figure out why the relay server is still posting to old url

[williamcodes]

Once I set the heatseeknyc.com domain to forward to heatseek.org, the readings stop coming through. I tried restarting the batch process and restarting the entire relay server. Baffling. Thoughts @hrldcpr ?

[hrldcpr]

Did you rebuild the docker image? I'm afk right now, but there should be instructions in the relay readme. On May 12, 2016 5:51 PM, "William Jeffries" notifications@github.com wrote:

Once I set the heatseeknyc.com domain to forward to heatseek.org, the readings stop coming through. I tried restarting the batch process and restarting the entire relay server. Baffling. Thoughts @hrldcpr https://github.com/hrldcpr ?

— You are receiving this because you were mentioned. Reply to this email directly or view it on GitHub https://github.com/heatseeknyc/relay/issues/27

[williamcodes]

No, that's probably what I was missing. This is what the readme has:

$ docker build --pull --tag app app
$ sudo systemctl restart app

So I'm trying this:

$ docker build --pull --tag batch batch

That doesn't seem to be working. Looks like the --pull flag gets the latest version of the image, so that makes sense. the --tag part is the "Repository name (and optionally a tag) for the image" and the final input is the URL or PATH. The path seems like perhaps the part I'm getting wrong. There's no folder here for batch, just for app. Let's try this:

$ docker build --pull --tag batch app

Okay that seems to have built successfully. Let's try restarting:

$ sudo systemctl restart app

Hmm, this is taking longer than usual. I suppose that's promising. Let's try forwarding the domain name again and see if we get errors. Currently we're getting lots of readings on the app server, which means the relay is sending successfully. adding the following to /opt/nginx/nginx.conf:

    server {
            listen 80;
            listen 443 ssl;
            server_name www.heatseeknyc.com heatseeknyc.com;
            return 301 $scheme://www.heatseek.org$request_uri;
    }

Restarting the nginx server:

$ sudo nginx -s reload

Okay, heatseeknyc.com appears to be successfully forwarding to heatseek.org in Chrome. Let's check the logs to make sure the readings from the relay server are still coming through.

Crap. There are no more readings coming in on the app server nginx log file. It seems like the redirect is preventing any readings from coming through. Perhaps I didn't rebuild the docker image properly? What are your thoughts @hrldcpr?

[hrldcpr]

Yeah, the batch module doesn't have its own docker image, unlike most of the other modules. Definitely a bit confusing, I'll see if I can add some clarification to the readme. You could maybe guess this by the fact that the module is inside the app/ directory, at app/batch.service, or you can know for sure by looking at the contents of app/batch.service, in which the commands use the app docker image.

So anyway, to update the batch module, you have to rebuild the app image:

docker build --pull --tag app app
sudo systemctl restart batch

I won't run that because I'm not sure what the state of the domains is, but that should work!

[williamcodes]

Excellent, worked like a charm. Thanks Harold!