5 ms·

I have used docker on MacOS. I don’t know about windows but it ran fine on my mac. Is there something particular that didn’t work for you?

geerlingguy 7y ago
There are a number of use cases where Docker for Mac and Windows are painful compared to Linux, and to overcome those problems (like fs performance on shared volumes) people spend a lot of time building hacky solutions.
jacobr1 7y ago
docker for osx uses a linux VM behind the scenes (transparently)
zxexz 7y ago
I use Docker on macOS daily for work. I love it, and am grateful for it, but the experience is NOTHING like using Docker on Linux. Docker for macOS runs in a virtual machine, and requires dedicating RAM to the daemon. You aren’t directly using the host OS’ kernel either, which can lead to wonky behavior, especially in regards to networking (IME).
- RMPR 7y ago
  Have you tried podman (https://podman.io/whatis.html https://podman.io/whatis.html)?
  - FireBeyond 7y ago
    How would podman help here? It doesn't run on macOS at all.
deergomoo 7y ago
- Bind mount performance is appalling, even with the delegated/cached modes. I understand the reasons for this, but it's still a major issue for some workloads. To use PHP development as an example, where a popular framework boot-up can read hundreds of files per request (as well as writing out to caches), it's not uncommon to see 5 second response times for a page that displays "hello world" and nothing else. Thankfully there are tools like docker-sync and Mutagen which, while it's one extra thing to set up, get you back to nearly native performance. - com.docker.hyperkit is nuts. I can have a single container idling in the background doing nothing, hear my laptop fans spin up, and know without checking that com.docker.hyperkit is using 200% CPU for no discernible reason. Restarting the daemon brings things back down...for a while. At the office we use Linux, but since I've been working from home on my Mac I've gone back to Vagrant for as many projects as I can. Heavier and far less easy to orchestrate yes, but I've found I actually end up with better and more predictable performance (and far less time with my laptop doing double duty as a space heater).

geerlingguy 7y ago

There are a number of use cases where Docker for Mac and Windows are painful compared to Linux, and to overcome those problems (like fs performance on shared volumes) people spend a lot of time building hacky solutions.

jacobr1 7y ago

docker for osx uses a linux VM behind the scenes (transparently)

zxexz 7y ago

I use Docker on macOS daily for work. I love it, and am grateful for it, but the experience is NOTHING like using Docker on Linux. Docker for macOS runs in a virtual machine, and requires dedicating RAM to the daemon. You aren’t directly using the host OS’ kernel either, which can lead to wonky behavior, especially in regards to networking (IME).

RMPR 7y ago
Have you tried podman (https://podman.io/whatis.html https://podman.io/whatis.html)?
- FireBeyond 7y ago
  How would podman help here? It doesn't run on macOS at all.

deergomoo 7y ago

- Bind mount performance is appalling, even with the delegated/cached modes. I understand the reasons for this, but it's still a major issue for some workloads. To use PHP development as an example, where a popular framework boot-up can read hundreds of files per request (as well as writing out to caches), it's not uncommon to see 5 second response times for a page that displays "hello world" and nothing else. Thankfully there are tools like docker-sync and Mutagen which, while it's one extra thing to set up, get you back to nearly native performance. - com.docker.hyperkit is nuts. I can have a single container idling in the background doing nothing, hear my laptop fans spin up, and know without checking that com.docker.hyperkit is using 200% CPU for no discernible reason. Restarting the daemon brings things back down...for a while. At the office we use Linux, but since I've been working from home on my Mac I've gone back to Vagrant for as many projects as I can. Heavier and far less easy to orchestrate yes, but I've found I actually end up with better and more predictable performance (and far less time with my laptop doing double duty as a space heater).