SQL Server 2017 is now in General Availability (GA) on Linux. To support an operating system and database system that runs on it requires a deep level of knowledge. Troubleshooting issues, knowing the tools for diagnostics, and becoming familiar with the interactions and compatibilities among components is like the study of religion. It requires a great deal of dedication and patience. I started with Sybase (which birthed SQL Server) on Solaris (a Unix variant) many years ago and I transitioned to SQL Server on Windows. I missed the Unix tools so much that I insisted on installing MKS Toolkit (Unix tools on Windows) wherever I went as a freelance web programmer. I succeeded in building many data warehouses using these tools and welcome the addition of SQL Server to the Linux world.
Had Powershell emerged with “here” document capabilities back in 1995, things would be different. Since Oracle now controls MySQL, and MariaDB is gaining momentum, I expect that the stability of the SQL Server database engine will gain some converts among the Linux community wishing for an economical vendor-supported product. Despite popular belief, you don’t need .NET to run the SQL Server database engine, and you don’t need SSRS or SSIS to deliver reporting solutions and high speed streamed data integrations. The only tools MS needs to port with the database engine are the command line versions of BCP.exe and SQLCMD.exe, which they have done.
The database development software is supported on four varieties of Linux here.