I am having reasonable success running SQL Server for Linux on Docker for Mac, as per the instructions on https://docs.microsoft.com/en-us/sql/linux/sql-server-linux-setup-docker. The article also notes that:
Docker on Mac Performance: Docker Machine vs Docker for Mac One of the key advantages of Docker is the ability to run Docker Images on any operating system. An image is a lightweight, stand-alone, executable package that includes everything needed to run a piece of software.
Volume mapping for Docker-machine on Mac with the SQL Server on Linux image is not supported at this time.
This, of course, makes it difficult to do any more than really simple testing on the Mac if you have to re-install the database every time you start up Docker.
Has anybody a suggested strategy for maintaining a database beyond the current session?
ManngoManngo2 Answers
This is a known issue: mapping volumes on Macs isn't supported yet. You can follow that Github issue for more news, like when it's fixed.
Until then, no, people aren't doing anything more than simple testing with that combination of tools (Mac, Docker, Linux, SQL Server).
If you need to do real, productive development on a platform that's supported today, use:
- The current supported version of SQL Server (2016, as of this writing)
- Running on a currently supported OS (Windows, as of this writing)
- If you insist on using a Mac as your hardware layer (as I do too), then the easiest way to bridge these technologies is a hypervisor like VMware Fusion or Parallels
@BrentOzar answered my question, but I have since found a workable solution.
An alternative approach is to use a Data Volume Container. Simply put:
In the above:
--name mssql:mssqlis an arbitrary name for your data container-e SA_PASSWORD=…: obviously…needs to be a real password--name sql-server:sql-serveris an arbitray name for my SQL Server container.
It works after successive restarts of Docker.
Docker For Mac Vs Parallels
Thanks also to @AaronBertrand for pointing me to https://github.com/Microsoft/mssql-docker/issues/12 and, from there to @twright-msft for pointing me to Data Volume Containers.
ManngoManngo