| *** aspiers has quit IRC | 00:02 | |
| *** whiteveil has quit IRC | 00:02 | |
| *** vnogin has quit IRC | 00:06 | |
| *** manheim has joined #openstack-ansible | 00:15 | |
| *** thorst has joined #openstack-ansible | 00:19 | |
| *** thorst has quit IRC | 00:19 | |
| *** manheim has quit IRC | 00:19 | |
| *** cpuga has quit IRC | 00:48 | |
| *** cpuga has joined #openstack-ansible | 00:49 | |
| *** cpuga has quit IRC | 00:53 | |
| *** schwicht has joined #openstack-ansible | 00:56 | |
| *** schwicht has quit IRC | 01:01 | |
| openstackgerrit | Marc Gariépy proposed openstack/openstack-ansible-tests stable/ocata: Disable PrivateDevices for galera and Memcached https://review.openstack.org/475093 | 01:08 |
|---|---|---|
| *** acormier has joined #openstack-ansible | 01:10 | |
| *** cNilesh has joined #openstack-ansible | 01:23 | |
| *** woodard_ has joined #openstack-ansible | 01:24 | |
| *** woodard has quit IRC | 01:24 | |
| *** cmart has joined #openstack-ansible | 01:48 | |
| *** acormier has quit IRC | 02:05 | |
| *** acormier has joined #openstack-ansible | 02:05 | |
| *** SerenaFeng has joined #openstack-ansible | 02:19 | |
| *** thorst has joined #openstack-ansible | 02:19 | |
| *** SerenaFeng has quit IRC | 02:23 | |
| *** thorst has quit IRC | 02:25 | |
| *** schwicht has joined #openstack-ansible | 02:32 | |
| *** galstrom_zzz is now known as galstrom | 02:40 | |
| *** penguintamer has quit IRC | 02:46 | |
| *** acormier has quit IRC | 03:05 | |
| *** acormier has joined #openstack-ansible | 03:05 | |
| *** markvoelker has quit IRC | 03:07 | |
| *** acormier has quit IRC | 03:09 | |
| *** woodard_ has quit IRC | 03:12 | |
| *** woodard has joined #openstack-ansible | 03:12 | |
| *** woodard has quit IRC | 03:16 | |
| *** ianychoi has quit IRC | 03:18 | |
| *** ianychoi has joined #openstack-ansible | 03:19 | |
| *** askb has quit IRC | 03:21 | |
| *** galstrom is now known as galstrom_zzz | 03:26 | |
| *** galstrom_zzz is now known as galstrom | 03:33 | |
| *** NWOutlier_ has joined #openstack-ansible | 03:43 | |
| *** schwicht has quit IRC | 03:50 | |
| *** bauruine has quit IRC | 03:53 | |
| *** thorst has joined #openstack-ansible | 04:00 | |
| *** galstrom is now known as galstrom_zzz | 04:01 | |
| *** bauruine has joined #openstack-ansible | 04:01 | |
| *** poopcat has quit IRC | 04:01 | |
| *** thorst has quit IRC | 04:04 | |
| *** exodusftw has quit IRC | 04:08 | |
| *** exodusftw has joined #openstack-ansible | 04:10 | |
| openstackgerrit | Merged openstack/openstack-ansible-tests stable/ocata: Disable PrivateDevices for galera and Memcached https://review.openstack.org/475093 | 04:14 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/ocata: Set PrivateDevices=false for Galera https://review.openstack.org/474314 | 04:19 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/ocata: Set PrivateDevices=false for MemcacheD https://review.openstack.org/475101 | 04:21 |
| cloudnull | aindlq: https://launchpad.net/openstack-ansible | 04:21 |
| cloudnull | all of our bug reporting goes on launchpad | 04:22 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/ocata: Set PrivateDevices=false for MemcacheD https://review.openstack.org/475101 | 04:22 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ceph_client stable/ocata: Use the host python interpreter when delegating to localhost https://review.openstack.org/475102 | 04:28 |
| *** lbragstad has quit IRC | 04:32 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-ceph_client stable/newton: Use the host python interpreter when delegating to localhost https://review.openstack.org/475103 | 04:33 |
| odyssey4me | cloudnull you not sleeping either? | 04:36 |
| cloudnull | nope. | 04:36 |
| cloudnull | generally never | 04:36 |
| cloudnull | what you doing? | 04:36 |
| odyssey4me | well, now that we have centos working again, I'm rechecking some things and doing some more fixes | 04:38 |
| odyssey4me | I was busy working on changes to make serialised upgrades for keystone work properly yesterday, and might get back to it. | 04:39 |
| openstackgerrit | Merged openstack/openstack-ansible-os_heat master: Deprecate rpc_backend option https://review.openstack.org/469780 | 04:40 |
| odyssey4me | you? | 04:40 |
| *** bauruine has quit IRC | 04:45 | |
| *** lbragstad has joined #openstack-ansible | 04:46 | |
| cloudnull | working on maas things | 04:52 |
| cloudnull | ive got master(xenial), ocata(xenial), newton(xenial|trusty), | 04:53 |
| cloudnull | those all work | 04:53 |
| cloudnull | mitaka is next | 04:53 |
| cloudnull | then i should be just about done. | 04:53 |
| odyssey4me | eish, that does not sound like fun | 04:53 |
| odyssey4me | I hope you have beer to compensate. | 04:54 |
| *** bauruine has joined #openstack-ansible | 04:55 | |
| *** lbragstad has quit IRC | 04:56 | |
| cloudnull | indeed. | 04:56 |
| cloudnull | many beers | 04:56 |
| odyssey4me | did you see http://lists.openstack.org/pipermail/openstack-dev/2017-June/118417.html ? | 04:57 |
| cloudnull | the last hurdle for me is the jinja2 pins for mitaka/liberty/kilo | 04:58 |
| cloudnull | no. | 04:58 |
| * cloudnull reading | 04:58 | |
| odyssey4me | EmilienM has been seeking feedback in the ML | 04:58 |
| odyssey4me | he's very keen on our response | 04:58 |
