On 07/16/14 18:54, John Ferlan wrote:
On 07/11/2014 08:20 AM, Peter Krempa wrote:
> Version 2 now splits the stuff into separate driver backend funcs.
>
> Peter Krempa (2):
> storage: wipe: Move helper code into storage backend
> storage: Split out volume wiping as separate backend function
>
> src/storage/storage_backend.c | 203 +++++++++++++++++++++++++++++++++
> src/storage/storage_backend.h | 12 ++
> src/storage/storage_backend_disk.c | 1 +
> src/storage/storage_backend_fs.c | 3 +
> src/storage/storage_backend_iscsi.c | 1 +
> src/storage/storage_backend_logical.c | 1 +
> src/storage/storage_backend_mpath.c | 1 +
> src/storage/storage_backend_scsi.c | 1 +
> src/storage/storage_driver.c | 205 +---------------------------------
> 9 files changed, 229 insertions(+), 199 deletions(-)
>
ACK both as they seem to fulfill the changes requested from
http://www.redhat.com/archives/libvir-list/2014-July/msg00556.html
Thanks; Pushed.
John
BTW: Since I was looking at a bug (bz 1091866) in this area - this
caught my attention... Not that these changes fix the problem, but
perhaps they'll help since as it seems "sparse snapshot" logical volumes
are "problematic"...
Well with this stuff pushed, you can easily add a wrapper function for
the logical pool that will check whether the volume is sparse and forbid
destroying it before relaying to the wiping function.
Peter