Allain,

Thanks a lot for the confirmation and clarification. Yes, the host driver is being redesigned. I’ll drop the patch accordingly.

BR,

Kailun

From: Legacy, Allain [mailto:Allain.Legacy@windriver.com]
Sent: Tuesday, November 6, 2018 8:16 PM
To: Qin, Kailun <kailun.qin@intel.com>; Peters, Matt <Matt.Peters@windriver.com>; starlingx-discuss@lists.starlingx.io
Subject: RE: [Starlingx-discuss] Analysis of patch 039705e for StarlingX upstreaming

I agree.   These changes can all be dropped.  

The keystone config opts change was needed for the host driver to communicate back to our sysinv subsystem, but, as far as I know, that is being redesigned to avoid this requirement.

Regards,

Allain

Allain Legacy, Software Developer, Wind River
direct 613.270.2279  fax 613.492.7870 skype allain.legacy

350 Terry Fox Drive, Suite 200, Ottawa, Ontario, K2K 2W5

From: Qin, Kailun [mailto:kailun.qin@intel.com]
Sent: Tuesday, November 06, 2018 2:10 AM
To: Peters, Matt; starlingx-discuss@lists.starlingx.io
Subject: [Starlingx-discuss] Analysis of patch 039705e for StarlingX upstreaming

Hi Matt,

We’re looking into the patch 039705e for upstreaming, which contains several base changes.

1. For the oslo.db related change, upstream has already dropped this override in neutron-server to make it use default values from oslo.db (see https://review.openstack.org/456419/ for detailed information).

2. For the keystone opts registration, I guess it is used for connecting to external services. From upstream’s perspective, this should be maintained along with the specific service to extend rather than make it globally.

3. For the others, I think they’re StarlingX specific.

Thus, I suggest that we should not upstream these changes.

What do you think? Let me know if any question, thanks!

BR,

Kailun