| odyssey4me | cloudnull wrt the jinja2 pins - is this the pins for the deployment host, or the target host? | 04:59 |
| cloudnull | deployment host | 04:59 |
| cloudnull | ansble 1.9 is busted with jinja >2.8 | 04:59 |
| cloudnull | we should pin that in the mitaka branch | 05:00 |
| cloudnull | given that we still have it | 05:00 |
| odyssey4me | yeah | 05:00 |
| odyssey4me | although it's EOL, they're actively doing the EOL branching | 05:00 |
| odyssey4me | ah, it's already pinned: https://github.com/openstack/requirements/blob/stable/mitaka/upper-constraints.txt#L7 | 05:01 |
| odyssey4me | and we use that for bootstrapping ansible: https://github.com/openstack/openstack-ansible/blob/stable/mitaka/scripts/bootstrap-ansible.sh#L106 | 05:01 |
| cloudnull | I dont think we use the constraints like we do in newton . | 05:01 |
| cloudnull | gate-check-commit dropped 2.9.x | 05:02 |
| odyssey4me | hmm, then we have a bug | 05:02 |
| cloudnull | indeed | 05:02 |
| odyssey4me | oh no: https://github.com/openstack/requirements/blob/5e111bc5f5423cad6d526a992f6f7d3520a5c1b2/upper-constraints.txt#L186 | 05:03 |
| odyssey4me | that's at the sha we're using for mitaka | 05:03 |
| openstackgerrit | Merged openstack/openstack-ansible-tests master: Make a "test_branch" variable set to master https://review.openstack.org/471440 | 05:03 |
| odyssey4me | oh sorry, that's our master pin | 05:04 |
| odyssey4me | <---- dumbass | 05:04 |
| odyssey4me | this is our mitaka pin: https://github.com/openstack/requirements/blob/2a6816dab055a70747f37a6184b0d0d196a73647/upper-constraints.txt#L7 | 05:05 |
| odyssey4me | so yeah, it should be that version | 05:05 |
| odyssey4me | I see: http://logs.openstack.org/42/466142/1/check/gate-openstack-ansible-openstack-ansible-aio-ubuntu-trusty/e6fed7b/console.html.gz#_2017-05-19_01_44_17_242856 | 05:07 |
| odyssey4me | but then http://logs.openstack.org/42/466142/1/check/gate-openstack-ansible-openstack-ansible-aio-ubuntu-trusty/e6fed7b/console.html.gz#_2017-05-19_01_44_20_752349 | 05:08 |
| *** markvoelker has joined #openstack-ansible | 05:09 | |
| odyssey4me | ah, we do better in master | 05:09 |
| odyssey4me | cloudnull ah yes, https://github.com/openstack/openstack-ansible/commit/c904de2db512728a4cd479fbd72bb19b432d08f4 got it all going right | 05:13 |
| odyssey4me | and we don't have that in mitaka - it's all on the host | 05:13 |
| odyssey4me | we can improve it a little by re-arranging some things | 05:14 |
| odyssey4me | but http://logs.openstack.org/42/466142/1/check/gate-openstack-ansible-openstack-ansible-aio-ubuntu-trusty/e6fed7b/console.html.gz#_2017-05-19_01_44_20_752349 shows it is installing the right thing, even if it installs the wrong thing first | 05:14 |
| openstackgerrit | Merged openstack/openstack-ansible-os_heat stable/ocata: Correct heat developer mode constraint https://review.openstack.org/474720 | 05:15 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_heat stable/newton: Correct heat developer mode constraint https://review.openstack.org/475105 | 05:15 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible stable/mitaka: pin jinja2 https://review.openstack.org/475106 | 05:17 |
| cloudnull | its the simple solution but it should get it done | 05:17 |
| odyssey4me | you won't be able to change the requirements.txt file like that | 05:18 |
| odyssey4me | it'll cause the global requirements bot to change it bac | 05:18 |
| odyssey4me | *back | 05:18 |
| odyssey4me | it's already pinned, as I said above | 05:19 |
| cloudnull | so http://git.openstack.org/cgit/openstack/requirements/plain/upper-constraints.txt?id=2a6816dab055a70747f37a6184b0d0d196a73647 is in fact using the right constraint , but something is dropping 2.9.x | 05:26 |
| odyssey4me | cloudnull from the gate job I linked above, the right version is being used | 05:26 |
| odyssey4me | it drops 2.9 first, then downgrades | 05:26 |
| odyssey4me | my review suggests how you can prevent the downgrade from being necessary | 05:27 |
| odyssey4me | the repo build uses 2.8 according to that gate job | 05:27 |
| cloudnull | https://github.com/openstack/openstack-ansible/blob/stable/mitaka/requirements.txt#L7 | 05:32 |
| cloudnull | so that's why its downgraded | 05:32 |
| cloudnull | but why its breaking swift due to jinja2.9 bugs is the question | 05:32 |
| odyssey4me | hmm, well - is this in the RPC gates or in a clean build? | 05:33 |
| cloudnull | straight osa build | 05:34 |
| cloudnull | clone mitaka run gate-check-commit | 05:34 |
| odyssey4me | because RPC installs ARA, which requires a higher version of jinja | 05:34 |
| odyssey4me | we recently removed that from mitaka/ocata because of this issue | 05:34 |
| odyssey4me | ok, then there's another known issue - let me check something | 05:34 |
| *** phalmos has quit IRC | 05:35 | |
| odyssey4me | https://github.com/openstack/openstack-ansible-os_swift/commit/3591f800e44fc8066008260024bea7159ff5981c | 05:35 |
| cloudnull | does that need to go back into mitak a? | 05:37 |
| odyssey4me | it shouldn't need to | 05:38 |
| odyssey4me | because we shouldn't be using that version of ansible | 05:38 |
| odyssey4me | it doesn't pick cleanly, unfortunately | 05:38 |
| odyssey4me | I though it might be worth a try | 05:38 |
