| openstackgerrit | Adam Gandelman proposed stackforge/akanda-rug: devstack: Update URL for appliance image to tarballs.o.o https://review.openstack.org/223364 | 01:17 |
|---|---|---|
| *** puranamr has joined #akanda | 01:22 | |
| openstackgerrit | Adam Gandelman proposed stackforge/akanda-rug: Enrich functional test suite https://review.openstack.org/219952 | 01:39 |
| *** puranamr has quit IRC | 01:53 | |
| *** ryanpetrello has quit IRC | 03:28 | |
| *** ryanpetrello has joined #akanda | 03:30 | |
| *** elo has joined #akanda | 03:37 | |
| *** sarob has quit IRC | 03:40 | |
| *** ryanpetrello has quit IRC | 03:41 | |
| *** adam_g has quit IRC | 03:45 | |
| *** adam_g has joined #akanda | 03:51 | |
| *** adam_g has quit IRC | 03:51 | |
| *** adam_g has joined #akanda | 03:51 | |
| *** elo has quit IRC | 03:56 | |
| *** sarob has joined #akanda | 04:03 | |
| *** ryanpetrello has joined #akanda | 04:04 | |
| *** elo has joined #akanda | 04:17 | |
| *** elo has quit IRC | 04:46 | |
| *** elo has joined #akanda | 04:47 | |
| *** elo has quit IRC | 05:27 | |
| *** elo has joined #akanda | 06:27 | |
| *** adam_g has quit IRC | 06:55 | |
| *** elo has quit IRC | 08:10 | |
| *** elo has joined #akanda | 08:16 | |
| *** elo has quit IRC | 08:42 | |
| *** puranamr has joined #akanda | 12:49 | |
| *** puranamr has quit IRC | 12:55 | |
| openstackgerrit | David Lenwell proposed stackforge/akanda-rug: Router driver and required changes to instance manager (WIP) https://review.openstack.org/215227 | 12:58 |
| openstackgerrit | David Lenwell proposed stackforge/akanda-rug: Router driver and required changes to instance manager (WIP) https://review.openstack.org/215227 | 13:00 |
| davidlenwell | progress on tests.. and missing stuff from previous commit ^^ | 13:01 |
| openstackgerrit | David Lenwell proposed stackforge/akanda-rug: Router driver and required changes to instance manager (WIP) https://review.openstack.org/215227 | 13:56 |
| *** puranamr has joined #akanda | 15:45 | |
| *** puranamr has quit IRC | 16:16 | |
| *** puranamr has joined #akanda | 16:17 | |
| *** puranamr has quit IRC | 16:49 | |
| *** puranamr has joined #akanda | 16:51 | |
| *** puranamr has quit IRC | 16:52 | |
| *** puranamr has joined #akanda | 16:52 | |
| *** cleverdevil has joined #akanda | 16:53 | |
| *** adam_g has joined #akanda | 17:20 | |
| *** adam_g has quit IRC | 17:20 | |
| *** adam_g has joined #akanda | 17:20 | |
| *** gsagie_ has joined #akanda | 17:21 | |
| *** puranamr_ has joined #akanda | 17:24 | |
| *** puranamr has quit IRC | 17:26 | |
| *** puranamr has joined #akanda | 17:29 | |
| *** puranamr_ has quit IRC | 17:31 | |
| *** puranamr has quit IRC | 17:42 | |
| *** puranamr has joined #akanda | 17:46 | |
| *** puranamr_ has joined #akanda | 18:04 | |
| *** puranamr has quit IRC | 18:06 | |
| *** puranamr has joined #akanda | 18:08 | |
| *** puranamr_ has quit IRC | 18:09 | |
| *** gsagie_ has quit IRC | 18:12 | |
| adam_g | davidlenwell, http://paste.ubuntu.com/12419463/ this gets me up tot he point where state machine is looping and there is a traceback in happening in notifications.py | 18:35 |
| *** puranamr has quit IRC | 18:36 | |
| davidlenwell | adan .. yeah .. that reflects what I was just about to push.. | 18:38 |
| davidlenwell | adam_g: ^ | 18:38 |
| adam_g | davidlenwell, i think your new ensure_cache() wrapper might be the problem | 18:38 |
| davidlenwell | adam_g: okay | 18:38 |
| davidlenwell | hmm..okay | 18:38 |
| davidlenwell | I'll back it out | 18:39 |
| davidlenwell | I have a better idea about how to make the clients thread safe anyways.. I'll deal with that later | 18:39 |
| *** puranamr has joined #akanda | 18:46 | |
| *** puranamr has quit IRC | 18:46 | |
| adam_g | davidlenwell, i say we avoid nice-to-have improvements for now | 18:46 |
| davidlenwell | yeah.. im with you | 18:47 |
| davidlenwell | we can have a conversation about my list of nice to haves later when we are passed this | 18:48 |
| adam_g | davidlenwell, https://review.openstack.org/#/c/215227/14/akanda/rug/instance_manager.py | 19:05 |
| adam_g | davidlenwell, where did the make_vrrp_ports callback go? | 19:05 |
| davidlenwell | adam_g: its supposed to be in the driver and triggered by the driver.pre_boot method .. not sure why its missing in that patch .. | 19:11 |
| davidlenwell | making sure its there on the next one .. | 19:12 |
| rods | adam_g Hi, I'm looking at the ansible playbook and it seems like the dns for the appliance should be set to 8.8.8.8 https://github.com/stackforge/akanda-appliance/blob/master/ansible/tasks/base.yml#L24 | 19:16 |
| adam_g | rods, does it not? | 19:17 |
| rods | but mounting your image at http://tarballs.openstack.org/akanda-appliance/images/akanda_appliance.qcow2 seems like dns is set to "nameserver 127.0.0.1" | 19:17 |
| adam_g | rods, hmm. | 19:18 |
| rods | I'm building my image to and mine is set to "nameserver 192.168.121.1" | 19:18 |
