Kadira Successfully connected but no data about meteor-apm-server

[funkyremi]

Thank you it's working after manually editing the default.conf file and restarting the service.

[jehartzog]

I spent a decent amount of time looking into this, and there was not a simple solution, at least many months ago. That was before MUP released it's nginx config features, so it may be doable now. I'll take another look into it, no promises though :D.

[lmachens]

Take a look at #14 (comment)

Looks like the mup hook didn't work for you https://github.com/lmachens/meteor-apm-server/blob/master/mup-placeholder.js#L57

@jehartzog did you notice issues with your solution?

[jehartzog]

It works fine for me, but I freely admin that fix is hacky and probably won't work in all cases. For example, it relies on the apm-engine app being listed first in the proxy file, which may not always happen.

If it doesn't work @funkyremi you'll need to manually edit the port in the /etc/nginx/conf.d/default.conf file in the docker container.

[vparpoil]

Hello there,
I just installed successfully the APM, but got this issue.
I fixed manually using these steps :

• docker ps to list containers
• docker exec -ti CONTAINER_ID /bin/bash to enter the docker container named mup-nginx-proxy
• manually edit /etc/nginx/conf.d/default.conf to change port 80 by port 11011 for the apm-engine upstream (I also had to install vi on the docker)
• nginx -t to test config, then service nginx reload
  I am pretty sure this fix won't stay after a new deployment using mup...

[lmachens]

Do you have any client side errors?

[funkyremi]

No client or server side error. Just the "Configure your app" page still displayed and no data available.

[lmachens]

Please check if you have the nginx issue described in 2. https://github.com/lmachens/meteor-apm-server#configuration-details
If the port shows 80 this might be the issue.

Did you setup the server with mup?

[funkyremi]

Ok that's it, the MUP hooks may be failing because I see port 80. Yes the server has been setup with MUP. What should I try to resolve this issue?

[Ehrs]

I am also having trouble with this default.conf file. The "solution" with manually editing default.conf is somehow automatically reverted after a while (around 8 hours). How can I find what is causing this? I manually edit the file and then run npx mup deploy to make it work for a while. I also tried docker restart mup-nginx-proxy at my apm server but that would instantly revert my changes and use port 80 again.

[jehartzog]

If the server restarts then the proxy settings are reset, same with restarting the proxy. Sounds like your server is probably restarting every 8 hours for some reason.

[Ehrs]

I have an uptime that says otherwise...

Am I missing some detail in how docker container works? I make manual changes outside the container then copying it to a default2.conf using

docker cp default.conf mup-nginx-proxy:/etc/nginx/conf.d/default.conf2

followed by

docker exec -t mup-nginx-proxy cp /etc/nginx/conf.d/default.conf2 /etc/nginx/conf.d/default.conf

[jehartzog]

I don't see anything wrong with how you're doing it, but I just directly edited using vi or vim, which I had to install in the container.

Other things which will reset the proxy config are running a mup deploy again, even if you server is up and running, so if you are pushing code updates that could also be causing the reset.

[Ehrs]

Made a lazy attempt to install vim in the container but it didn't work first try so I did it like this. :)

After manually editing the file, should I just restart the container mup-nginx-proxy? That makes the most sense to me but it didn't really work.

[jehartzog]

Restarting the container will end up with your config changes being reset. Instead just tell the container to reload nginx:

docker exec mup-nginx-proxy nginx -s reload

A longer discussion about all this stuff is here.

[lmachens]

I think we need a better solution here. If the nginx container restarts and overwrites the config, this error will happen again.
Maybe we can use the mup advanced proxy settings?

// This config can add additional upstream and server blocks if needed.
nginxConfig: './path/to/config',

Or use env vars for nginx proxy to change the nginx config?

// Environment variables for nginx proxy
env: {
 DEFAULT_HOST: 'foo.bar.com'
}

I think the issue is that we expose different ports for different domains.
There is a discussion on nginx-proxy/nginx-proxy#59.

Ideas/PRs are welcome :).

[lmachens]

@vparpoil see discussion in #14