| sean-k-mooney[m] | admin1: no the host value is the rpc endpoint name and maps to the name of the compute service that manages the host the instance is associated with. the node value is the name of the compute node, for ironic the node value is the uuid of the corresponding ironic baremetal node, for all other virt drivers the node value is the same as the host value which is set by the host config option which defaults to socket.gethostname(). | 01:28 |
|---|---|---|
| sean-k-mooney[m] | so while the value of host and node is often the same they are conceptualy not the same. | 01:28 |
| sean-k-mooney[m] | ierdem__: once you create a vm form an image you could modify the content of the instance root disk via first boot scripts or logging in so the signigure of instance snapshot would not have any releation to the orginal image. it does not hold true that if an image is trusted the vm is trusted because the instance can be modifed and there for any trust fo the orginal image cannot be assumed to tansfer to any snapshot of the | 01:37 |
| sean-k-mooney[m] | image. first-boot script i.e. via user-data and cloud-init, ssh key injections, modifcation via login and other operatiosn such as boot from rescue disk mean that there is a treat vector and any kind of trust that existed for the orginal image need to be reverifed from scratch for any snapshots | 01:37 |
| sean-k-mooney[m] | Etua: you have two options. either extend the functional test suite to mimic the senarios you want to test or take a more real world approch and deploy a large number of compute services using the fake driver. if you can deploy the fake driver in a container that make it much simpler to run many of the on a limited number of hosts. | 01:40 |
| sean-k-mooney[m] | in either case you will need to write code to build the senairios and execute your tests. | 01:41 |
| sean-k-mooney[m] | the schduler is more or less independent form the rest of nova and can be deployed on its own. its only api is the internal rpc api which is normally invoked by the nova conductor. if you create the api and cell dbs and a rabbitmq broker then populate the cell db with fake compute nodes and set up the cell->host mappings you could fake the compute nodes that way and then do your testing by creating flavors and invoking the rpc | 01:46 |
| sean-k-mooney[m] | api. | 01:46 |
| sean-k-mooney[m] | it is not a trivial thing to do however. | 01:46 |
| sean-k-mooney[m] | its not hard just invovled/tedious to do that | 01:47 |
| sean-k-mooney[m] | the functional tests have fixtures that automate much of this but its not reusable independetly. | 01:49 |
| opendevreview | Ghanshyam Mann proposed openstack/python-novaclient stable/xena: Make tox.ini tox 4.0.0 compatible https://review.opendev.org/c/openstack/python-novaclient/+/868720 | 03:12 |
| opendevreview | Ghanshyam Mann proposed openstack/python-novaclient stable/wallaby: Make tox.ini tox 4.0.0 compatible https://review.opendev.org/c/openstack/python-novaclient/+/868721 | 03:17 |
| opendevreview | Ghanshyam Mann proposed openstack/nova stable/zed: DNM: testing requirement-checks job on stable branch https://review.opendev.org/c/openstack/nova/+/868852 | 04:08 |
| opendevreview | Ghanshyam Mann proposed openstack/nova stable/xena: DNM: testing requirement-checks job on stable branch https://review.opendev.org/c/openstack/nova/+/868853 | 04:21 |
| opendevreview | Ghanshyam Mann proposed openstack/nova stable/xena: DNM: testing requirement-checks job on stable branch https://review.opendev.org/c/openstack/nova/+/868853 | 05:05 |
| *** bhagyashris is now known as bhagyashris|ruck | 06:30 | |
| *** mtreinish_ is now known as mtreinish | 15:24 | |
| opendevreview | Jean-Sébastien Bevilacqua proposed openstack/nova master: Add Lustre support to nova https://review.opendev.org/c/openstack/nova/+/853786 | 15:34 |
| *** mtreinish_ is now known as mtreinish | 15:34 | |
| *** mtreinish_ is now known as mtreinish | 15:42 | |
| opendevreview | Jean-Sébastien Bevilacqua proposed openstack/nova master: Add Lustre support to nova https://review.opendev.org/c/openstack/nova/+/853786 | 17:12 |
| opendevreview | Ghanshyam Mann proposed openstack/python-novaclient stable/xena: Make tox.ini tox 4.0.0 compatible https://review.opendev.org/c/openstack/python-novaclient/+/868720 | 18:19 |
| opendevreview | Ghanshyam Mann proposed openstack/python-novaclient stable/wallaby: Make tox.ini tox 4.0.0 compatible https://review.opendev.org/c/openstack/python-novaclient/+/868721 | 18:19 |
| opendevreview | Ghanshyam Mann proposed openstack/placement stable/yoga: Make tox.ini tox 4.0.0 compatible https://review.opendev.org/c/openstack/placement/+/868735 | 23:10 |
| opendevreview | Ghanshyam Mann proposed openstack/placement stable/xena: Make tox.ini tox 4.0.0 compatible https://review.opendev.org/c/openstack/placement/+/868736 | 23:36 |
| opendevreview | Ghanshyam Mann proposed openstack/placement stable/wallaby: Make tox.ini tox 4.0.0 compatible https://review.opendev.org/c/openstack/placement/+/868737 | 23:37 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!