| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_placement master: [WIP] Create base files to install placement https://review.openstack.org/618820 | 00:12 |
|---|---|---|
| *** cshen has joined #openstack-ansible | 00:21 | |
| *** cshen has quit IRC | 00:25 | |
| *** chhagarw has joined #openstack-ansible | 00:37 | |
| *** chhagarw has quit IRC | 00:39 | |
| *** eumel8 has quit IRC | 00:52 | |
| *** jawad_axd has joined #openstack-ansible | 01:05 | |
| *** jawad_axd has quit IRC | 01:09 | |
| *** jawad_axd has joined #openstack-ansible | 01:26 | |
| *** jawad_axd has quit IRC | 01:30 | |
| *** jawad_axd has joined #openstack-ansible | 02:07 | |
| *** jawad_axd has quit IRC | 02:11 | |
| *** cshen has joined #openstack-ansible | 02:22 | |
| *** cshen has quit IRC | 02:26 | |
| *** jawad_axd has joined #openstack-ansible | 02:49 | |
| *** jawad_axd has quit IRC | 02:53 | |
| *** ianychoi has quit IRC | 02:57 | |
| *** ianychoi has joined #openstack-ansible | 02:57 | |
| *** udesale has joined #openstack-ansible | 03:13 | |
| *** vnogin has joined #openstack-ansible | 04:17 | |
| *** vnogin has quit IRC | 04:21 | |
| *** cshen has joined #openstack-ansible | 04:22 | |
| *** cshen has quit IRC | 04:26 | |
| *** ivve has joined #openstack-ansible | 04:35 | |
| *** radeks__ has joined #openstack-ansible | 04:47 | |
| *** radeks_ has quit IRC | 04:50 | |
| *** radeks__ has quit IRC | 05:00 | |
| *** d3n14l has joined #openstack-ansible | 05:42 | |
| *** pcaruana has quit IRC | 05:44 | |
| *** cshen has joined #openstack-ansible | 06:01 | |
| *** cshen has quit IRC | 06:06 | |
| *** chhagarw has joined #openstack-ansible | 06:23 | |
| *** chandankumar has joined #openstack-ansible | 06:23 | |
| *** chandankumar is now known as chkumar|rover | 06:25 | |
| *** chkumar|rover is now known as chkumar|ruck | 06:25 | |
| *** pcaruana has joined #openstack-ansible | 07:08 | |
| *** rmart04 has joined #openstack-ansible | 07:11 | |
| *** rmart04 has quit IRC | 07:19 | |
| *** cshen has joined #openstack-ansible | 07:24 | |
| *** mma has joined #openstack-ansible | 07:44 | |
| *** mma has quit IRC | 07:46 | |
| *** mma has joined #openstack-ansible | 07:46 | |
| *** rmart04 has joined #openstack-ansible | 07:59 | |
| *** jawad_axd has joined #openstack-ansible | 08:02 | |
| *** rpittau has joined #openstack-ansible | 08:02 | |
| *** rmart04 has quit IRC | 08:11 | |
| *** frippe75 has quit IRC | 08:30 | |
| *** DanyC has joined #openstack-ansible | 08:34 | |
| Miouge | Good morning everyone | 08:35 |
| *** dcdamien has joined #openstack-ansible | 08:47 | |
| *** DanyC has quit IRC | 09:08 | |
| *** DanyC has joined #openstack-ansible | 09:09 | |
| *** DanyC has quit IRC | 09:09 | |
| odyssey4me | dmsimard thank you :) we try | 09:12 |
| odyssey4me | o/ Miouge | 09:12 |
| odyssey4me | Does anyone have any idea why there's suddenly a bunch of timeouts going on in gates? | 09:13 |
| odyssey4me | it seems it's all related to the galera packages - 2 hrs to install the client: http://logs.openstack.org/35/619535/1/gate/openstack-ansible-deploy-aio_metal-opensuse-423/087f55d/job-output.txt.gz#_2018-11-22_14_38_29_504369 | 09:14 |
| *** tosky has joined #openstack-ansible | 09:15 | |
| *** emine__ has joined #openstack-ansible | 09:26 | |
| *** cshen has quit IRC | 09:35 | |
| *** d3n14l has quit IRC | 09:44 | |
| openstackgerrit | MaoyangLiu proposed openstack/openstack-ansible-os_neutron master: 1.modify the http link to https link 2. fix the Chinese punctuation https://review.openstack.org/619688 | 09:46 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-galera_server stable/queens: Use OpenStack infrastructure MariaDB mirrors https://review.openstack.org/619690 | 09:55 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-galera_server stable/pike: Use OpenStack infrastructure MariaDB mirrors https://review.openstack.org/619691 | 09:56 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-galera_client stable/queens: Use OpenStack infrastructure MariaDB mirrors https://review.openstack.org/619694 | 09:57 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-galera_client stable/pike: Use OpenStack infrastructure MariaDB mirrors https://review.openstack.org/619695 | 09:58 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests stable/queens: Add MariaDB infrastructure mirrors https://review.openstack.org/619696 | 10:00 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests stable/pike: Add MariaDB infrastructure mirrors https://review.openstack.org/619697 | 10:01 |
