Powerwall gateway resets... issues with cookieproxy?

[jasonacox]

First of all, thank you for this great Powerwall Monitor dashboard!

I have a Powerwall+ with Backup Gateway. I am running the powerwall_monitor stack and also have an instance of the local network Powerwall web portal to see the Tesla power flow animation on another screen (would love to somehow add that to a dashboard panel).

While running the powerwall_monitor and the portal, after several hours, the portal would start to slow down and crash, appearing to restart. It would be unavailable then would prompt for login. I looked at the cookieproxy code and can't tell if it is creating a new session with each call or caching the cookie for set period of time. But I did notice that it was firing off GET requests at a high frequency. I wondered if a cooke and API call cache (something like 5s) would help with the Powerwall API server stability.

I've been working on a python based API for the Powerwall, creatively called, pypowerwall. I have it cache the session cookie data as well as API calls (5s TTL). I added an ultra simple proxy that I containerized as jasonacox/pypowerall. I replaced cookieproxy with this in powerwall.yml and re-deployed via docker-compose. I've been running it all week without any session restart issues on the portal.

I forked this repo and updated it for pypowerwall instructions in case anyone else is experiencing the same thing: https://github.com/jasonacox/powerwall_monitor - If there is interest, I will submit a pull request that allows users to pick cookieproxy or pypowerwall for their proxy layer.

Thanks again for this great tool!!!

BTW - The cookieproxy project looks great, way more advanced and adaptable, but it seems to be causing problems with my Powerwall. I decided on a replacement that was simple and restricted to basic read-only API calls (for aggregates and soe) which is all telegraf needs.