On Wed, Mar 7, 2018 at 3:37 PM, Richard W.M. Jones <rjones@redhat.com> wrote:

On Wed, Mar 07, 2018 at 03:12:58PM +0200, Arik Hadas wrote:
> If for some predefined period of time no new disk is added or an upload
> doesn't make any progress (assuming the uploads are done sequentially), to
> fail the import operation and that would roll back the resources (disks,
> VMs) that were created as part of the import process.

At the moment we're actually trying to remove the disk on failure.
However the disk_service.remove() API does nothing (doesn't even
fail). Perhaps because the transfer isn't finalized on the error
path?

That's weird, it should fail to acquire the disk's lock.

In any case, I think that removing the disk this way would be wrong as we also store an entity in the image_transfer table that should be removed (otherwise, another attempt to upload the same disk would probably fail).

Daniel/Nir, I can't find a way to cancel an ongoing upload-image operation through the API (as we have in the webadmin), am I missing it or is it missing?

Anyway the code - which needs review - is:

https://www.redhat.com/archives/libguestfs/2018-March/msg00024.html

Rich.

--
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
Read my programming and virtualization blog: http://rwmj.wordpress.com
Fedora Windows cross-compiler. Compile Windows programs, test, and
build Windows installers. Over 100 libraries supported.
http://fedoraproject.org/wiki/MinGW