| openstackgerrit | Merged openstack/openstack-ansible master: Add MariaDB infrastructure mirrors https://review.openstack.org/619535 | 10:14 |
| jrosser | i don't seem to be able to use the tempest role tags usefully with playbooks/os-temptest-install.yml | 10:20 |
| jrosser | i just want to rerun the tests but it's a bit all/nothing when using --tags | 10:20 |
| *** cshen has joined #openstack-ansible | 10:22 | |
| *** chkumar|ruck has quit IRC | 10:22 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Move git SHA pins to group_vars https://review.openstack.org/619629 | 10:28 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Add MariaDB infrastructure mirrors https://review.openstack.org/619714 | 10:29 |
| *** chandan_kumar has joined #openstack-ansible | 10:35 | |
| *** chandan_kumar is now known as chkumar|ruck | 10:35 | |
| *** electrofelix has joined #openstack-ansible | 10:38 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/queens: Add MariaDB infrastructure mirrors and disable nested virt https://review.openstack.org/619718 | 10:42 |
| *** mcela1 has joined #openstack-ansible | 10:44 | |
| mcela1 | :jamesdenton We succesfully re-installed everything and used a new "clean" vlan (nodes did not have an interface on this vlan, so br-vlan.1324 was created by neutron). And everything appears to work correclty. On another note, we are having problems with neutron and floating IP's, since they lose connection intermitently. We suspect that this might be a problem with keepalive/HA configuration, because the output of "neutron l3 | 10:47 |
| mcela1 | -agent-list-hosting-router myrouter" shows 3 routers in "active" mode | 10:47 |
| *** udesale has quit IRC | 11:07 | |
| openstackgerrit | Jonathan Herlin proposed openstack/openstack-ansible master: Add a check that keepalived interfaces have IPs https://review.openstack.org/619092 | 11:11 |
| *** ahosam has joined #openstack-ansible | 11:20 | |
| *** flaviosr has joined #openstack-ansible | 11:25 | |
| *** rpittau_ has joined #openstack-ansible | 11:31 | |
| *** rpittau has quit IRC | 11:35 | |
| *** mma has quit IRC | 11:36 | |
| *** mma has joined #openstack-ansible | 11:36 | |
| *** rpittau_ is now known as rpittau | 11:43 | |
| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_placement master: [WIP] Create base files to install placement https://review.openstack.org/618820 | 11:53 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: [docs] Clean up the AIO user story https://review.openstack.org/619522 | 11:53 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Implement AIO changes to allow test usage by roles https://review.openstack.org/551791 | 11:53 |
| nowster | bootstrap-aio.sh still dying at: | 11:59 |
| nowster | TASK [systemd_service : Create TEMP run dir] *********************************** | 11:59 |
| nowster | fatal: [localhost]: FAILED! => {"msg": "'dict object' has no attribute 'interface'"} | 11:59 |
| nowster | Any ideas for tracking this down. | 11:59 |
| nowster | Server is a bare metal Intel core i7. | 11:59 |
| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_placement master: [WIP] Create base files to install placement https://review.openstack.org/618820 | 12:02 |
| *** ahosam has quit IRC | 12:03 | |
| *** cshen has quit IRC | 12:14 | |
| odyssey4me | nowster what distro is that, and is that a fresh host with no changes before the git checkout and running bootstrap-ansible/bootstrap-aio ? | 12:19 |
| *** shardy has joined #openstack-ansible | 12:20 | |
| odyssey4me | nowster it seems likely that it doesn't relate to the task, but instead there's something looking at the current facts and it's not finding the right fact | 12:20 |
| nowster | Fresh host, ubuntu bionic | 12:21 |
| odyssey4me | nowster does the host have eth0 ? | 12:21 |
| nowster | no, it's called enp2s0 | 12:21 |
| nowster | but all the other networking has set itself up | 12:22 |
| nowster | br-vlan, etc. | 12:22 |
| *** cshen has joined #openstack-ansible | 12:22 | |
| odyssey4me | and that's using eth1, right? | 12:22 |
| odyssey4me | (which is a dummy interface) | 12:22 |
| nowster | there's no eth1 | 12:22 |
| odyssey4me | hmm, ok - let me finish up my current patch then I'll take a peek - but I suspect the issue has something to do with all that | 12:24 |
| nowster | I have: | 12:26 |
| nowster | lo, enp2s0, br-vlan, br-storage, br-vxlan, br-mgmt, dummy-lbaas, dummy-dbaas, du | 12:26 |
| nowster | mmy-vlan, dummy-storage, eth14@br-lbaas-veth, br-lbaas-veth@eth14, eth13@br-dbaa | 12:26 |