| odyssey4me | maybe it's leftovers from the downgrade | 05:42 |
| *** markvoelker has quit IRC | 05:43 | |
| cloudnull | that's what im thinking | 05:48 |
| cloudnull | make the change in requirements and global requirement pins should solve that | 05:49 |
| cloudnull | **maybe | 05:50 |
| odyssey4me | if you just move the PIP_OPTS thing higher up as suggested in your patch, it should prevent the need for the downgrade and cover more pins | 05:51 |
| *** NWOutlier_ has quit IRC | 05:53 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible stable/mitaka: pin jinja2 https://review.openstack.org/475106 | 05:56 |
| *** armaan has joined #openstack-ansible | 05:56 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_cinder stable/ocata: Optimise the distro package installation https://review.openstack.org/472147 | 06:01 |
| *** thorst has joined #openstack-ansible | 06:01 | |
| *** thorst has quit IRC | 06:05 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_neutron master: Normalise distro package installation https://review.openstack.org/472261 | 06:15 |
| cloudnull | ok. im going to pass out. | 06:17 |
| cloudnull | talk to you all soon. | 06:17 |
| odyssey4me | have a good night's rest cloudnull | 06:18 |
| *** T42X has quit IRC | 06:19 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron stable/ocata: Optimise the distro package installation https://review.openstack.org/475107 | 06:26 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron stable/ocata: Optimise the distro package installation https://review.openstack.org/475107 | 06:28 |
| *** yolanda has joined #openstack-ansible | 06:31 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron stable/newton: Optimise the distro package installation https://review.openstack.org/475109 | 06:33 |
| *** gouthamr has quit IRC | 06:35 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron stable/newton: Optimise the distro package installation https://review.openstack.org/475109 | 06:35 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron stable/newton: Optimise the distro package installation https://review.openstack.org/475109 | 06:36 |
| *** markvoelker has joined #openstack-ansible | 06:40 | |
| *** armaan has quit IRC | 06:55 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-plugins stable/ocata: connection: ssh: Clear environment when connecting to LXC containers https://review.openstack.org/475113 | 07:06 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-plugins stable/newton: connection: ssh: Clear environment when connecting to LXC containers https://review.openstack.org/475114 | 07:06 |
| *** bauruine has quit IRC | 07:11 | |
| *** markvoelker has quit IRC | 07:13 | |
| *** bauruine has joined #openstack-ansible | 07:14 | |
| *** bauruine has quit IRC | 07:29 | |
| *** manheim has joined #openstack-ansible | 07:30 | |
| *** bauruine has joined #openstack-ansible | 07:34 | |
| *** manheim has quit IRC | 07:35 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_keystone master: Implement serialisable rolling upgrade https://review.openstack.org/475115 | 07:39 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_keystone master: Implement serialisable rolling upgrade https://review.openstack.org/475115 | 07:41 |
| *** manheim has joined #openstack-ansible | 07:47 | |
| *** thorst has joined #openstack-ansible | 08:02 | |
| *** manheim has quit IRC | 08:05 | |
| *** thorst has quit IRC | 08:06 | |
| *** armaan has joined #openstack-ansible | 08:07 | |
| *** markvoelker has joined #openstack-ansible | 08:10 | |
| *** gtrxcb has joined #openstack-ansible | 08:12 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Implement rolling upgrades for keystone https://review.openstack.org/469605 | 08:15 |
| *** armaan has quit IRC | 08:16 | |
| *** Manheim has joined #openstack-ansible | 08:18 | |
| *** whiteveil has joined #openstack-ansible | 08:18 | |
| *** whiteveil has quit IRC | 08:24 | |
| *** markvoelker has quit IRC | 08:43 | |
| *** pbandark has joined #openstack-ansible | 08:46 | |
| *** armaan has joined #openstack-ansible | 08:46 | |
| *** gtrxcb has quit IRC | 08:47 | |
| *** bauruine has quit IRC | 08:47 | |
| *** bauruine has joined #openstack-ansible | 08:53 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Implement rolling upgrades for keystone https://review.openstack.org/469605 | 09:24 |
| *** armaan has quit IRC | 09:34 | |
| *** armaan has joined #openstack-ansible | 09:35 | |
| *** armaan has quit IRC | 09:41 | |
| *** armaan has joined #openstack-ansible | 09:41 | |
| *** bauruine has quit IRC | 09:50 | |
| *** bauruine has joined #openstack-ansible | 09:57 | |
| *** bauruine has quit IRC | 09:58 | |
| *** thorst has joined #openstack-ansible | 10:02 | |
| *** bauruine has joined #openstack-ansible | 10:04 | |
| *** vnogin has joined #openstack-ansible | 10:09 | |
| *** pbandark has quit IRC | 10:09 | |
| *** thorst has quit IRC | 10:09 | |
| *** vnogin has quit IRC | 10:13 | |
| *** armaan has quit IRC | 10:31 | |
| *** armaan has joined #openstack-ansible | 10:32 | |
| *** cNilesh has quit IRC | 10:35 | |
| *** cNilesh has joined #openstack-ansible | 10:36 | |
| *** armaan has quit IRC | 10:37 | |
| *** vnogin has joined #openstack-ansible | 10:38 | |
| *** markvoelker has joined #openstack-ansible | 10:41 | |
