Comment 21 for bug 1684272

I made another test here https://review.openstack.org/#/c/466311/ removing another suspect Heat commit (https://review.openstack.org/#/c/435171/). We can see the result here: http://logs.openstack.org/11/466311/1/check-tripleo/gate-tripleo-ci-centos-7-ovb-updates/a3a49fc/logs/postci.txt.gz

We can see that it has a noticeable effect on the service chain creation:

ControllerServiceChain 108.0

Compared to 384 on a similar job previously. But! The overall deployment time is not impacted. It's because we move calculation of the attributes to the resource creation, instead of later on. So while the commit that we already reverted had an effect, I don't think that one does, and the change in timing on the chain creation is a false alarm.