| nowster | s-veth, br-dbaas-veth@eth13, eth12@br-vlan-veth, br-vlan-veth@eth12, br-lbaas, b | 12:26 |
| nowster | r-dbaas, dummy-vxlan, dummy-mgmt | 12:26 |
| nowster | sorry... lo, enp2s0, br-vlan, br-storage, br-vxlan, br-mgmt, dummy-lbaas, dummy-dbaas, dummy-vlan, dummy-storage, eth14@br-lbaas-veth, br-lbaas-veth@eth14, eth13@br-dbaas-veth, br-dbaas-veth@eth13, eth12@br-vlan-veth, br-vlan-veth@eth12, br-lbaas, br-dbaas, dummy-vxlan, dummy-mgmt | 12:26 |
| *** pcaruana has quit IRC | 12:26 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Implement AIO changes to allow test usage by roles https://review.openstack.org/551791 | 12:31 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Ensure AIO container_tech/install_method vars are namespaced https://review.openstack.org/619233 | 12:36 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: [docs] Clean up the AIO user story https://review.openstack.org/619522 | 12:36 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Implement AIO changes to allow test usage by roles https://review.openstack.org/551791 | 12:36 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Implement AIO changes to allow test usage by roles https://review.openstack.org/551791 | 12:41 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Move git SHA pins to group_vars https://review.openstack.org/619629 | 12:44 |
| odyssey4me | mnaser morning - could I get eyes on https://review.openstack.org/619691 & https://review.openstack.org/619697 too please | 12:49 |
| odyssey4me | also https://review.openstack.org/619714 | 12:49 |
| odyssey4me | jrosser evrardjp are either of you around? or any other cores? | 12:50 |
| jrosser | Sort of :) | 12:50 |
| *** noonedeadpunk has quit IRC | 12:50 | |
| *** noonedeadpunk[h] has joined #openstack-ansible | 12:50 | |
| odyssey4me | jrosser it'd be helpful to get reviews for https://review.openstack.org/#/q/topic:osa-mariadb-mirrors+status:open asap to unblock queens/pike from all the timeouts due to slow mariadb mirrors | 12:51 |
| *** noonedeadpunk[h] is now known as noonedeadpunk | 12:52 | |
| odyssey4me | nowster ok, I've kicked off a vagrant-based bionic image and have confirmed it has enp0s3 instead of eth0 - so hopefully I hit the same issue and can work it out | 12:55 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Add bionic to Vagrantfile https://review.openstack.org/619727 | 12:57 |
| chkumar|ruck | jrosser, Hello | 12:59 |
| chkumar|ruck | jrosser, http://logs.openstack.org/36/614336/3/check/openstack-ansible-functional-ubuntu-bionic/c29dc71/logs/ara-report/result/0c446a67-a6eb-434d-87c3-04e8bddace3d/ | 13:00 |
| chkumar|ruck | jrosser, here barbican tempest plugin is there | 13:00 |
| chkumar|ruck | jrosser, but if we check distro jobs | 13:00 |
| chkumar|ruck | jrosser, http://logs.openstack.org/36/614336/3/check/openstack-ansible-functional-distro_install-ubuntu-bionic/312df80/logs/ara-report/result/b418aed2-7a67-4a04-b2a4-55569df06b97/ | 13:00 |
| chkumar|ruck | jrosser, no plugin is found here | 13:00 |
| chkumar|ruck | jrosser, so no test discovered so regex passed to tempest | 13:01 |
| chkumar|ruck | so no tempest run | 13:01 |
| chkumar|ruck | jrosser, it is pretty much expected | 13:01 |
| odyssey4me | chkumar|ruck jrosser yes, for ubuntu there are no plugin packages - so we actually need to make all ubuntu distro jobs use the source based tempest install | 13:01 |
| chkumar|ruck | jrosser, on distro side we need to add barbican-tempest-plugin package | 13:01 |
| odyssey4me | https://github.com/openstack/openstack-ansible-os_tempest/blob/master/vars/ubuntu.yml#L29-L33 | 13:02 |
| chkumar|ruck | jrosser, and rest is ok | 13:02 |
| chkumar|ruck | odyssey4me, can I change here tempest_install_method to source | 13:02 |
| chkumar|ruck | will it overrides the value? | 13:02 |
| odyssey4me | chkumar|ruck where? | 13:03 |
| chkumar|ruck | odyssey4me, https://github.com/openstack/openstack-ansible-os_tempest/blob/master/vars/ubuntu.yml#L29 | 13:04 |
| chkumar|ruck | ? | 13:05 |
| chkumar|ruck | or we need to handle overrides some where else? | 13:05 |
| odyssey4me | chkumar|ruck yeah, I'm looking to see what'd be best | 13:05 |
| chkumar|ruck | let me push a patch and see how it goes | 13:06 |
| odyssey4me | I guess that might work. | 13:06 |
| openstackgerrit | Chandan Kumar proposed openstack/openstack-ansible-os_tempest master: Use tempest_install_method to source for ubuntu distro job https://review.openstack.org/619729 | 13:08 |
| chkumar|ruck | jrosser, ^^ | 13:08 |
| jrosser | Cool, thanks for taking a look | 13:09 |