| *** cNilesh has quit IRC | 10:43 | |
| *** manheim_ has joined #openstack-ansible | 11:05 | |
| *** manheim_ has quit IRC | 11:10 | |
| *** markvoelker has quit IRC | 11:14 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_keystone master: Optimise the distro package installation https://review.openstack.org/471451 | 11:34 |
| *** Manheim has quit IRC | 11:40 | |
| *** lbragstad has joined #openstack-ansible | 11:42 | |
| *** manheim has joined #openstack-ansible | 11:43 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_keystone master: Fix rolling upgrade test https://review.openstack.org/475040 | 11:44 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_keystone master: Implement serialisable rolling upgrade https://review.openstack.org/475115 | 11:44 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_keystone stable/ocata: Optimise the distro package installation https://review.openstack.org/475128 | 11:48 |
| *** kylek3h has joined #openstack-ansible | 11:48 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_keystone master: Deprecate rpc_backend option https://review.openstack.org/469782 | 11:49 |
| *** lbragstad has quit IRC | 11:53 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_keystone stable/newton: Optimise the distro package installation https://review.openstack.org/475129 | 11:54 |
| *** T42X has joined #openstack-ansible | 11:57 | |
| *** thorst has joined #openstack-ansible | 12:06 | |
| *** thorst has quit IRC | 12:10 | |
| *** markvoelker has joined #openstack-ansible | 12:11 | |
| *** manheim has quit IRC | 12:15 | |
| *** manheim has joined #openstack-ansible | 12:22 | |
| *** pbandark has joined #openstack-ansible | 12:25 | |
| *** kylek3h_ has joined #openstack-ansible | 12:25 | |
| *** kylek3h has quit IRC | 12:28 | |
| *** armaan has joined #openstack-ansible | 12:31 | |
| *** markvoelker has quit IRC | 12:45 | |
| *** lenserpgo has quit IRC | 12:48 | |
| *** manheim has quit IRC | 12:48 | |
| *** thorst has joined #openstack-ansible | 13:06 | |
| *** thorst has quit IRC | 13:11 | |
| *** thorst has joined #openstack-ansible | 13:20 | |
| *** vnogin has quit IRC | 13:20 | |
| *** thorst has quit IRC | 13:21 | |
| *** manheim has joined #openstack-ansible | 13:27 | |
| *** vnogin has joined #openstack-ansible | 13:31 | |
| *** hamza has joined #openstack-ansible | 13:31 | |
| *** markvoelker has joined #openstack-ansible | 13:42 | |
| *** vnogin has quit IRC | 13:48 | |
| *** manheim has quit IRC | 14:07 | |
| *** pbandark1 has joined #openstack-ansible | 14:08 | |
| *** pbandark has quit IRC | 14:08 | |
| *** pbandark1 is now known as pbandark | 14:08 | |
| *** manheim has joined #openstack-ansible | 14:09 | |
| *** Mannheim has joined #openstack-ansible | 14:10 | |
| *** manheim has quit IRC | 14:14 | |
| *** markvoelker has quit IRC | 14:16 | |
| *** hamza has quit IRC | 14:44 | |
| *** manheim has joined #openstack-ansible | 14:45 | |
| *** acormier has joined #openstack-ansible | 14:58 | |
| *** SerenaFeng has joined #openstack-ansible | 14:59 | |
| *** acormier_ has joined #openstack-ansible | 15:02 | |
| *** acormier has quit IRC | 15:05 | |
| *** vnogin has joined #openstack-ansible | 15:06 | |
| *** vnogin has quit IRC | 15:06 | |
| *** markvoelker has joined #openstack-ansible | 15:13 | |
| *** thorst has joined #openstack-ansible | 15:21 | |
| *** thorst has quit IRC | 15:25 | |
| *** raginbajin has quit IRC | 15:33 | |
| *** raginbajin has joined #openstack-ansible | 15:43 | |
| *** markvoelker has quit IRC | 15:46 | |
| *** SerenaFeng has quit IRC | 15:52 | |
| *** cpuga has joined #openstack-ansible | 15:56 | |
| *** thorst has joined #openstack-ansible | 15:56 | |
| *** thorst has quit IRC | 15:58 | |
| *** cpuga has quit IRC | 16:00 | |
| *** cNilesh has joined #openstack-ansible | 16:08 | |
| *** acormier_ has quit IRC | 16:10 | |
| *** acormier has joined #openstack-ansible | 16:10 | |
| *** cNilesh has quit IRC | 16:13 | |
| *** acormier has quit IRC | 16:15 | |
| *** acormier has joined #openstack-ansible | 16:18 | |
| *** acormier has quit IRC | 16:18 | |
| *** acormier has joined #openstack-ansible | 16:19 | |
| *** manheim has quit IRC | 16:19 | |
| *** pbandark has quit IRC | 16:33 | |
| *** cNilesh has joined #openstack-ansible | 16:37 | |
| *** acormier has quit IRC | 16:43 | |
| *** armaan has quit IRC | 16:45 | |
| *** acormier has joined #openstack-ansible | 16:57 | |
| *** acormier has quit IRC | 16:57 | |
| *** acormier has joined #openstack-ansible | 16:58 | |
| *** cNilesh has quit IRC | 16:59 | |
| *** openstackgerrit has quit IRC | 17:18 | |
| *** markvoelker has joined #openstack-ansible | 17:43 | |
| *** cpuga has joined #openstack-ansible | 17:57 | |
| *** thorst has joined #openstack-ansible | 17:59 | |
| *** cpuga has quit IRC | 18:02 | |
| *** thorst has quit IRC | 18:04 | |
| *** markvoelker has quit IRC | 18:16 | |
| *** manheim has joined #openstack-ansible | 18:20 | |
| *** woodard has joined #openstack-ansible | 18:20 | |
| *** manheim has quit IRC | 18:24 | |
| *** pcaruana has joined #openstack-ansible | 18:24 | |
| *** pcaruana has quit IRC | 18:25 | |
| *** boxrick1 has quit IRC | 18:48 | |
| *** fdegir has quit IRC | 18:48 | |
| *** rackertom has quit IRC | 18:48 | |
| *** serverascode has quit IRC | 18:48 | |
