pyPowerwall Proxy Stopped (getting more frequent) #506
-
I am still having trouble figuring out why my dashboard stops data collection. (This is a follow on question from #498) Just to recap the installation. It is running on a virtual Linux Mint install. Co-hosted on that VM is my own LAMP install that is querying the Fronius inverters and doing separate weather data collection. I don't think the two installs are clashing, but I cannot prove that conclusively. This is the tail end of the logs from a recent reinstall. The warnings about not being able to collect vitals is expected.
Powerwall seems to work fine for a period, then stops collecting data, shortly after midnight (each time). There is no other clues that I can see, except the very next line of the log file says ...
Running a curl command gives the following output.
Running the verify script produces this result
Whatever the problem is, it seems to be taking down the weather container as well. Re-run the upgrade script and everything starts working again. The verify script produces the following result.
There is one other bit of info. I suspect it is completely unrelated and a red herring for this particular problem. My Security Onion installation (different VM on the same server) is occasionally generating Suricata alerts against outbound traffic from the powerwall VM server to the client PC running the web page. Security Onion is sitting on the LAN with a mirrored traffic feed. Here is an extract of the alert.
This is not happening all of the time. It comes in little batches of about 10 seconds duration. Then just stops. I do think it is a red hearing, but, shrug ... Thoughts as to the next place to look for logs/clues? |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 9 replies
-
This is key. This is caused by pypowerwall getting a KeyboardInterurpt (^C) or SystemExit interrupt. There could be other failure reasons, but in any case I would ask:
If I were you, I would set up another installation of the Powerwall Dashboard on a non-Mint system to see if this is specific to that VM setup. I would look at the Mint system logs to see if it shows anything. Also, confirm that the docker compose shows the correct restart configuration. |
Beta Was this translation helpful? Give feedback.
With the docker and containerd errors in the logs, it does seem like something is unhealthy about the docker setup. You could paste all of those logs in an LLM like ChatGPT to see what it says.
As for re-installing or creating a new VM, my recommendation would be to:
./compose-dash.sh stop
)chown -R myuser Powerwall-Dashboard/*
)setup.sh
- The inflxudb container should mount and use he old database you copied over.