| odyssey4me | ah, it won't actually help for the role test though, because extra vars have a higher precedence that anything, and the distro install is being set via an extra var | 13:09 |
| chkumar|ruck | jrosser, odyssey4me https://review.openstack.org/#/c/619024/ | 13:09 |
| odyssey4me | so https://github.com/openstack/openstack-ansible-tests/blob/master/test-distro_install-vars.yml#L54 would have to be set via some jinja which sets it to source for ubuntu, but distro for all others | 13:10 |
| odyssey4me | lemme do a patch for that quick | 13:10 |
| chkumar|ruck | odyssey4me, sure | 13:10 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests master: Always use source installs for tempest on ubuntu https://review.openstack.org/619730 | 13:14 |
| odyssey4me | that should do the trick | 13:14 |
| evrardjp | odyssey4me: hey I am there | 13:18 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Always use source installs for tempest on ubuntu https://review.openstack.org/619732 | 13:19 |
| evrardjp | odyssey4me: should I update the a-r-r for rocky with the new galera_server role? | 13:23 |
| odyssey4me | evrardjp no need - those all merged before rocky was cut | 13:23 |
| *** sep has quit IRC | 13:24 | |
| evrardjp | it's okay | 13:24 |
| evrardjp | yeah | 13:24 |
| evrardjp | I saw it | 13:24 |
| evrardjp | changed my bad review with a better one | 13:24 |
| evrardjp | thanks there | 13:24 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Implement AIO changes to allow test usage by roles https://review.openstack.org/551791 | 13:26 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Implement AIO changes to allow test usage by roles https://review.openstack.org/551791 | 13:27 |
| odyssey4me | evrardjp tyvm | 13:27 |
| odyssey4me | jrosser also tyvm | 13:28 |
| *** sep has joined #openstack-ansible | 13:30 | |
| odyssey4me | nowster odd, I don't get any failures | 13:31 |
| chkumar|ruck | odyssey4me, https://github.com/openstack/openstack-ansible-os_tempest/blob/master/defaults/main.yml#L19 | 13:37 |
| odyssey4me | chkumar|ruck yes? | 13:37 |
| chkumar|ruck | odyssey4me, let's depend on tempest binary for everything instead of stestr? | 13:37 |
| odyssey4me | chkumar|ruck https://github.com/openstack/openstack-ansible-os_tempest/commit/74f49fc12f3cd5a67cb05aa6cff0fa12c3855e74 | 13:37 |
| chkumar|ruck | tempest run --subunit also generates subunit result | 13:37 |
| odyssey4me | chkumar|ruck oh really? if that's the case, then yeah - go ahead and patch it to remove that extra variable | 13:38 |
| chkumar|ruck | odyssey4me, since there will be no support for tempest from ubuntu side there we are good to go | 13:38 |
| *** mcela1 has left #openstack-ansible | 13:39 | |
| nowster | odyssey4me: any ideas where to look for where it's trying to retrieve that info? | 13:39 |
| odyssey4me | nowster I'm guessing it's cached facts, but I have no idea what's trying to use it - which would be handy. | 13:40 |
| odyssey4me | nowster if you remove the /etc/openstack_deploy/ansible_facts folder, then try bootstrapping again, does it work? | 13:41 |
| nowster | I'm going to start from bare metal again. | 13:43 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Implement AIO changes to allow test usage by roles https://review.openstack.org/551791 | 13:56 |
| *** speedmann has joined #openstack-ansible | 13:58 | |
| nowster | odyssey4me: did "rm -rf ansible_facts"... still fails. | 13:58 |
| odyssey4me | nowster ok, that's odd - is python installed on the host? I wonder if there's some sort of dependency missing | 13:59 |
| openstackgerrit | Merged openstack/openstack-ansible-tests stable/queens: Add MariaDB infrastructure mirrors https://review.openstack.org/619696 | 14:00 |
| nowster | # python -V | 14:00 |
| nowster | Python 2.7.15rc1 | 14:00 |
| nowster | In the facts dir... localhost has: | 14:01 |
| nowster | "ansible_default_ipv4": {}, | 14:01 |
| nowster | "ansible_default_ipv6": {}, | 14:01 |
| odyssey4me | oh, that'd be a problem | 14:01 |
| mgariepy | http://logs.openstack.org/54/619354/1/check/openstack-ansible-deploy-aio_distro_lxc-ubuntu-bionic/372515b/logs/openstack/aio1_cinder_api_container-d64d2f9a/cinder/cinder-volume.log.txt.gz#_2018-11-22_21_30_55_280 | 14:01 |
| mgariepy | cinder issue on rocky ? | 14:01 |
| odyssey4me | nowster does the host have a default route set? | 14:01 |
| nowster | yes | 14:01 |
| nowster | default via 10.35.7.254 dev enp2s0 proto dhcp src 10.35.6.147 metric 100 | 14:02 |