| *** bgmccollum has quit IRC | 18:48 | |
| *** homerp has quit IRC | 18:48 | |
| *** kencjohnston_ has quit IRC | 18:48 | |
| *** mudpuppy has quit IRC | 18:48 | |
| *** phealy has quit IRC | 18:48 | |
| *** homerp has joined #openstack-ansible | 18:48 | |
| *** bgmccollum_ has joined #openstack-ansible | 18:49 | |
| *** mudpuppy has joined #openstack-ansible | 18:49 | |
| *** phealy has joined #openstack-ansible | 18:49 | |
| *** kencjohnston has joined #openstack-ansible | 18:49 | |
| *** rackertom has joined #openstack-ansible | 18:49 | |
| *** boxrick1 has joined #openstack-ansible | 18:49 | |
| *** serverascode has joined #openstack-ansible | 18:50 | |
| *** acormier has quit IRC | 18:50 | |
| *** fdegir has joined #openstack-ansible | 18:50 | |
| *** acormier has joined #openstack-ansible | 18:50 | |
| *** strattao has quit IRC | 18:51 | |
| *** strattao has joined #openstack-ansible | 18:54 | |
| *** acormier has quit IRC | 18:54 | |
| *** eaxiom has joined #openstack-ansible | 19:06 | |
| *** bauruine has quit IRC | 19:06 | |
| eaxiom | hi all, | 19:06 |
| eaxiom | I am trying to install openstack on two physical nodes. (a compute node and a controller/network node) . Both nodes have two nics installed. | 19:07 |
| eaxiom | I am not using a separate storage network . do i still need to mention a network in the openstack_user_config.yml file? | 19:08 |
| *** khomkrit1499 has joined #openstack-ansible | 19:11 | |
| *** khomkrit1499 has quit IRC | 19:11 | |
| *** bauruine has joined #openstack-ansible | 19:11 | |
| *** markvoelker has joined #openstack-ansible | 19:14 | |
| *** vnogin has joined #openstack-ansible | 19:30 | |
| ivve | eaxiom: depends on how you want to set up your network | 19:32 |
| eaxiom | ivve: I would like to use only two one nic . In that case what configurations should i enable | 19:33 |
| eaxiom | oops. i mean I would like to use only singe nic on both servers | 19:34 |
| ivve | do you want them bonded? | 19:37 |
| ivve | lacp, active/passive? will you use vlans? | 19:37 |
| ivve | or just one flat network with many subnets | 19:37 |
| ivve | or just one subnet for all of it | 19:38 |
| ivve | i guess the correct question would be, how do you want to separate your traffic and what equipment do you have :) | 19:38 |
| eaxiom | yes I will use a single nic and no bonding . and i would like to use a flat network rather than vlans | 19:39 |
| ivve | okay | 19:39 |
| ivve | lvm storage on the compute node then? | 19:39 |
| eaxiom | yes thats correct | 19:39 |
| ivve | then configuration should be quite easy, just use the same CIDR for container, tunnel and storage networks | 19:40 |
| eaxiom | oh okay . so that i means i can duplicate the same cider for all three network types | 19:41 |
| ivve | configure that one network interface on each node to a bridge | 19:41 |
| ivve | yea | 19:41 |
| eaxiom | alright.. | 19:41 |
| ivve | it just has to be large enough | 19:41 |
| eaxiom | its /24 | 19:41 |
| ivve | and then just use that very bridge for all | 19:41 |
| ivve | should be enough | 19:41 |
| ivve | remember to use flat when setting the network stanzas | 19:42 |
| ivve | basically you should be able to use one of the example configs and just add your compute node | 19:42 |
| *** vnogin has quit IRC | 19:43 | |
| eaxiom | sure.. got it .. which one would you recomend .. Appendix A or B please? | 19:43 |
| ivve | A | 19:44 |
| ivve | thats the one i meant | 19:44 |
| ivve | it basically suits your needs quite close | 19:44 |
| ivve | just ignore the vlan stuff | 19:45 |
| eaxiom | Alright will try that.. with modifications (setting same cidr and bridge name for all) | 19:45 |
| eaxiom | okay sure | 19:45 |
| ivve | (you could still use more than one subnet if you wish) | 19:45 |
| eaxiom | vlan and vxlan .. | 19:45 |
| ivve | the example uses a vlan for storage, a vlan for mgmt and a vlan to tunnel vxlans in it | 19:45 |
| ivve | inceptionstyle :) | 19:46 |
| ivve | basically when doing the network config you should need 1 stanza for that interface you are using | 19:46 |
| ivve | meaning 1 for the physical nic + bridge for that nic | 19:47 |
| *** markvoelker has quit IRC | 19:47 | |
| *** eaxiom has quit IRC | 19:49 | |
| *** eaxiom has joined #openstack-ansible | 19:57 | |
| eaxiom | ivve: sorry I got cut off. | 19:58 |
| *** vnogin has joined #openstack-ansible | 19:58 | |
| eaxiom | Let me try this and see how it goes.. | 19:58 |
| eaxiom | and thanks for your support :) | 19:58 |
| ivve | eaxiom: happy to help :) | 19:59 |
| *** armaan has joined #openstack-ansible | 20:05 | |
| eaxiom | ivve: another question : Under the provider networks section , do I need to configure all three networks and set the same bridge name "bridge-mgmt" for all or I can configure just one network ? | 20:06 |
| eaxiom | and set the type as flat | 20:06 |
| *** diliprenkila1 has joined #openstack-ansible | 20:07 | |
| ivve | since you only have 1 nic you should be able to create only 1 network and define type as flat, net_name as flat and put in all the groupbinds in that very network stanza | 20:09 |
| diliprenkila1 | Do any one knows about "msg": "'unicode object' has no attribute 'name' error while installing swift | 20:09 |
| eaxiom | ivve: sure, makes sense.. | 20:11 |
