New manual install get nginx 502 error

[mediclab]

After new magento 2.2.6 install with sample data, nginx get 502 error

app_1            | 2019/01/16 08:01:11 [info] 8#8: *1 client closed connection while waiting for request, client: 172.19.0.1, server: 0.0.0.0:8443
app_1            | 2019/01/16 08:01:11 [info] 7#7: *2 client closed connection while waiting for request, client: 172.19.0.1, server: 0.0.0.0:8443
app_1            | 2019/01/16 08:01:11 [crit] 8#8: *3 connect() to unix:/sock/docker.sock failed (2: No such file or directory) while connecting to upstream, client: 172.19.0.1, server: , request: "GET /admin/ HTTP/1.1", upstream: "fastcgi://unix:/sock/docker.sock:", host: "magento.local"
app_1            | 172.19.0.1 - - [16/Jan/2019:08:01:11 +0000] "GET /admin/ HTTP/1.1" 502 576 "-" "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Ubuntu Chromium/71.0.3578.98 Chrome/71.0.3578.98 Safari/537.36" "-"
app_1            | 2019/01/16 08:01:11 [crit] 8#8: *3 connect() to unix:/sock/docker.sock failed (2: No such file or directory) while connecting to upstream, client: 172.19.0.1, server: , request: "GET /favicon.ico HTTP/1.1", upstream: "fastcgi://unix:/sock/docker.sock:", host: "magento.local", referrer: "https://magento.local/admin/"
app_1            | 172.19.0.1 - - [16/Jan/2019:08:01:11 +0000] "GET /favicon.ico HTTP/1.1" 502 576 "https://magento.local/admin/" "Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Ubuntu Chromium/71.0.3578.98 Chrome/71.0.3578.98 Safari/537.36" "-"

medic84-note:~/Documents/Work/magento-docker$ sudo docker-compose exec phpfpm ls /sock -la
total 8
drwxr-xr-x 2 app  app  4096 Dec 24 15:02 .
drwxr-xr-x 1 root root 4096 Jan 16 08:21 ..

[mark-gene]

I also have repeated this issue when trying to do a manual install

[docker77]

Same to me. Any workarounds?

[markshust]

are you installing to a directory in which you previously had an old docker-magento install? if so i believe you need to remove the associated data volumes because of the old config being run against them. you can manually remove them or run bin/remove, then try reinstalling again.

[mediclab]

are you installing to a directory in which you previously had an old docker-magento install? if so i believe you need to remove the associated data volumes because of the old config being run against them. you can manually remove them or run bin/remove, then try reinstalling again.

No. It's a really new install. And after error, re-install app with bin/remove also not working.

[mark-gene]

I think it is something to do with a default php-fpm config file which overrides and sets php-fpm to listen on port 9000. I believe the file is zz-docker.conf

docker-library/php#241

^May not be the full reason, but if I rebuilt the image locally using your Dockerfile and added a line to remove this I then got it working as expected.

[markshust]

i'm unable to replicate. i tried an automated install and a manual install, and both successfully installed.

i want to verify that you are completely purging your system before trying again

there should be nothing returned as a result for:

containers: docker ps -a
images: docker images
volumes: docker volume ls

note that bin/remove will remove all associated containers for the project, and bin/removevolumes will remove all associated volumes. you need to run both to completely purge a project. perhaps something is laying around.

fyi i made no changes to port 9000 or anything wtih zz-docker.conf. appears to work as is.

[markshust]

also, i want to verify everyone is using the exact steps laid out at https://github.com/markoshust/docker-magento#quick-setup

these are the steps i'm taking to replicate a successful install (either automated or manual, both successfully installed for me).

[lucbard]

I have the same issue trying to manually install Magento CE 2.2.6 here is my setup before starting the install:
docker container ls -a CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
docker images -a REPOSITORY TAG IMAGE ID CREATED SIZE
docker volume ls DRIVER VOLUME NAME

Then here are the commands that I ran for manual setup:
curl -s https://raw.githubusercontent.com/markshust/docker-magento/master/lib/template | bash -s -- magento-2

bin/download 2.2.6

I had to change the docker.composer.yml this :
image: markoshust/magento-php:7.1-fpm-0

because it failed to find the 7.2-fpm-3 package and even with 7.2-fpm-2 I had a bunch of errors that it did not meet the PHP version requirements.

bin/setup mydomain.dev

Other then that the setup goes very well with no errors. My only problem is that I am getting the 502 Bad Gateway error when trying to reach the site.

The bin/magento command works fine BTW.

I ran the command docker logs -f ea2891582dfd (this is the nginx container id)

And this is what I see:

172.19.0.1 - - [17/Jul/2019:21:26:54 +0000] "GET / HTTP/1.1" 502 174 "-" "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_14_5) AppleWebKit/605.1.15 (KHTML, like Gecko) Version/12.1.1 Safari/605.1.15" "-"
2019/07/17 21:26:54 [crit] 7#7: *6 connect() to unix:/sock/docker.sock failed (2: No such file or directory) while connecting to upstream, client: 172.19.0.1, server: , request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/sock/docker.sock:", host: "mydomain.dev"
2019/07/17 21:27:30 [info] 7#7: *6 client 172.19.0.1 closed keepalive connection

Any help would be appreciated.

Thanks,
Luc

[vernard]

This is also happening to me, and I found a solution.

I containerized a whole Magento installation inside a custom image and it's loading all of the files upon docker-compose up -d.

I found that there's an error when I initially compose my container which says UnixHTTPConnectionPool(host='localhost', port=None): Read timed out. (read timeout=60)

But when I run docker-compose up -d again, it just goes up fine.

I'm guessing it is related to that.

So what I did is extended the timeout by adding the following environment variables:

export DOCKER_CLIENT_TIMEOUT=120
export COMPOSE_HTTP_TIMEOUT=120

And after that, I refreshed my containers including the volume.

I ran this command: docker-compose down && docker volume prune -f && docker-compose up -d (BE CAREFUL, THIS REMOVES YOUR VOLUME AND YOU'LL LOSE ALL DATA IN YOUR CONTAINER)

EDIT:

Alternatively, I created an .env file (same directory as the docker-compose.yml file) and put these as variables:

DOCKER_CLIENT_TIMEOUT=120
COMPOSE_HTTP_TIMEOUT=120

That way, I don't have to re-set it everytime it fails.