| odyssey4me | nowster hmm, I wonder why ansible is not seeing that as the default address then | 14:02 |
| odyssey4me | nowster I think you should be able to work around that by setting this variable when running the bootstrap: https://github.com/openstack/openstack-ansible/blob/master/tests/roles/bootstrap-host/defaults/main.yml#L206 | 14:03 |
| odyssey4me | nowster export BOOTSTRAP_OPTS="bootstrap_host_public_address=<the address usable for access from outside the box>" | 14:04 |
| odyssey4me | I suspect that it doesn't allow a dhcp address to be used - it likely needs to be static | 14:05 |
| odyssey4me | mgariepy well, that looks like some clean up is required | 14:05 |
| mgariepy | https://review.openstack.org/#/c/617053/2 | 14:05 |
| mgariepy | found it. | 14:05 |
| mgariepy | i guess it needs to merge also ;) | 14:06 |
| nowster | odyssey4me: maybe it doesn't like rfc1918 addresses | 14:06 |
| odyssey4me | oh dear http://logs.openstack.org/14/619714/1/check/openstack-ansible-deploy-distro_ceph-opensuse-423/f25435d/job-output.txt.gz#_2018-11-23_11_57_27_135387 | 14:08 |
| *** chkumar|ruck has quit IRC | 14:15 | |
| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_placement master: [WIP] Create base files to install placement https://review.openstack.org/618820 | 14:22 |
| odyssey4me | cores could I get some eyes on https://review.openstack.org/619218 to remove some cruft ? | 14:22 |
| *** jawad_axd has quit IRC | 14:26 | |
| openstackgerrit | Merged openstack/openstack-ansible-tests stable/pike: Add MariaDB infrastructure mirrors https://review.openstack.org/619697 | 14:29 |
| mnaser | im stuck on https://review.openstack.org/#/c/608433/ and not sure why it's not able to access the vip | 14:30 |
| mnaser | help appreciated :) i'll try to spin up a vm with it soon but if anyone sees something glaring.. | 14:30 |
| *** jawad_axd has joined #openstack-ansible | 14:32 | |
| nowster | odyssey4me: if I run "ansible -c local -m setup localhost" it gets the right info | 14:32 |
| odyssey4me | nowster awesome :) | 14:32 |
| nowster | But why doesn't it get it ordinarily? | 14:33 |
| odyssey4me | nowster what did you change to make it work? | 14:33 |
| nowster | and then when you re-run bootstrap-aio.sh it fails again | 14:33 |
| nowster | and has wiped the facts | 14:33 |
| nowster | nothing was changed to get ansible guessing correctly | 14:34 |
| jrosser | mnaser: have you had to deal with this? nova.compute.manager [instance: 2159468e-f5bd-4330-abde-b6e66835f5f2] UEFINotSupported: UEFI is not supported | 14:34 |
| odyssey4me | hmm, ok - check the cached facts to see whether it has the same contents? | 14:34 |
| mnaser | jrosser: eh yes, you have to use a custom image to boot stuff on arm | 14:34 |
| jrosser | doh | 14:35 |
| nowster | odyssey4me: in /etc/openstack_deploy/ansible_facts ? | 14:35 |
| mnaser | is there a cirros arm image? | 14:35 |
| jrosser | thats what happens with cirros arm64 | 14:35 |
| jrosser | unless theres something else that needs fixing too...... | 14:35 |
| *** jawad_axd has quit IRC | 14:36 | |
| odyssey4me | nowster yep, there should be a file per host with the facts cached in each file | 14:41 |
| nowster | odyssey4me: but if I've wiped that directory? | 14:42 |
| odyssey4me | nowster run the bootstrap-aio again and it should recreate it | 14:42 |
| nowster | yes, and it creates it with the empty default.ipv4 | 14:42 |
| odyssey4me | nowster ok, now run ansible -m setup localhost - it should update the cached facts | 14:43 |
| nowster | done that... and bootstrap-aio.sh overwrites them with the wrong ones | 14:43 |
| odyssey4me | ugh, 'ansible-playbook' gathers facts differently when set to in the play, versus the setup module | 14:44 |
| odyssey4me | clearly it's doing something different, and it's not meeting some condition which results in the fact not being populated | 14:45 |
| odyssey4me | could you try and set that interface to be a static address? or override the bootstrap-host role default as suggested? | 14:45 |
| nowster | actually, that worked. | 14:45 |
| odyssey4me | yeah, unfortunately playbook fact gathering is not as good as using the setup module in a task | 14:46 |
| nowster | running bootstrap-aio.sh, having it fail, then "ansible -c local -m setup localhost", then boostrap-aio.sh again, works | 14:46 |
| nowster | odyssey4me: pointer to the playbook stuff and I'll investigate? | 14:47 |
| odyssey4me | nowster it's an ansible bug - not an OSA bug | 14:47 |
| nowster | oh, right. | 14:47 |
