GHLocalApi
GHLocalApi copied to clipboard
port 8443 seems to be filtered on latest fuchsia build
Nest Hub on latest Fuchsia build:
Nmap scan report for 10.10.10.12
Host is up (0.0071s latency).
Not shown: 65527 closed ports
PORT STATE SERVICE
22/tcp filtered ssh
8008/tcp open http
8009/tcp open ajp13
8012/tcp open unknown
8443/tcp filtered https-alt
9000/tcp open cslistener
10001/tcp open scp-config
10101/tcp open ezmeeting-2
MAC Address: 1C:F2:9A:33:0E:9C (Google)
after downgrading to stable build:
Nmap scan report for 10.10.10.12
Host is up (0.0074s latency).
Not shown: 65527 closed ports
PORT STATE SERVICE
22/tcp filtered ssh
8008/tcp open http
8009/tcp open ajp13
8012/tcp open unknown
8443/tcp open https-alt
9000/tcp open cslistener
10001/tcp open scp-config
10101/tcp open ezmeeting-2
MAC Address: 1C:F2:9A:33:0E:9C (Google)
Hey thanks for posting this! Does the API still work on 8443? Filtered doesn't necessarily mean closed. Maybe they've added some extra checks. I don't have a nest hub to verify.
My og Google Home still has it open:
Nmap scan report for 192.168.0.16
Host is up (0.012s latency).
Not shown: 995 closed ports
PORT STATE SERVICE VERSION
8008/tcp open http?
8009/tcp open ssl/ajp13?
8443/tcp open ssl/https-alt?
9000/tcp open ssl/cslistener?
10001/tcp open ssl/scp-config?
Does the API still work on 8443? Filtered doesn't necessarily mean closed.
doesn't work for me anymore, it times out
Update:
it seems to be a network timeout rather than a filtered port, I dug a bit deeper and realized that my api calls ran often times, but took very long to complete. see below comparison of an android and fuchsia device. It took the fuchsia device 2 minutes to reply. I cannot tell if thats an issue with fuchsia or my network. I think we can close the case for now.
phixion@cray:~# echo -e '\nandroid device\n' && time curl https://10.10.10.12:8443/setup/eureka_info --insecure -s | jq -r .cast_build_revision
android device
1.52.260996
real 0m0.253s
user 0m0.042s
sys 0m0.004s
phixion@cray:~# echo -e '\nfuchsia device\n' && time curl https://10.10.10.13:8443/setup/eureka_info --insecure -s | jq -r .cast_build_revision
fuchsia device
1.54.250118
real 2m9.426s
user 0m0.006s
sys 0m0.005s
2 minutes?! Have no clue why that could happen. Does the Google Home app work normally (without delays) when changing these settings?
It works fine for a while after powercycling the device, it seems the device is throttling my requests after some time.
To give some background: I use your work to connect a BT speaker to my Nesthub when I enter my flat. It seems that the software I use for homeautomation is trying to fetch the status of the BT device too often which either clogs the wifi network or the device itself, since it was working fine before I suspect Fuchsia to be the reason.
Hm interesting, that could be it. Think I'll try scripting repeated calls and counting until it slows down (if it does on a Google Home).
I've found out that the home automation checks the state of the switch every 30 seconds. So it ran this every 30 seconds:
#!/bin/sh
curl --insecure -s -H "cast-local-authorization-token: $LOCAL_AUTH_TOKEN" \
https://10.10.10.12:8443/setup/bluetooth/get_bonded | jq '.[] | select(.name == "My cool soundbar") | {connected} | .[]' > switch_state.log;
if grep 'true' switch_state.log
then
exit 0
else
exit 1
fi
Checking the state only every 5th minute instantly solved my issue.
When it's throttled you also lose lots of functionality in the google Home app. You can see that the device is online, however you cannot start casting to it, change settings or reboot it