Re: Temporary file access API

Robert Haas  wrote:

> On Mon, Apr 11, 2022 at 4:05 AM Antonin Houska  wrote:
> > There are't really that many kinds of files to encrypt:
> >
> > https://wiki.postgresql.org/wiki/Transparent_Data_Encryption#List_of_the_files_that_contain_user_data
> >
> > (And pg_stat/* files should be removed from the list.)
> 
> This kind of gets into some theoretical questions. Like, do we think
> that it's an information leak if people can look at how many
> transactions are committing and aborting in pg_xact_status? In theory
> it could be, but I know it's been argued that that's too much of a
> side channel. I'm not sure I believe that, but it's arguable.

I was referring to the fact that the statistics are no longer stored in files:

https://git.postgresql.org/gitweb/?p=postgresql.git;a=commit;h=5891c7a8ed8f2d3d577e7eea34dacff12d7b6bbd

> Similarly, the argument that global/pg_internal.init doesn't contain
> user data relies on the theory that the only table data that will make
> its way into the file is for system catalogs. I guess that's not user
> data *exactly* but ... are we sure that's how we want to roll here?

Yes, this is worth attention.

> I really don't know how you can argue that pg_dynshmem/mmap.NNNNNNN
> doesn't contain user data - surely it can.

Good point. Since postgres does not control writing into this file, it's a
special case though. (Maybe TDE will have to reject to start if
dynamic_shared_memory_type is set to mmap and the instance is encrypted.)

Thanks.

-- 
Antonin Houska
Web: https://www.cybertec-postgresql.com