Closed Pubec closed 3 months ago
@Pubec, while investigation is in progress, I'd like to ask a few questions:
In common situations, you don't need to run bash bin/upgrade_containers.sh
, unless the install script instructed you to do so.
@Pubec, I tried to reproduced the issue but didn't replicated it. I answered y
to all questions except the question that asks if I'd like to install the experimental version of Anthias. master
is at 6b8b0d0
.
I also didn't get an HTTP 504 from the settings page in experimental@84bd4ab
.
To answer the questions:
Logs:
docker logs screenly-anthias-nginx-1
produces no logsdocker logs screenly-antihas-server-1
produces:
[2024-08-07 07:38:30 +0200] [20] [INFO] Starting gunicorn 22.0.0
[2024-08-07 07:38:30 +0200] [20] [INFO] Listening at: http://0.0.0.0:8080 (20)
[2024-08-07 07:38:30 +0200] [20] [INFO] Using worker: gthread
[2024-08-07 07:38:30 +0200] [22] [INFO] Booting worker with pid: 22
host_agent_service is not ready after 60 retries
host_agent_service is not ready after 60 retries
host_agent_service is not ready after 60 retries
host_agent_service is not ready after 60 retries
docker logs screenly-anthias-websocker-1
produces no logsThanks, @Pubec. With your logs, I already found the root cause: host_agent_service is not ready after 60 retries
usually happens when retrieval of the device's IP address failed. I'll let you know when it's fixed.
In the meantime, I'd like to request specific logs from the system journal. You can do so by running sudo journalctl -fu anthias-host-agent
, specifically during the time that the browser attempts to load the settings page.
Would you also be able to provide a screenshot of your browser at http://<ip_address>/splash-page
?
Thanks @nicomiguelino . I've noticed the same error, and I've tried checking the journalctl -fu anthias-host-agent
, but that provides me zero output, like there was nothing going on, even though anthias is running and working.
I can not get to /splash-page
, but I can to main menu:
@Pubec, I recently created #2014 in an attempt to fix the issue. I'll keep you posted. Do you get a white screen or something on /splash-page
?
@nicomiguelino I get a 504 Gateway Time-out aswell as at /settings
Could it possibly be a network manager issue, as I used nmtui
to set fixed ip, with 23-bit netmask.
@Pubec, I added a snippet in the settings view/controller that gets the IP. I'm not sure if that's a Network Manager issue. Do you get similar issues if the IP addresses are not fixed?
@nicomiguelino should I checkout to to nicomiguelino:fix-2013
branch and try to rebuild containers?
I can try removing connection from Network Manager with fixed IP.
@Pubec, no not yet. I still have to have it approved and merged. You can try disabling fixed IP first.
@nicomiguelino I've tried with fixed IP removal, and rebooted, but the problem persists. Only /
and /system-info
is accessible.
@Pubec, I'll see what I can do. I an make some changes, just for you to test if out quickly as soon as possible, even if it's still not merged. I'll give you the command later, just make sure to back up your data first before proceeding.
@Pubec, once you've backed up your data, you can run the following to install a version of Anthias with my patches:
bash <(curl -sL https://raw.githubusercontent.com/nicomiguelino/Anthias/test-fix-2013/bin/install.sh)
Do take note that this is only temporary. If this went well, I'll have the pull request merge, then I'll let you know when changes to master
is ready.
Whilst installing I did notice an error:
E: Conflicting values set for option Signed-By regarding source https://download.docker.com/linux/debian/ bookworm: /etc/apt/keyrings/docker.asc !=
E: The list of sources could not be read
Will get back with new updates as soon as docker pull finishes
Update finished, with error persisting: docker logs screenly-anthias-server-1
produces host_agent_service is not ready after 60 retries
@Pubec, you've said earlier that you got nothing when you ran journalctl -fu anthias-host-agent
. It could be that Anthias is not properly and fully installed in your device.
You can also check if anthias-host-agent.service
(or similar) is present inside the /etc/systemd/system
directory.
If it's not too much on your end, can you do the installation from scratch and take note of the console output from start to finish (until the part where the script prompts you to reboot now or later). At that point, don't reboot your Pi yet, as you still need to get those console logs.
@nicomiguelino, unfortunately, there is nothing similar in /etc/systemd/system
directory.
For a fresh install, should I reinstall Raspberry Pi OS as well?
@Pubec, yes it's recommended to reinstall Raspberry Pi OS at this point.
Here's a screenshot of my terminal (that shows the contents of /etc/systemd/system
) for reference:
With fresh install everyting works as expected, with anthias-host-agent.service
located in /etc/systemd/system
. I guess an installation failure.
@Pubec. I'll be closing this for now. If the issue occurs again, please re-open this and include logs where the failure happened, for instance, in the installation script.
Overview of the Issue
Opening
settings
page produces 504 Gateway Time-out.Reproduction Steps
$ bash <(curl -sL https://install-anthias.srly.io)
bash bin/upgrade_containers.sh
http://<ip>/settings
Environment