| odyssey4me | There are a few bugs around the difference between running the setup module and the gather_facts option. It relates to using smart gathering IIRC. All in all, it's one of the frustrations of inconsistencies. | 14:48 |
| odyssey4me | A workaround could be to have the playbook use the setup module prior to running the role. | 14:48 |
| odyssey4me | ie add pre_tasks here https://github.com/openstack/openstack-ansible/blob/master/tests/bootstrap-aio.yml#L20 with a task that runs the setup module | 14:49 |
| *** cshen has quit IRC | 14:49 | |
| odyssey4me | using this gather_subset argument: https://github.com/openstack/openstack-ansible/blob/master/scripts/gate-check-commit.sh#L143 | 14:50 |
| *** mma has quit IRC | 14:50 | |
| odyssey4me | the argument just reduces the time taken by excluding unnecessary fact gathering (ohai & facter) | 14:50 |
| openstackgerrit | Merged openstack/openstack-ansible-galera_client stable/pike: Use OpenStack infrastructure MariaDB mirrors https://review.openstack.org/619695 | 14:51 |
| openstackgerrit | Merged openstack/openstack-ansible-galera_client stable/queens: Use OpenStack infrastructure MariaDB mirrors https://review.openstack.org/619694 | 14:51 |
| *** mma has joined #openstack-ansible | 14:51 | |
| nowster | jrosser: Have now got (with your patches and suggestions) to the xml library missing failure point | 14:53 |
| nowster | probably missing "python-lxml" somewhere | 14:54 |
| jrosser | nowster: ok - i've not addressed that yet so just apt installed libxslt1.1 into the utility container | 14:54 |
| jrosser | lxc-attach to the container, apt to install and it'll get a bit further | 14:55 |
| jrosser | did you see my question to mnaser earlier about the UEFI images? that will be the next thing to fail | 14:55 |
| *** mma has quit IRC | 14:56 | |
| *** rgogunskiy has quit IRC | 15:04 | |
| openstackgerrit | Merged openstack/openstack-ansible-galera_server stable/queens: Use OpenStack infrastructure MariaDB mirrors https://review.openstack.org/619690 | 15:10 |
| nowster | jrosser: ta | 15:12 |
| nowster | not spotted that one (UEFI) | 15:12 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/queens: Add MariaDB infrastructure mirrors and disable nested virt https://review.openstack.org/619718 | 15:12 |
| openstackgerrit | Merged openstack/openstack-ansible-galera_server stable/pike: Use OpenStack infrastructure MariaDB mirrors https://review.openstack.org/619691 | 15:13 |
| nowster | erk... Cannot initiate the connection to ubuntu-cloud.archive.canonical.com:80 (2001:67c:1360:8c01::23). - connect (101: Network is unreachable)" | 15:13 |
| nowster | this is x86-64 with no ipv6 connectivity | 15:14 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/pike: Fix ansible-sshd version https://review.openstack.org/618995 | 15:18 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/pike: Fix ansible-sshd version https://review.openstack.org/618995 | 15:19 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/pike: Fix ansible-sshd version https://review.openstack.org/618995 | 15:24 |
| odyssey4me | cores - it'd be good to get reviews for https://review.openstack.org/618995 asap as it's a gate unblocker for pike | 15:25 |
| *** cshen has joined #openstack-ansible | 15:27 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Move git SHA pins to group_vars https://review.openstack.org/619629 | 15:28 |
| *** jawad_axd has joined #openstack-ansible | 15:30 | |
| odyssey4me | evrardjp will https://review.openstack.org/#/c/619629/5 be ok for your release tooling? | 15:31 |
| *** jawad_axd has quit IRC | 15:33 | |
| *** jawad_axd has joined #openstack-ansible | 15:34 | |
| evrardjp | odyssey4me: this will require slight adaptations, but I have bigger concerns by adding all those to group vars | 15:34 |
| odyssey4me | evrardjp like? | 15:34 |
| odyssey4me | (bearing in mind that this is an interim measure to unblock getting rid of repo_build) | 15:35 |
| evrardjp | well here those variables were loaded only when needed, so workers didn't have to carry those | 15:35 |
| evrardjp | yeah | 15:35 |
| evrardjp | well | 15:35 |
| *** shardy has quit IRC | 15:36 | |
| odyssey4me | I can turn some attention to killing more of the other group vars to compensate if this is a concern. | 15:36 |
| evrardjp | I guess my question is what are the next steps -- I am curious about this temp step | 15:38 |
| evrardjp | odyssey4me: I thought we could directly remove -- let's say glance sha -- from the integrated, with a depends on the sha in the glance role | 15:39 |