| eaxiom | thanks again | 20:11 |
| ivve | eaxiom: hmm actually i became a bit unsure now.. you might want to have atleast 2 actually | 20:12 |
| ivve | one raw and one flat for your setup | 20:12 |
| eaxiom | you mean the raw for storage ? | 20:14 |
| *** thorst has joined #openstack-ansible | 20:16 | |
| *** thorst has quit IRC | 20:16 | |
| ivve | for mgmt bridge | 20:16 |
| ivve | im having a look at the aio installation | 20:16 |
| ivve | which should be closer to your setup | 20:16 |
| eaxiom | sure .. | 20:17 |
| ivve | i guess you could have all of them and use all 3 bridges | 20:18 |
| ivve | but under the same nic and use flat for all except mgmt which would use raw | 20:18 |
| ivve | i guess that would be easiest | 20:18 |
| ivve | you could still use vxlans too | 20:19 |
| eaxiom | ah. okay . let me look into that | 20:19 |
| ivve | basically use appendix A config, remove the br-vlan which isn't flat | 20:20 |
| ivve | the rest should be fine (except for your network CIDR and of course /etc/network/interfaces config) | 20:21 |
| ivve | then when you have deployed it, creating provider networks would be flat and tenant would still be able to create vxlans that are isolated | 20:21 |
| eaxiom | okay.. lemme try that | 20:21 |
| *** markvoelker has joined #openstack-ansible | 20:44 | |
| eaxiom | here is the update : I tried to use same ip against the storage_host and compute and it gave me an error: | 20:49 |
| eaxiom | Both host:compute1 and host:storage1 have address:10.1.0.6 assigned. Cannot assign same ip to both hosts | 20:49 |
| ivve | yeah just name storage1 to compute1 | 20:49 |
| ivve | or both to host1 or something | 20:49 |
| eaxiom | okay | 20:50 |
| ivve | this setup is built for 3 separate hosts, you have 2 | 20:51 |
| ivve | eaxiom: also, im guessing you read the prereqs.. you need to create the cinder_volume vg manually if im not totally off here | 20:52 |
| ivve | s/cinder_volume/cinder-volumes/ | 20:53 |
| ivve | and since you're using lvm you probably have to do is_metal: true in env.d/cinder.yml | 20:53 |
| *** eaxiom has quit IRC | 20:55 | |
| *** eaxiom has joined #openstack-ansible | 20:58 | |
| eaxiom | ivve: yes, i have configured cinder_volumes on compute1 already . and is_metal is already true by default in playbooks/env.d/cinder.yml | 20:59 |
| ivve | then you should be good to go :) | 21:00 |
| eaxiom | yeps : ) now going to try and see | 21:02 |
| ivve | is it a test or prod with only 1 infra? :D | 21:02 |
| ivve | guess you could add metering and metrics if you are interested in those, they are not in the appendix A config | 21:03 |
| eaxiom | currently it is a test. however we are planning to add more infra hosts and compute hosts into this setup in next stage | 21:03 |
| ivve | just add these: "metering-infra_hosts: , metering-alarm_hosts: , metrics_hosts: , metering-compute_hosts:" | 21:04 |
| ivve | then you want those stanzas | 21:04 |
| eaxiom | sure let me do that | 21:04 |
| ivve | maybe even log_hosts: too | 21:04 |
| ivve | sets up rsyslog for the environemnt | 21:04 |
| ivve | those are the basic services, if you want more there are ofcourse more stuff.. like trove-infra_hosts: or magnum-infra_hosts or dnsaas_hosts ... etc :) | 21:07 |
| ivve | depends on what you want | 21:07 |
| eaxiom | Yes, currently those additional services are not required at this time . maybe at a later stage we will need to integrate more hosts and then those services too | 21:08 |
| *** woodard has quit IRC | 21:08 | |
| ivve | i also want to point out that you might need to do some configuration to your horizon config | 21:08 |
| eaxiom | currently if I can set a basic openstack up and running we are good | 21:08 |
| ivve | with lvm you should be fine though | 21:09 |
| eaxiom | emm. like what kind of customizations. would you please indicate some | 21:09 |
| ivve | there are some bugs with the horizon upload, if you have issues look for 'direct' and change it to 'legacy' | 21:09 |
| ivve | also url upload is disabled by default if thats something you are looking for | 21:09 |
| eaxiom | oh okay .. i ll note this point.. | 21:09 |
| ivve | add this to the local_settings.py | 21:10 |
| ivve | IMAGES_ALLOW_LOCATION = True | 21:10 |
| eaxiom | i have starting running the first playbook | 21:10 |
| ivve | cool | 21:10 |
| ivve | i guess thats pretty much it, then you have to customize as you go | 21:11 |
| eaxiom | currently its progressing.. openstack-ansible setup-hosts.yml | 21:11 |
| ivve | aye if hosts complete usually there are no issue with setup-openstack | 21:12 |
| ivve | however its the longest one | 21:12 |
| ivve | my setup takes a little over 2 hours to complete | 21:12 |
| ivve | and its quite good hardware with all flash :P | 21:12 |
| ivve | all 3 playbook counted | 21:13 |
| eaxiom | yes i can understand :) in fact deploying openstack manually took me 7-8 hours | 21:13 |
| ivve | ouch | 21:13 |
| ivve | i tried that once | 21:14 |
| ivve | doing it on a large environment.. or even one like mine.. i would never dare | 21:14 |
| ivve | especially with all the haproxy shit.. holy configs | 21:14 |
| eaxiom | i created bash scripts of my own to speed up then . it was for openstack juno on centos | 21:14 |
| ivve | ive tried devstack & rdo & packstack | 21:15 |
