order to minimize this? It’s been happening more frequently since latest build (I did a dirty upgrade)

 

[L&LD]

Thanks To all for the provided information. It seems that I’ll need to implement some way out that includes a power cycle.
Is there anything I could do script-wise in order to minimize this? It’s been happening more frequently since latest build (I did a dirty upgrade)

Maybe test if a monthly, weekly, bi-weekly, every other day or daily reboot schedule via amtm helps ?

Or, more 'drastic'!

Maybe perform an M&M Config?

 

[andresmorago]

Maybe test if a monthly, weekly, bi-weekly, every other day or daily reboot schedule via amtm helps ?

Or, more 'drastic'!

Maybe perform an M&M Config?

thanks!