Re: Not cleaning up tempfiles for multipart?

[Hongli Lai <hongli@phusion.nl>]

On Mar 8, 3:42 pm, James Tucker <jftuc...@gmail.com> wrote:
> Just to be clear, I don't disagree at all that this needs to be fixed. I think handlers/servers should close any tempfiles they create, and the requirement for middleware should be the same.

Glad we're on the same page. :) Phusion Passenger does clean up its
own tempfiles. As for the link you posted, I think that's for cleaning
up the rack.input buffer file, not multipart files.


Charles Nutter:
> I don't think it should be specified behavior that a tempfile *ever*
> exists on disk. The mechanism of temporary storage for a large
> incoming post should be a black box. Consider systems that won't
> actually have a writable filesystem, like GAE; there, the "tempfile"
> data would be stored in memory or in BigTable.

Agreed. However I think it should be *possible* to access the on-disk
file directly as an optimization for apps that need it, they just
musn't rely on it to be available. For example if your app deals with
2 GB file uploads then you really want to avoid that extra copy when
storing the file upload in a storage location on the same filesystem,
whenever possible.