| ivve | rdo deployment method is really wierd | 21:15 |
| ivve | makes no sense | 21:15 |
| eaxiom | that saved me a couple of hours . .but i agree with you eventually what i needed was something more reliable so i am trying openstack-ansible | 21:16 |
| ivve | the packstack is nice however but i don't like ovs | 21:16 |
| ivve | devstack is just like packstack but worse :P | 21:16 |
| ivve | and OSA(D) is just so got damn great! | 21:16 |
| eaxiom | i tried bot devstack and rdo none of them suit my needs | 21:16 |
| *** markvoelker has quit IRC | 21:17 | |
| eaxiom | devstack has its own issues | 21:17 |
| eaxiom | you have to unstack before a reboot | 21:17 |
| ivve | wish i had more time so i could help with this project, there is so much to do :) | 21:17 |
| *** manheim has joined #openstack-ansible | 21:17 | |
| eaxiom | yes you are quite right . | 21:17 |
| eaxiom | In fact I also wish to contribute to this as i also like python programming and write some code . | 21:18 |
| ivve | its a great project and the guys working on it deserve tons of cred for their hard (and great) work | 21:18 |
| eaxiom | i wrote a plugin for nagios infact a few plugins . they are up on nagios_exchange | 21:18 |
| eaxiom | agreed fully agreed | 21:18 |
| ivve | cool, well im not a programmer. just sysarch/sysadmin | 21:19 |
| ivve | i usually dont do more than bash/ansible | 21:19 |
| eaxiom | me too . not a hard core developer but a sysadmin who can write some code too :) | 21:19 |
| ivve | but i know some python, i guess i just need to learn libs and i could get going.. if i had the time :) | 21:19 |
| ivve | or if work pushed in that direction of course | 21:20 |
| eaxiom | sure.. it all needs time and consistency otherwise we start forgetting it | 21:20 |
| *** vnogin has quit IRC | 21:20 | |
| ivve | but i work quite lonely with developers, so im surrounded by people who know 10+ languages.. if i need something done i ask them :) | 21:21 |
| eaxiom | oh yes i can feel the comfort you have once you site among a group of coders who answer your questions | 21:22 |
| ivve | its great working with open source software when you have those kind of resources, they can quickly look for and correct bugs | 21:23 |
| eaxiom | Yes, and that saves a lot of time and gives relief .. | 21:25 |
| eaxiom | reduces mental stress.. indeed | 21:26 |
| ivve | yeah after a good amount of years in this profession i've developed a good amount of hate towards proprietary software | 21:27 |
| ivve | not to mention all the stupid licensemodels | 21:28 |
| eaxiom | I agree with you and same feelings here | 21:29 |
| ivve | or even proprietary hardware | 21:29 |
| ivve | yuck :P | 21:29 |
| eaxiom | yes especially the networking devices.. they trap the user | 21:31 |
| ivve | l3/l2 switches & x86 hardware is all you need | 21:31 |
| *** acormier has joined #openstack-ansible | 21:31 | |
| *** acormier has quit IRC | 21:31 | |
| ivve | i mean you could build own switches as well.. but im not sure about that | 21:31 |
| *** acormier has joined #openstack-ansible | 21:32 | |
| eaxiom | yes you can | 21:32 |
| ivve | but firewalls/ids/ips and stuff like that i agree | 21:32 |
| eaxiom | there are software packages like quagga | 21:32 |
| eaxiom | i have got a chance to work on that side | 21:32 |
| eaxiom | you can even build routers | 21:32 |
| ivve | i was quite recently at an interview and apparently they built their own switches at some point | 21:33 |
| eaxiom | you can run your own versions of routing protocols like ospf , rip , bgp etc | 21:33 |
| eaxiom | building switches is even more simple with linux network namespaces | 21:34 |
| eaxiom | you can use default linux bridging or ovs | 21:34 |
| ivve | i dig bridges :) | 21:34 |
| eaxiom | you just need a lot of physical nics | 21:34 |
| ivve | yeah thats the thing. space & power and the hardware below that | 21:35 |
| eaxiom | yes | 21:35 |
| eaxiom | emm. the playbook completed execution . however i got some errors at the end | 21:36 |
| eaxiom | while executingthe task TASK [lxc_container_create : Drop container network file (interfaces)] | 21:36 |
| eaxiom | "Failed to get information on remote file (/proc/30437/root/etc/network/interfaces.d/eth1.cfg): MODULE FAILURE"} | 21:36 |
| *** Mannheim has quit IRC | 21:36 | |
| eaxiom | i think i am not using the right phycial interface name here ? | 21:37 |
| ivve | could you paste the config | 21:37 |
| *** manheim has quit IRC | 21:37 | |
| palendae | https://bugs.launchpad.net/openstack-ansible/+bug/1696802 | 21:37 |
| openstack | Launchpad bug 1696802 in openstack-ansible "lxc-containers-create fails when run on a host with existing containers" [Critical,Confirmed] | 21:37 |
| eaxiom | sure.. | 21:39 |
| ivve | btw which one is it that creates the elk stack containers | 21:39 |
| ivve | its not log_hosts: | 21:39 |
| eaxiom | its the infra1 | 21:39 |
| eaxiom | provider_networks: - network: container_bridge: "br-mgmt" container_type: "veth" container_interface: "eth1" ip_from_q: "container" type: "raw" group_binds: - all_containers - hosts is_container_address: true is_ssh_address: true | 21:39 |
| palendae | eaxiom, See the bug I linked | 21:40 |
| eaxiom | hmm this was not user friendly .. | 21:40 |
