So I upgraded to Electric Eel yesterday and everything went very smoothly after waiting out some Docker image pull timeouts. I then started experiencing some very weird behavior throughout my network.
Everything worked great for about an hour and then my internet specifically would drop throughout the entire house. I could still connect locally to everything on the network, but the external internet connection was dead. I connected directly to my router and I had internet access so it wasn’t that. But the mesh network connected to it loses all internet connection. Rebooting fixes the issue but again, only for an hour. The only thing I’ve been able to do to retain internet is to disconnect my truenas system completely from the network which obviously isn’t ideal.
Has anybody experienced this since the update? The most frustrating thing is that I’m not seeing anything weird in the logs of my mesh network or the server. Any thoughts?
Sorry that I didn’t add any actual system data, I had to rush out the house to get to work. Here’s what I have:
• TrueNAS Version: Scale 24.10
• Motherboard: ASRock Z690 Extreme
• RAM Qty: 2x 16 GB DDR4 3200
• CPU Make/Model: Intel i5-12600k
• On-Board-NIC: 1 x 2.5 Gigabit LAN 10/100/1000/2500 Mb/s (Dragon RTL8125BG) *This only runs at 1 Gigabit but that’s all I need and worked prior to the update
The ethernet cable to the system is unplugged and my wife confirms the internet is staying connected. Any idea where to look? I imagine something about how docker handles the container networking tanking internet connectivity.
VPN weirdness is the first thing I thought of. I have tailscale running but I turned it off just to check with no luck. I also have a cloudflare tunnel running. I didn’t think to turn that off yet. With everything working for about an hour before tanking, it’s taking a while to work through without changing too many things at once.
I looked at my ARP tables and didn’t see a conflict immediately but I’ll look closer when I get back home.
I’m thinking the same. I run pfblockerng on my router, so I’ll triple check. I do know that Home Assistant changed it’s socket after the migration to docker so perhaps something else was changed that I’m missing.
So I’ve been banging my head against the wall trying to figure this out. What I’ve tried so far.
Disabled cloudlfare tunnel
Disabled tailscale
Gave static IPs to my Deco X20 mesh satellites
Changed DNS (from quad9 to 8.8.8.8)
Reset the X20 satellites
Nothing has worked for more than 2ish hours. Now the last thing I noticed just now is that both the name of my server and “docker” appear in my wifi list when I have my server plugged in. Both disappear when I unplug the server. When i look in the arp table of my pfsense router, I only see the name of my server. Could this be the cause of my issues? The server has a static IP for about year now so I wonder if the docker containers are messing with that.
Finally got some time to troubleshoot and after doing a lot, turns out it’s Home Assistant. I looked everywhere else since it was working fine before EE but deactivating all apps then reactivating one by one showed me that HA was the culprit for my network issues.
No idea how to fix that but since it’s not a Truenas issue, I’ll mark this as resolved. Thanks for the help!
I got very same issue. All network went down after installing electric eel. Reinstalled as fresh install and no HOAC and still the same issue. I had to reinstall previous version Dragonfish to keep using truenas scale. Electric eel is a nightmare product so far! The other issue was inability to have individual IPs for apps which is a great feature of Dragonfish.
