Update nginx-proxy for M1 Mac Support

[jmjones88]

Describe the bug
When the nginx-proxy container starts on an M1 Mac, you get the following error:

runtime: failed to create new OS thread (have 2 already; errno=22)
fatal error: newosproc

Steps to Reproduce

1. Start the global containers
2. Look at the Error in Docker Desktop

Expected behavior
The container should start without issue

Environment information

• Device: MacBook Air (M1, 2020)
• OS: MacOS 11.2.3

Additional context
The issue has been solved temporarily on the main repo nginx-proxy/nginx-proxy#1550

[jeremiahbratton]

I am having the same issue, and as I do some research on my end on how to get things working I found this issue. I thought it might be helpful to note that there is a conversation that relates to this happening in an issue for wp-local-docker-v2

For anyone searching for a fix like me this might help close the loop for existing knowledge.

Issue Link: 10up/wp-local-docker-v2#188

[dustinrue]

Hi everyone. I have updated and rebuilt the proxy so that it supports linux/amd64 and linux/arm64. This should provide support for the M1 chip. It would be great if someone could test this. Here is what you need to do to test:

• Issue 10updocker stop all and make sure all docker containers that are part of wp-local-docker global servers are destroyed (check docker ps -a)
• Edit the ~/.wplocaldocker/global/docker-compose.yml file and update the image entry for the gateway from 10up/nginx-proxy to 10up/nginx-proxy:multiarch
• Start an environment and confirm this updated image works as expected

I need to see this tested on both an x86 system as well as an arm64 (M1) system.

Thank you!

[jeremiahbratton]

@dustinrue updated and made the change. The gateway is still having trouble after a 10updocker start. I am on an M1 Mac Mini

Creating global_dns_1 ... 
Creating global_mysql_1 ... 
Creating global_gateway_1 ... 
Creating global_mailcatcher_1 ... 
Host is already in use by another container
Creating global_gateway_1     ... error

ERROR: for global_gateway_1  Cannot start service gateway: driver failed programming external connectivity on endpoint global_gateway_1 (d0a9772620d0babb49af60479c6abc59277766db553d67b7a208535d7520b7a0): Bind for 127.0.0.1:443 failed: port is already allocated
Creating global_mailcatcher_1 ... done
Creating global_mysql_1       ... done
Creating global_phpmyadmin_1  ... 
Creating global_dns_1         ... done
Creating global_phpmyadmin_1  ... done

ERROR: for gateway  Cannot start service gateway: driver failed programming external connectivity on endpoint global_gateway_1 (d0a9772620d0babb49af60479c6abc59277766db553d67b7a208535d7520b7a0): Bind for 127.0.0.1:443 failed: port is already allocated
Encountered errors while bringing up the project.

and from the logs in docker
global_gateway_1 exited with code 128 and Error invoking remote method 'docker-start-container': Error: (HTTP code 500) server error - driver failed programming external connectivity on endpoint global_gateway_1 (95e4d50c8991d11d52ca3ac83f10d58191843066af21045a5e5227676183944f): Bind for 127.0.0.1:443 failed: port is already allocated

[dustinrue]

Make sure you don't have something using that port already. Docker containers, some other local dev system

[jeremiahbratton]

@dustinrue That I can see there is nothing else that would be running and using that port. No other containers are running. I am a little out of my depth with things like this, it is possible I am not looking at the right thing. Do you have any steps you would perform that could help me get more information for you?

[dustinrue]

Make sure you do 10updocker stop all and then docker ps. It should come back with nothing. If it doesn't...you'll have to resolve that. If it is empty you can try telnet localhost 443 to see if something responds. If something responds (it looks like this):

telnet localhost 443
Trying ::1...
telnet: connect to address ::1: Connection refused
Trying 127.0.0.1...
Connected to localhost.
Escape character is '^]'.

Then something has that port open. You can determine using this command

sudo lsof -i -P -n | grep -i listen

[jeremiahbratton]

@dustinrue Thank you for that... ok, found it. Something must have gone wrong during the first wind up after updating. Going into docker desktop and deleting the 10up/nginx-proxy:multiarch container then doing another 10updocker start got it running. The mySQL container is still failing so I can't fully confirm the setup. But I can see a green light for 10up/nginx-proxy:multiarch in docker desktop

Also, I can at least get a 503 at the project URL so that is definitely progress.

[dustinrue]

Yup. The MySQL container needs to be swapped out as well and I’ll see what I can do about it in the coming days. Rumor has it you can use MariaDB but this is untested

[jeremiahbratton]

I was given that recommendation in the other thread (10up/wp-local-docker-v2#188) but doesn't seem to be as simple as just changing mysql:5 to mariadb but I digress... Thank you for your work and time.

[jeremiahbratton]

Update: I guess it was as simple as changing mysql:5 to mariadb in docker-compose.yml

I can now confirm for you that after the update, and the tweak to the mysql image I can access the local environment.

[azatecas]

@dustinrue thanks it now works great and i can be on the same workflow as the rest of my team!