| rods | akanda-linux[vagrant@devstack-develop:/mnt/image/etc] $ ifconfig | 19:18 |
| rods | eth0 Link encap:Ethernet HWaddr 52:54:00:b6:e1:e3 | 19:18 |
| rods | inet addr:192.168.121.94 Bcast:192.168.121.255 Mask:255.255.255.0 | 19:18 |
| adam_g | rods, you mean when you build yours locally, the resulting image gets it set to the hosts NS? | 19:19 |
| rods | the image that I build locally has dns set to 192.168.121.1 | 19:20 |
| adam_g | i wonder if there might be a another DIB element running later and overwriting the ansible change | 19:20 |
| rods | but the image I get from http://tarballs.openstack.org/akanda-appliance/images/akanda_appliance.qcow2 | 19:20 |
| adam_g | right. they should be set to 8.8.8.8 in both cases, as per the ansible playbook | 19:20 |
| rods | has dns set to 127.0.0.1 | 19:20 |
| rods | yep | 19:20 |
| adam_g | would you mind filing a bug ? i will take a look later | 19:21 |
| rods | sure, I'll dig into the issue a little more to see if it's something I can fix | 19:22 |
| adam_g | davidlenwell, how would that work tho? the callback returns ports that we pipe into the nova create request as NICs to be attached at boot time | 19:22 |
| davidlenwell | the callback will be a staticmethod that lives in the driver so that it can differ from driver to driver and still be passed into the nova client | 19:23 |
| adam_g | davidlenwell, is that available somwhere to push up? | 19:28 |
| davidlenwell | its about to be | 19:29 |
| davidlenwell | just a sec .. | 19:29 |
| adam_g | k | 19:29 |
| openstackgerrit | David Lenwell proposed stackforge/akanda-rug: Router driver and required changes to instance manager (WIP) https://review.openstack.org/215227 | 19:55 |
| davidlenwell | adam_g: ^^ this has the callback function the way it should work.. I have not tested it yet because im in the middle of backing off the ensure_cache wrapper thing.. but will get to testing it in a moment.. | 19:56 |
| adam_g | ok | 19:59 |
| adam_g | unfortunately the worker_context and its neutron client are lost by that time that would get called, AFAICS | 20:00 |
| openstackgerrit | David Lenwell proposed stackforge/akanda-rug: Router driver and required changes to instance manager (WIP) https://review.openstack.org/215227 | 20:01 |
| davidlenwell | yeah.. I made a booboo.. that should work.. however .. like I said.. I'll test it in a minute | 20:01 |
| davidlenwell | adam_g: ^ | 20:01 |
| adam_g | davidlenwell, https://review.openstack.org/#/c/215227/15..16/akanda/rug/instance_manager.py | 20:03 |
| adam_g | thats passing the return values from the callback and not the reference to the callback itself | 20:03 |
| davidlenwell | yes.. if you look at what its returning | 20:03 |
| davidlenwell | look at https://review.openstack.org/#/c/215227/16/akanda/rug/drivers/router.py | 20:04 |
| davidlenwell | the return value of that function is the callback function | 20:04 |
| davidlenwell | I think it should work fine .. I'll confirm in a few | 20:05 |
| davidlenwell | just finishing up the other thing | 20:05 |
| adam_g | davidlenwell, k | 20:06 |
| *** cleverdevil has quit IRC | 20:18 | |
| *** cleverdevil has joined #akanda | 20:22 | |
| *** cleverdevil has quit IRC | 20:25 | |
| *** cleverdevil has joined #akanda | 20:27 | |
| *** cleverdevil- has joined #akanda | 21:08 | |
| *** cleverdevil has quit IRC | 21:11 | |
| rods | adam_g so looks like the disk-image-create command is making a copy of the resolv.conf file here https://github.com/openstack/diskimage-builder/blob/master/lib/common-functions#L254 | 21:20 |
| rods | which is called before the extra-data phase https://github.com/openstack/diskimage-builder/blob/master/bin/disk-image-create#L219 | 21:21 |
| rods | making some changes on the file depending on the distribuiton | 21:22 |
| rods | and then restoring the file here https://github.com/openstack/diskimage-builder/blob/master/lib/img-functions#L106 | 21:23 |
| rods | after the finalise phase https://github.com/openstack/diskimage-builder/blob/master/bin/disk-image-create#L279 | 21:23 |
| adam_g | rods, oh nice find. i figured it was something like this | 21:23 |
| rods | looks like a workaround could be to change -L $TMP_MOUNT_PATH/etc/resolv.conf.ORIG file in the finalise phase | 21:24 |
| rods | before it gets resotred | 21:25 |
| *** puranamr has joined #akanda | 21:28 | |
| *** puranamr has quit IRC | 21:33 | |
| *** puranamr has joined #akanda | 21:36 | |
| openstackgerrit | Adam Gandelman proposed stackforge/akanda-rug: Add akanda-pez-service RPC server and client (WIP) https://review.openstack.org/221997 | 21:41 |
| openstackgerrit | Adam Gandelman proposed stackforge/akanda-rug: Split out instance provisioning to on_demand and pez (WIP) https://review.openstack.org/221999 | 21:41 |
| *** puranamr has quit IRC | 22:04 | |
| *** cleverdevil- has quit IRC | 23:01 | |
| *** openstackgerrit has quit IRC | 23:16 | |
| *** openstackgerrit has joined #akanda | 23:16 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!