[PATCH v2 5/6] tests: unit tests for iSCSI iSER

[Daniel P. Berrangé]

On Fri, May 15, 2020 at 10:26:22AM +0200, Peter Krempa wrote:
> On Fri, May 15, 2020 at 09:03:39 +0100, Daniel Berrange wrote:
> > On Fri, May 15, 2020 at 11:46:04AM +0800, Han Han wrote:
> > > Signed-off-by: Han Han <hhan at redhat.com>
> > > ---
> > >  .../qemuxml2argvdata/disk-network-iscsi.args  |  8 +++-
> > >  .../disk-network-iscsi.x86_64-2.12.0.args     |  7 ++-
> > >  .../disk-network-iscsi.x86_64-latest.args     | 45 +++++++++++--------
> > >  tests/qemuxml2argvdata/disk-network-iscsi.xml |  9 ++++
> > >  tests/qemuxml2argvtest.c                      |  5 ++-
> > >  .../qemuxml2xmloutdata/disk-network-iscsi.xml | 10 +++++
> > >  tests/qemuxml2xmltest.c                       |  4 +-
> > >  tests/virstoragetest.c                        | 16 +++++++
> > >  8 files changed, 80 insertions(+), 24 deletions(-)
> > > diff --git a/tests/qemuxml2argvdata/disk-network-iscsi.xml b/tests/qemuxml2argvdata/disk-network-iscsi.xml
> > > index 8a55f1b2..9812b4b1 100644
> > > --- a/tests/qemuxml2argvdata/disk-network-iscsi.xml
> > > +++ b/tests/qemuxml2argvdata/disk-network-iscsi.xml
> > > @@ -55,6 +56,14 @@
> > >        </source>
> > >        <target dev='sda' bus='scsi'/>
> > >      </disk>
> > > +    <disk type='network' device='lun'>
> > > +      <driver name='qemu' type='raw'/>
> > > +      <source protocol='iscsi' name='iqn.1992-01.com.example/1'>
> > > +        <host name='example.org'/>
> > > +        <iser/>
> > > +      </source>
> > > +      <target dev='sdb' bus='scsi'/>
> > > +    </disk>
> > 
> > I'm thinking would be better to have this represented as
> > an attribute for the host, as we're basically choosing between
> > TCP and RMDA as the network link, over which ISCSI is run eg
> > 
> >       <source protocol='iscsi' name='iqn.1992-01.com.example/1'>
> >         <host name='example.org' transport="tcp|rdma"/>
> >       </source>
> 
> I specifically rejected this in the last review. The issue is that
> 'rdma' is a wrong term. It's not even used by qemu any more.
> Additionally iser runs on top of a variety of transport depending on
> what technology you have so encoding it as a transport is wrong.
> 
> See https://www.redhat.com/archives/libvir-list/2020-May/msg00397.html

Ah, thanks for the explanation. I agree with this.

> > If we really want to keep the "iser" terminology, then it would
> > be as a replacement for protocol="iscsi" instead g
> > 
> >       <source protocol='iser' name='iqn.1992-01.com.example/1'>
> 
> Yes. IMO this is the only way to go. Specifically it will also make
> other things like disk type='volume' les weird IMO.

Yeah, I think this makes sense, given that iser is technically a
different protocol to "iscsi", even if they shared 90% commonality.

Regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|