Common practice is to put TEMPDB files on a dedicated disk and size TEMPDB to fill the entire disk.

When TEMPDB grows unexpectedly it can quickly use all the space on the disk and is a common cause of SQL Server issues.

We have great observability about TEMPDB issues in SQL Monitor now but no default alerts.

We have some metrics we can alert on such as TEMPDB Free Bytes, but a literal value is not much use in a mixed environment where some servers have 1TB+ TempDB databases and some have only 30GB.

We collect a custom percentage free disk space metric and put in warnings/errors when this hits 70%/90%.

I think this would be useful to include in the base product.