| robcresswell | ivve: Horizons legacy upload is pretty limited in some scenarios. | 21:40 |
| ivve | eaxiom: since you ran it twice, did you run the lxc-destroy-containers.xml? | 21:40 |
| eaxiom | palendae: sure. let me have a look | 21:40 |
| eaxiom | no i did not run that .. should that be an issue? | 21:40 |
| ivve | there is a few things that has to be erased if you perform a rerun | 21:41 |
| ivve | there should be a small doc about it | 21:41 |
| ivve | lemme see | 21:41 |
| palendae | It shouldn't be, but there's currently that bug that causes LXC creation to fail that no one's looked at | 21:41 |
| palendae | https://bugs.launchpad.net/openstack-ansible/+bug/1696802 | 21:41 |
| openstack | Launchpad bug 1696802 in openstack-ansible "lxc-containers-create fails when run on a host with existing containers" [Critical,Confirmed] | 21:41 |
| eaxiom | oh okay. | 21:42 |
| ivve | if you get far enough and have made some config errors you could end up with apt proxy, cached stuff in /var/lib/lxc | 21:42 |
| ivve | root/.pip | 21:43 |
| *** manheim has joined #openstack-ansible | 21:43 | |
| *** manheim has quit IRC | 21:43 | |
| ivve | robcresswell: yea but with ceph glance doesn't work with 'direct', at least not in ocata | 21:43 |
| *** manheim has joined #openstack-ansible | 21:43 | |
| ivve | unless i did something wrong | 21:44 |
| eaxiom | i am running openstack-ansible lxc-containers-destroy.yml now | 21:44 |
| eaxiom | that was successful so let me try running the setup hosts again | 21:45 |
| eaxiom | but before i run it , let me change the parameter container_interface: "eth1" to the physical interface on infra1 | 21:45 |
| ivve | i'd love to have an elk stack for all the rsyslogs, jeez its hard to troubleshoot with all the logs the stack creates | 21:46 |
| palendae | We have a teammate currently breaking that role out | 21:46 |
| ivve | container_interface is the interface on the inside of the lxc container | 21:46 |
| palendae | I don't know his status though | 21:46 |
| ivve | eaxiom: you can basically name it to whatever in that network stanze as long as it doesn't conflict with another (afaik) | 21:47 |
| eaxiom | oh okay | 21:47 |
| eaxiom | then let me try to run it again | 21:48 |
| ivve | eaxiom: btw you running stable/ocata or some other branch? | 21:48 |
| ivve | palendae: when you say breaking out, you mean its already in? :) | 21:48 |
| eaxiom | that is an important question : I am using a branch -b 15.1.5 | 21:49 |
| ivve | i think thats the master | 21:49 |
| eaxiom | hmm. what would you suggest | 21:49 |
| ivve | i'd go with the stable/ocata unless you're testing stuff i guess. its a preference | 21:50 |
| palendae | ivve, https://github.com/rcbops/rpc-openstack/pull/2332 | 21:50 |
| eaxiom | sure | 21:50 |
| eaxiom | I will use stable/ocata | 21:51 |
| eaxiom | let me try that now.. | 21:51 |
| ivve | palendae: ah yes i i've seen that.. i was looking at it the other day when troubleshooting some stuff and drowning in logs. then i solved it and went on with other stuff :) | 21:51 |
| ivve | it would totally rock to have an elk in the log_hosts | 21:53 |
| robcresswell | ivve: Oh, maybe, I've not tested it myself. | 21:53 |
| ivve | i had problems uploading images anyway via horizon when using direct, via cli worked like a charm, so i switched to legacy and it worked | 21:57 |
| *** acormier has quit IRC | 21:59 | |
| *** acormier has joined #openstack-ansible | 22:00 | |
| *** cpuga has joined #openstack-ansible | 22:00 | |
| *** acormier_ has joined #openstack-ansible | 22:03 | |
| *** bauruine has quit IRC | 22:03 | |
| *** acormier_ has quit IRC | 22:04 | |
| *** acormier_ has joined #openstack-ansible | 22:04 | |
| *** acormier has quit IRC | 22:04 | |
| *** cpuga has quit IRC | 22:05 | |
| *** bauruine has joined #openstack-ansible | 22:09 | |
| eaxiom | ivve: I think 15.0.0 is the first stable release for ocata.. as per reference : https://docs.openstack.org/developer/openstack-ansible/ | 22:11 |
| eaxiom | and as per the deployment guide 15.1.5 is the latest stable release | 22:13 |
| eaxiom | https://docs.openstack.org/project-deploy-guide/openstack-ansible/ocata/deploymenthost.html#configure-the-operating-system-ubuntu | 22:13 |
| ivve | alright | 22:14 |
| eaxiom | so I think i am using the right version | 22:15 |
| ivve | agreed | 22:15 |
| eaxiom | hmm. cool | 22:15 |
| eaxiom | trying to bootstrap now | 22:16 |
| *** thorst has joined #openstack-ansible | 22:17 | |
| *** thorst has quit IRC | 22:22 | |
| *** manheim has quit IRC | 22:24 | |
| eaxiom | bootstrap was successful . now trying to run the setup hosts playbook | 22:25 |
| *** armaan has quit IRC | 22:27 | |
| ivve | cool | 22:28 |
| ivve | im off to sleep, good luck with the deploy! | 22:29 |
| eaxiom | sure | 22:35 |
| eaxiom | have a nice sleep | 22:35 |
| eaxiom | i ll be trying my luck | 22:35 |
| eaxiom | thanks a alot :) | 22:36 |
| *** markvoelker has joined #openstack-ansible | 23:14 | |
| *** manheim has joined #openstack-ansible | 23:15 | |
| *** manheim has quit IRC | 23:20 | |
| *** gouthamr has joined #openstack-ansible | 23:31 | |
| *** markvoelker has quit IRC | 23:48 | |
| *** thorst has joined #openstack-ansible | 23:48 | |
| *** thorst has quit IRC | 23:52 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!