Hello Docker Community!
I’m coming up empty on diagnosing an issue driving me to drink. I was hopeful of getting some help debugging the problem.
I have a docker container built from the official PHP images here https://hub.docker.com/r/circlical/php74app-behat-multi/
With the advent of new Apple Silicon Macs, I figured it’d be a cinch to use buildx to compile versions of this image - but the image isn’t behaving as expected. Seems to compile ok, but on the M1 Macs, it just ‘jams’.
I’m using docker compose to raise the image
version: "2"
services:
php:
image: circlical/php74app-behat-multi:latest
volumes:
- ./docker/php.ini:/usr/local/etc/php/php.ini:delegated
- ./docker/php-fpm.conf:/usr/local/etc/php-fpm.conf:delegated
- ./docker/php-log/:/var/log/php/:delegated
- ./build/behat/coverage/:/behat-coverage/:delegated
- nfsmount:/code:cached
networks:
- lemonade-testnet
networks:
lemonade-testnet:
driver_opts:
com.docker.network.driver.mtu: 1450
driver: bridge
volumes:
nfsmount:
driver: local
driver_opts:
type: nfs
o: addr=host.docker.internal,rw,nolock,hard,nointr,nfsvers=3
device: ":${PWD}"
The filesystem is using nfsd, and I can confirm that it seems to mount as expected (can access all files from within the PHP container).
On my Intel Mac, I docker compose up and interact with it as one normally would. I can docker container exec -it $(docker-compose ps -q php) /bin/bash
to get into the container, and execute code from the command line with no issue.
On the Apple Silicon Mac, however, executing this same described process just hangs (dead) - even CTRL+C won’t work. I’m not getting any backtraces with gdb, and if I strace the process, the output always freezes at the same spot and never reaches an exit code (just hung, dead in the middle of a read operation).
I’ve tried manipulating the amount of memory, reducing to 1 CPU, and tried every permutation of Virtualization Framework and gRPC/VirtioFS, and it’s always the same result.
Can you recommend some help toward debugging the problem? Anyone seen this?
Appreciated!
Affected machine is an Apple M1, 16 GB on Mac Ventura 13.2