PVOutput stopped working today, corresponding with maintaince on their side

My PVOutput uploader stopped working today. I looked on their site and found this planned maintenance, which lines up exactly with when my upload stopped (9AM EDT or 13:00 UTC). The reason they give on the thread is “Upgrades to the HTTPs certificate and load balancer server.”

This is what my IotaWatt logs show, after a restart:

5/03/26 22:28:42 PVoutput: Unrecognized HTTP completion, getSystemService <html>
<head><title>502 Bad Gateway</ti
5/03/26 22:29:42 PVoutput: Unrecognized HTTP completion, getSystemService <html>
<head><title>502 Bad Gateway</ti
5/03/26 22:30:42 PVoutput: Unrecognized HTTP completion, getSystemService <html>
<head><title>502 Bad Gateway</ti

I did see this thread here from 2019, but a restart didn’t seem to fix it for me. At least a reset from the UI, I can power cycle it later and see if that makes any difference.

This was a tricky issue. Thanks to @DaleSwanson for pointing out the planned outage, which was key to getting to the root of the problem.

First, this issue only appears to affect users that are using the IoTaWatt HTTPSproxy feature to forward IoTaWatt HTTP requests to HTTP using NGINX. Unsecure HTTP access appears to be working fine, which is probably why there was no pile-on to this post. Fortunately, I run my home IoTaWatt using the HTTPS proxy so I was able to see the issue firsthand as my uploader stopped at the same time. Special thanks to chatGPT for helping to diagnose the issue and saving me a ton of time getting back up to speed on NGINX to remedy the situation.

What appears to have happened is PVoutput has put into place a layer in their stack that requires the name of the service requested (pvoutut.org) to be sent in a header with the proxy request. Absent that, the pvoutput server doesn’t seem to know what certificate to present to NGINX in order to establish a secure connection.

I’ve added instructions in the HTTPSproxy configuration available on the IoTaWatt.com website. Users will need to update their /etc/nginx/sites-enabled/HTTPSproxy.

sudo wget -P /etc/nginx/sites-enabled https://iotawatt.com/download/HTTPSproxy

sudo nginx -s reload

@DaleSwanson let me know if this works for you.

1 Like

@overeasy Thanks for the fast response, this did work for me, although a couple notes.

First, when I just ran sudo wget -P /etc/nginx/sites-enabled https://iotawatt.com/download/HTTPSproxy it created a second file /etc/nginx/sites-enabled/HTTPSproxy.1, and then when I ran sudo nginx -s reload I got this error nginx: [emerg] a duplicate default server for 0.0.0.0:9000 in /etc/nginx/sites-enabled/HTTPSproxy.1:26 because there were two files. I just deleted the old file (and the new one with the wrong name) and reran the wget command. sudo rm /etc/nginx/sites-enabled/HTTPSproxy

Second, after that I had to go into the IoTaWatt UI status section and click Start next to PVoutput, because it was stopped. This one though, I think might have just been on me, as I was trying to start and stop it to debug it last night.

Thanks again for the quick fix.

1 Like