| odyssey4me | well, with this in place, the roles can simply use those vars - so repo build should be able to go away entirely... all I need is to change the integrated repository's pip arguments | 15:39 |
| odyssey4me | getting the in-role pinning done is going to take quite a bit of time, and I'm not sure that's going to get done by release | 15:40 |
| evrardjp | why not? What's the issue? | 15:40 |
| evrardjp | I mean we can move role per role | 15:40 |
| evrardjp | I thought that was the plan | 15:40 |
| odyssey4me | I don't have enough time to work through them all in time for release. | 15:40 |
| evrardjp | but you don't have to be alone | 15:41 |
| odyssey4me | Doing this allows each role to do its builds, and us to remove repo build and the py pkgs plugin... then we can move the pinning into the roles as a next step. | 15:41 |
| evrardjp | as long as we have an example with one role, that gets removed from the integrated, and mvoed to role | 15:41 |
| evrardjp | ok that's the thing, I thought this was the other way around | 15:41 |
| evrardjp | we moved pins towards roles, slowly, removing the corresponding things of integrated | 15:42 |
| evrardjp | and then, we nothing remains, remove repo build and py_pkgs | 15:42 |
| evrardjp | I guess it's all about the same | 15:42 |
| odyssey4me | I think this will get better results more quickly. | 15:42 |
| odyssey4me | It also helps us to switch to using the integrated build for role tests more quickly. | 15:42 |
| evrardjp | I don't see the speed gain there, but I might not have your oversight/vision | 15:43 |
| evrardjp | I will adapt the tooling anyway :) | 15:44 |
| odyssey4me | well, I'm also kinda just feeling this out as I go - this way seemed a simple way to make progress | 15:44 |
| odyssey4me | the in-role pinning is going to involve more difficult work which also involves proposal bot things which I'm not entirely sure how best to tackle | 15:45 |
| odyssey4me | as far as I can see, this is an easier change to follow | 15:45 |
| evrardjp | proposal bot is the automation of the process, like we'd do proposal bot for bumping shas | 15:48 |
| evrardjp | I'd not take that into account for now | 15:48 |
| evrardjp | separate problem domains :) | 15:48 |
| odyssey4me | basically I'm saying that starting in the roles with this is harder than starting here as far as I can see. I'll run a trial now, but I think with this and some more group_var changes on top, repo_build is dead. | 15:51 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible master: Move git SHA pins to group_vars https://review.openstack.org/619629 | 15:51 |
| evrardjp | odyssey4me: yeah but you have to unravel all the pain of the removal of repo_build, because of its intertwined nature | 15:56 |
| odyssey4me | yep, working on that now | 15:56 |
| evrardjp | if you want I can work on the bumping of the shas in the roles, as this is not dependant of your work | 15:56 |
| odyssey4me | I think a few bits will have to be done before that. | 15:58 |
| evrardjp | I guess that's the part I don't get :) | 15:58 |
| odyssey4me | 1. we have to wean the roles away from using git on the repo server, so the 'smart-sources' patches cloudnull started need to be finalised and merged so that we're using the wheel contents rather than fetching from git | 15:59 |
| *** pcaruana has joined #openstack-ansible | 16:00 | |
| evrardjp | what I'd do is ... 1) move role x to be developer_mode: True by default (and therefore only build locally), 2) match shas between integrated and role, 3) remove all mention of developer_mode, 4) repeat | 16:01 |
| evrardjp | as 1) 2) 3) is effectively the role tests IIRC | 16:01 |
| odyssey4me | 2. we need to decide how important offline source-based installs are for now... it's easier to leave that out initially, then focus on it later... otherwise we have to work out a process for getting all the git repositories and putting them onto the repo server in the python_venv_build role or somewhere else, and how to switch between using that or going online | 16:01 |
| evrardjp | I guess offline source-based installs don't matter anymore | 16:01 |
| odyssey4me | 3. we have to make developer mode usable in the new world | 16:02 |
| odyssey4me | I've got a plan written out for developer mode. I just need to execute on it. | 16:03 |
| evrardjp | that's the thing, I have the impression it's a lot of work for a few people again, and I just don't want burnout of people due to stresses of bringing broken things | 16:03 |
| *** jawad_axd has quit IRC | 16:04 | |
| evrardjp | well I'm no ptl anymore though, so I shouldn't judge that :p | 16:04 |
