No Access to Frontend via IP Address, Only SSH, IP ping Successful

Several days ago I had a prolonged power outage. After, I was no longer able to access HA via frontend IP address using either the internal IP, hassio.local, duckdns address, or HA phone app. I assumed a corrupt SD on Raspberry Pi 3 B+ and restored a backup successfully, but it did not restore webpage function. I am able to SSH to the device and an IP ping is successful. I can view and edit config files via Samba. Here is my HTTP config (previously working).
Screenshot (266)

Ha core check shows intact config file. Attempted on multiple browsers (Firefox, Edge, etc). No success.

I checked the IP provider gateway and the public IP address is unchanged. All port forwarding is unaltered. Console command ā€˜ha dns infoā€™ shows proper host dns address. Could there be a problem with host address, since it displays ā€¦32.3 vs 32.1?
Screenshot (262)

When I initiate console command ā€˜ha core restartā€™, the following error appears in the log:
TypeError: _update() got an unexpected keyword argument ā€˜external_urlā€™

Does anyone have any suggestions on how I can regain access to the frontend? I have no control over any devices with this issue. Please and THANK YOU!!!

I also checked that the UTC time was correct using the following steps:

Connect Raspberry Pi with HDMI monitor and USB keyboard
Login with root
Type ā€œloginā€ and it will show prompt#
Type ā€œdateā€

Could it be possible that some certification has expired? How would I verify?

I also deleted the IP browsers cache without resolution. In the Windows command prompt, nslookup homeassistant.local returns

Screenshot (269)

Neither of the IP addresses listed are my local IP, VPN, or any other association.

can you verify if thereā€™s a service up on port 8123?

nmap -sV -p 8123 172.30.32.3

Your nslookup seems to return a wrong ip, you could try and add it to your hosts file just for testing or temporary workaround

EDIT: you should not add protocol when using nslookup > do it without https://

To test for invalid cert you could use curl, it has a parameter to ignore invalid ssl. Curl is not in windows by default but you can find it, same goes for nmap

Thank you for the suggestions Kurt! The nmap command returned the following:
Screenshot (276)

The assigned internal IP ā€¦50.222 returned a reply, but the listed dns host address does not. TCP closed??? I donā€™t know why the listed host address would even be 172.30ā€¦, it is not an address range assigned by my router DHCP range. Do you think this maybe the issue. When HA was working, I remember seeing occasional 172ā€¦ addresses listed in my logs when there were failed login attempts internally, so assumed it was preassigned/default addresses with ha core/supervisor coding. I tried accessing HA via the 172 addresses as well without success.

172.30.32.1:8123, 172.30.32.2:8123, 172.30.32.3:8123

Once again, thank you for the feedback and any further suggestions!

2 different issues here:

DNS: your dns lookup is wrong > you should not do nslookup https://IP but only ever IP. Iā€™m not going into that, your HA is not on an external IP

So thatā€™s not the root cause hereā€¦

An nmap service scan on IP (-sV) returns no running services > so the service is not runningā€¦ no point in searching for a cert error for a service that is not running then

A quick re-read of your problem says an error with external_url. Probably service is not starting because of that.

If you still have access through smb to your config, make a backup and make a new one, one without the cert and !secret. If no access through smb, do it through ssh, or even console. Just a simple config just to get back onto your web interface

Thanks Kurt!

Returns the same foreign IP address. Not my network or public IP. Not even same region. I researched with IPlookup and donā€™t know whoā€™s address this is or why its there?

Screenshot (278)

How did you come to this conclusion? The internal IP (192.168.50.222) that I usually connect to the UI and SSH does return a reply: 1 IP address (1 host up). Are you saying that because the TCP status is closed, that there are no services running?

Screenshot (279)

I commented out all of the configuration setting in HTTP: and restarted the devices without any resolution. The problem still exist. I remember reading that there was a configuration change with the most recent updates, where the new settings would be external_url and internal_url. Is it possible that the code from the newest version was retained when I rolled back to the older 108.1 that has the backup? With that said, this same problem existed before I loaded a fresh install with newest version, hence the reason I even restored & wiped from backup.

Once more thanks!

Not really willing to go into your network setup but the IP belongs to Akamai btw:

NetRange: 23.192.0.0 - 23.223.255.255
CIDR: 23.192.0.0/11
NetName: AKAMAI

I got the IP from the info you posted: HA DNS INFO 172.30.32.3

172.30 seems the IP from one of the docker containersā€¦

Just for information, what is your local ip from the computer your are testing this from? For example if my local Ip is 192.168.0.x/24 The HA is on same LAN.

Just do a ping sweep on your pc (nmap -sP 192.168.0.1/24) replace with your IP to see hosts up on your networkā€¦ then service scan on 8123 to see if service is up.

Sorry gonna have to leave now will check up on thread later

Screenshot (279)

Service is up.

Sorry but no, it is closed

Again, if I were you I would backup my yaml, and make a new one, without !secret and certs

really gotta go now, good luck!

If you did a rollback from 0.111.x to 0.108.x, look in .storage, probably external and internal URL are still in the json.

@francisp Thanks for the reply!

I can not seem to locate any json files in .storage with external/internal URL? Which file would it be located?

The http file has the following settings:

Screenshot (281)

Would it benefit me to manually assign the internal IP address here, turn off IP ban and disable SSL? The router has been set to assign a fixed internal IP to HA. I just want local access at this point and could care less if itā€™s encrypted. Is there a way to remove LetsEncrypt and DuckDNS settings/addon from console? Just in case it maybe causing a conflict. Thanks!!!

Thanks Kurt!

1 Like

Another observation, my backup .storage file has much more data that the restored from snapshot .storage file. I would think that all information would be transferred, such as lovelace, mobile_app, hacs etc, but this clearly is not the case.

Would I be better served coping all the data files from the backup .storage to the current .storage file? Are there any settings in this directory that could crash SSH and console connectivity? That is my only access at this time.

@francisp You are brilliant!!! Thank you, thank you, thank you! I removed the external/internal data from core file and was able to perform a ha core restart in console without error this time. I scoured the forum for frontend loading issues, but did not think to check for HA Docker container issues???

Screenshot (288)

The UI loaded after reboot, albeit in a very unorganized/not restored layout, but at least I can access it again via local IP for further troubleshooting. Iā€™m back on the right track thanks to your input!

Also, thanks for pointing me in the direction of the .storage file. For whatever reason, the restored snapshot did not copy all of these files and thus the reason why the UI layout was scattered. I was able to transfer the missing files manually, which saved me tons of time having to reorganize.

Once more I canā€™t thank you enough for the feedback!!! This is issue is solved. Cheers!