Any way to automatically reboot/restart modem after loss of net?

[FORANE]

Followup on this old issue of modem dropping offline then staying off, it is now resolved.
Not the root cause of it dropping offline, but a workaround that has proven 100% effective so far.
Installed one of these:
https://www.amazon.com/dp/B0792S1DGZ/ref=cm_sw_em_r_mt_dp_U_GVJFDbYED62PD
Occasionally the modem will go down transiently, and will be well into the reboot cycle by the time I go to check on it.
Would be nice to know the cause, but at least there is a workaround.

 

[flyingron]

Back when Embarq/CenturyLink/Sprint/Whatever was being really crappy, I actually put my modem on a cheapo house timer to just power cycle it every night.

Eventually, they replaced the bargain basement DSLAM so I could get rid of that. The next moronic thing was it took a formal FCC complaint to get someone who actually had a clue to respond with a software bug that was in their OEM'd router. Nobody at the manufacturer would talk to be because the router was OEM'd to embarq.

ME: Hi, my modem goes into a funny state when it's rebooted that causes it to drop all packets for 3 seconds ever 40.
TECH: Let's start by rebooting your mode.
ME: That's what causes it, it's not going to fix it.

Eventually, I get a call from the president's office in response to my complaint which has some guy Paul (they would not give me any contact information) call em with the caller id of the main support number to walk me through getting around the bugs in the autonegotiation software.

I always loved this XKCD cartoon:

 

[Let'sgoflying!]

"My computer has nothing to do with it"
That's hilarious to me at this moment because for many months we had a modem that would drop off at work, several times a day (I put in a remote power switch to reboot it as a quick 'fix' and that was going so well we forgot about further troubleshooting - until...)

...2 weeks ago, one of the workstations started dying at 4:06 every afternoon.
I popped in a new one and..
Not once have we had to reset that modem.

 

[Spring Ford]

...2 weeks ago, one of the workstations started dying at 4:06 every afternoon.
I popped in a new one and..
Not once have we had to reset that modem.

I can propose a failure model for that, if it was a router and not a dumb modem - no hard evidence here of course but I quite like it.

Assuming -
You use the router for DHCP and retain the often used default setting of using the router as the DNS server.

The faulty PC has malicious code on it that was making excessive DNS requests which in turn caused the router to run out of memory or other resource.

Or some other possibly deliberate attack on the router, as opposed to accidentally running it out of resource.

 

[denverpilot]

I can propose a failure model for that, if it was a router and not a dumb modem - no hard evidence here of course but I quite like it.

Assuming -
You use the router for DHCP and retain the often used default setting of using the router as the DNS server.

The faulty PC has malicious code on it that was making excessive DNS requests which in turn caused the router to run out of memory or other resource.

Or some other possibly deliberate attack on the router, as opposed to accidentally running it out of resource.

Awww you missed the really “fun” possibility that actually happened to my company once long long ago....

Leave routers set inadvertently to accept RIP routing updates from anything on the network.

User machine decides it’s the router of choice for the whole world every so often.

RIP storm ensues with every router in the company arguing over who has a route to the Internet and each other.

Ahhh RIP. Such a lovely protocol. LOL.

Of course a different company repeated the stupidity with OSPF and then again later with whatever Cisco’s proprietary routing protocol du jour was — and those hit whole data centers, one with paying customers in it...