Jump to content

Local IP failure


Recommended Posts

Really strange thing happened today .

Bear in mind I hadn’t been in IT mode today ….

 

Firstly my chrome remote login to my HA server wouldn’t work I.e wasn’t online .

I rebooted HA even though it’s headless but no joy .

Tried logging into ip cameras . Browser just hangs - no error .

So reboot hub ; reboot mesh network .

Still no joy .

Reboot netgear poe switches .

 

Still nothing !

 

Can browse web pages no issue I.e wan - but nothing local .

 

Tried WiFi just to hub ( no mesh ) same issue .

 

Completely stumped !

No idea what else to check / do .

 

20 minutes later ( through  no effort of my own ) all started working !! 🙄🙄🙄

Edited by pocster
Link to comment
Share on other sites

21 minutes ago, joth said:

Do you use IP addresses or mDNS or something else for local name resolution? Most often it's name lookups that screw up my local LAN connectivity. Unifi has its odd hiccup with this too

 

Just IP 

Link to comment
Share on other sites

Could it be DHCP lease expiry? I'll swear some of my stuff loses it's IP address every so often and takes a while to renegotiate. Similar experience to yours of stuff being unable to connect over the LAN while WAN unaffected. I seem to recall I could ping them on their IP address though which doesn't make sense.

  • Like 2
Link to comment
Share on other sites

11 hours ago, Radian said:

Could it be DHCP lease expiry? I'll swear some of my stuff loses it's IP address every so often and takes a while to renegotiate. Similar experience to yours of stuff being unable to connect over the LAN while WAN unaffected. I seem to recall I could ping them on their IP address though which doesn't make sense.

Maybe . But like you I could ping everything ….

Link to comment
Share on other sites

42 minutes ago, SteamyTea said:

I thought @pocster was using MAC addresses, but was having trouble.

MAC address is assigned an IP address . Some things are dchp so get allocated ; others static .

All very strange that *everything* failed . I’ll put it down to a hub bug … 🙄

Link to comment
Share on other sites

57 minutes ago, pocster said:

Maybe . But like you I could ping everything ….

 

Yes, this must be a big clue. I literally had this problem last Saturday. The Headless Raspberry Pi in my equipment cupboard hosting MQTT and node-red lost touch with everything on the LAN. It's wired on Ethernet so normally bullet-proof. It was listed as a device in the connected devices tab of the router control panel and I could ping its IP no problem. But neither HTTP on port 80, MQTT on port 1880 or SSH on port 22 would respond. Regular HTTP POSTs coming from a Python script to a cloud based MYSQL server were still ongoing so like everything else, WAN access was OK for the Pi, but not LAN.

 

The issue persisted over a complete reboot of router and satellites and after about 30 minutes of faffing around, I was all set to haul a monitor and keyboard over to the Headless Pi when one last attempt at SSH'ing in succeeded. I have no idea if the reboot had helped restore things (after a silly long delay) or if it would have recovered anyway.

 

My main router is a Netgear Orbi RBR50 with three additional satellites. Of course this is different to your setup but all these things are based on the same open-source code running on Linux so it could well be the result of a bug in a common library.

Link to comment
Share on other sites

15 minutes ago, Radian said:

Should have added that other things on the LAN such as Rpi security cameras, and a Z-wave gateway were all unreachable as well.

It was strange for me for other reasons:

 

Every raspberry pi "didn't exist" i.e. couldn't list their IP's nor could I ping them ( everything else existed though ).

I haven't rebooted any pi's before or after the event.

 

Blu Iris must of been running as it was triggering lighting which means the ip camera must be working and indeed homeseer z wave was working to turn the light on.

 

So it was like everything is working but you can't access anything.

 

Don't like these situation especially when magic fixes them. Have ordered a portable HDMI screen so at least if it happens again I can plug that into the HA pc as it's headless. If that ever goes down ( it hasn't; but windows always decides to do an update at some point ) then we have no lights, no heating, no audio etc. etc. SWMBO approval factor then drops rapidly!!

Link to comment
Share on other sites

1 hour ago, Radian said:

It was listed as a device in the connected devices tab of the router control panel and I could ping its IP no problem. But neither HTTP on port 80, MQTT on port 1880 or SSH on port 22 would respond. Regular HTTP POSTs coming from a Python script to a cloud based MYSQL server were still ongoing so like everything else, WAN access was OK for the Pi, but not LAN.

This sounds like DoS / DDoS protection is kicking in. Is there any logs in the router about potential threats? Did you have a failing script/process somewhere that was rapidly reattempting TCP connections or something?

 

 

Link to comment
Share on other sites

  • 1 month later...

Ha! It happened again just now. All my raspberry pi's were unreachable other than by ping and had stopped logging out to a remote database just like the last time - Saturday 6th Novemeber. Rebooted the router and about half an hour later it was all working fine again. Obviously doesn't like Saturday's.

  • Like 1
Link to comment
Share on other sites

10 minutes ago, Radian said:

Ha! It happened again just now. All my raspberry pi's were unreachable other than by ping and had stopped logging out to a remote database just like the last time - Saturday 6th Novemeber. Rebooted the router and about half an hour later it was all working fine again. Obviously doesn't like Saturday's.

It is the Chinese, or Russians.  Or MI5/6, NSA.

Or could be that they cost a tenner.

Link to comment
Share on other sites

31 minutes ago, Radian said:

Ha! It happened again just now. All my raspberry pi's were unreachable other than by ping and had stopped logging out to a remote database just like the last time - Saturday 6th Novemeber. Rebooted the router and about half an hour later it was all working fine again. Obviously doesn't like Saturday's.

What about looking at a monthly pattern (well it is close), did it also happen on the 6th December but you not notice it, or could it be the router leases timing out / not renewing somehow - there seems to be quite a a bit, no pun intended, on this on the RPi forums.

Link to comment
Share on other sites

23 minutes ago, JohnMo said:

Smart stuff, great until...

 

You're sounding like my millennial Son. Funny how us old fogies are more into the hi-tech scene 😁

 

I jest. It'a a valid view - unnecessary complexity is not a good thing. Anything critical to the smooth running of a household has to be self-contained in my opinion. As soon as you reach beyond your walls for services you're at the mercy of others so I'm quite envious of off-griders. The best I can afford to do is build my own smart services for lighting, heating, security etc. and not buy into the commercial offerings for these things.

 

it's also the reason I prefer, if practical, to code a microcontroller in preference to using a SBC like the Pi for simple tasks as even maintaining a Pi has complexities whereas the code running on a microcontroller is ridiculously bulletproof. I think this is one small miracle that's often overlooked - I've still got PIC chips that are running code that I flashed into them in the early 90's reliably performing the same functions for over 30 years now. Even stupid mechanical things will fall over on these timescales.

  • Like 1
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...