SQL Server – Best Practices for Placing Database Files in a Shared Folder

What I would do is to benchmark the performance before and after the changes. There should be a performance gain after tempdb is moved to another drive. Use DMVs like sys.dm_io_virtual_file_stats to see the read and write wait times for the DB files.

Use the perfmon physical disk counters: Avg. Disk sec/Read, Avg. Disk sec/Write, Disk Reads/sec, Disk Writes/sec. The change in these metrics will let you know if the disks are physically separate or are just the same disks which are logically separated.

Want a third party disk benchmarking tool for your disks, so that the SAN admins cannot accuse SQL server of its results? Use SQLIO. Although there is SQL in the name, this tool does not require SQL server to run and is not related to SQL server.

This is a good question, and I look forward to see the other answers.

this isn't exactly how filestream and filetables work, you can't specify the name of an existing share as SQL server will try and create the share that you specify in SQL Server Configuration Manager. This is because the share is not mapped directly to a folder on the filesystem like a normal share, but an NTFS data container that lives in the filestream filegroup that is created for the database.

There is a misconception with filestream/filetable that because the data is stored in the filesystem, it is just a normal folder, this isn't entirely correct.

When you create your database, you specify that one of the filegroups is a filestream filegroup, this isn't like a normal filegroup and is actually a series of NTFS folders called Data Containers, this is where your files actually get stored, you can't modify this folder directly, but you can look at the contents (not that there's much human readable stuff in here though, you can find your actual files in here though if you look hard enough through the GUID folders and oddly named files :)).

The file share that you use to copy the files into the table is actually a representation of the data inside these data containers that is presented to Windows explorer via a filter driver which uses the streaming APIs to make the changes to the data containers and add the rows to the filetable etc.

So if you want to specify the location that the files are physically stored, you need to specify that location as where you want the filestream filegroup stored, then you can change the share name to whatever you want.