Open Stack

On 06.06.2017 13:49, Sean Dague wrote:
> Some good progress has been made so far on Global Request ID work in the
> core IaaS layer, here is where we stand.
> 
> STATUS
> 
> oslo.context / oslo.middleware - everything DONE
> 
> devstack logging additional global_request_id - DONE
> 
> cinder:
> - client supports global_request_id - DONE
> - Cinder calls Nova with global_request_id - TODO (waiting on Novaclient
> release)
> - Cinder calls Glance with global_request_id - TODO
> 
> neutron:
> - client supports global_request_id - IN PROGRESS (this landed,
> released, but the neutron client release had to be blocked for unrelated
> issues).
> - Neutron calls Nova with global_request_id - TODO (waiting on
> Novaclient release)
> 
> nova:
> - Convert to oslo.middleware (to accept global_request_id) - DONE
> - client supports global_request_id - IN PROGRESS (waiting for release
> here - https://review.openstack.org/#/c/471323/)
> - Nova calls cinder with global_request_id - DONE
> - Nova calls neutron with global_request_id - TODO (waiting on working
> neutronclient release)
> - Nova calls Glance with global request id - IN PROGRESS (review needs
> final +2 here https://review.openstack.org/#/c/467242/)
> 
> glance:
> - client supports global_request_id - DONE
> - Glance supports setting global_request_id - IN REVIEW
> (https://review.openstack.org/#/c/468443/) *(some debate on this).
> 
> 
> Everything except the last glance change is uncontroversial, and it's
> just mechanics and project management to get things through in the
> correct order.
> 
> 
> The Glance support for global_request_id has hit a bump in the review
> process as there is a concern that it's changing the API. Though from an
> end user perspective that's not happening, it's just changing which
> field things get logged into. We'll see if we can work through that.
> 
> 	-Sean
> 

FWIW, I *really* appreciate such crisp updates on the ML. Thanks a lot
for your time writing them.

-- 
Regards, Markus Zoeller (markus_z)