Hi, On 2022-02-11 16:19:12 -0500, Robert Haas wrote: > I somewhat hope we never end up with THREE strategies for creating a new > database, but now that I think about it, we might. Somebody might want to > use a fancy FS primitive that clones a directory at the FS level, or > something.
I think that'd be a great, and pretty easy to implement, feature. But it seems like it'd be mostly orthogonal to the "WAL log data" vs "checkpoint data" question? On the primary / single node system using "WAL log data" with "COW file copy" would work well. I bet using COW file copies would speed up our own regression tests noticeably - on slower systems we spend a fair bit of time and space creating template0 and postgres, with the bulk of the data never changing. Template databases are also fairly commonly used by application developers to avoid the cost of rerunning all the setup DDL & initial data loading for different tests. Making that measurably cheaper would be a significant win. Greetings, Andres Freund