| evrardjp | I just personally care :p | 16:04 |
| *** dave-mccowan has joined #openstack-ansible | 16:05 | |
| odyssey4me | well, lemme see how this patch plays out in the gate - I'll switch back to working on developer mode for a role to see how my plan plays out | 16:05 |
| odyssey4me | I have the feeling that this is all going to have to come together anyway. | 16:05 |
| odyssey4me | I switched focus to the integrated build to try and get it to be easier to use for role tests - and I think that's largely done now. I just need to do a trial test for one of the roles to see how well it works. | 16:06 |
| evrardjp | maybe the big bang is not too bad here | 16:06 |
| odyssey4me | speaking of which https://review.openstack.org/#/c/551791/41 just passed :) | 16:09 |
| odyssey4me | it'd be good to get eyes on that, along with https://review.openstack.org/#/c/619522/5 | 16:10 |
| *** blinkiz has quit IRC | 16:20 | |
| *** jawad_axd has joined #openstack-ansible | 16:24 | |
| *** jawad_axd has quit IRC | 16:27 | |
| *** jawad_axd has joined #openstack-ansible | 16:27 | |
| nowster | I now have master running on x86_64. Had to re-run some of the playbooks due to not failing over from IPv6 to IPv4 correctly. | 16:28 |
| noonedeadpunk | I try to configure a QoS for neutron and it seems, that qos extension is present for linuxbridge plugin only on compute node https://github.com/openstack/openstack-ansible-os_neutron/blob/master/templates/plugins/ml2/linuxbridge_agent.ini.j2#L30, while https://docs.openstack.org/neutron/rocky/admin/config-qos.html#configuration says it should present both on compute and networking node | 16:28 |
| noonedeadpunk | Does anyone have working QoS with linuxbridge agent? | 16:29 |
| *** DanyC has joined #openstack-ansible | 16:30 | |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible-os_neutron master: Add port_forwarding to neutron configuration. https://review.openstack.org/619640 | 16:33 |
| mgariepy | mbuil, added the link. | 16:33 |
| *** rpittau has quit IRC | 16:45 | |
| *** dave-mccowan has quit IRC | 16:45 | |
| *** cshen has quit IRC | 16:46 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance master: [WIP] Add integrated deploy/upgrade jobs https://review.openstack.org/619765 | 16:46 |
| *** cshen has joined #openstack-ansible | 16:48 | |
| *** DanyC has quit IRC | 16:49 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance master: [WIP] Try using a dynamic scenario https://review.openstack.org/619768 | 16:53 |
| noonedeadpunk | Also "append['qos']" looks pretty strange for me... I thought, that "()" must be used for jinja append | 16:53 |
| openstackgerrit | Merged openstack/openstack-ansible master: Remove unnecessary octavia scenario AIO bootstrap https://review.openstack.org/619218 | 16:53 |
| *** DanyC has joined #openstack-ansible | 16:54 | |
| openstackgerrit | Dmitriy Rabotjagov (noonedeadpunk) proposed openstack/openstack-ansible-os_neutron master: Fix linuxbridge agent extensions https://review.openstack.org/619771 | 17:04 |
| *** DanyC has quit IRC | 17:06 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance master: [WIP] Try using a dynamic scenario https://review.openstack.org/619768 | 17:06 |
| odyssey4me | cores - https://review.openstack.org/#/c/619233/ is ready for review, which helps us move towards using the integrated repo for all tests | 17:21 |
| odyssey4me | also https://review.openstack.org/619522 which updates the AIO user story to make a bit more sense | 17:22 |
| odyssey4me | this whole chain would be good to get eyes on https://review.openstack.org/#/q/topic:use-integrated-repo-for-all-tests+(status:open+OR+status:merged) | 17:22 |
| *** rmart04 has joined #openstack-ansible | 17:34 | |
| odyssey4me | mnaser just did a test patch for glance role testing using the integrated repo, and https://review.openstack.org/#/c/619765/1 has done all the right things (just installs keystone, haproxy, glance) and passed tempest :) | 17:34 |
| mnaser | odyssey4me: that is so beyond exciting | 17:35 |
| mnaser | yay for not breaking things | 17:35 |
| *** emine has joined #openstack-ansible | 17:35 | |
| odyssey4me | so yeah, I'll work some more on it next week - but it's a good start... I'm stoked that it's working for a basic test at least | 17:37 |
| odyssey4me | much the same run time as the current role tests too, which is good :) | 17:37 |
| *** Adri2000 has quit IRC | 17:38 | |
| *** emine__ has quit IRC | 17:38 | |
| *** jmccrory has quit IRC | 17:38 | |
| odyssey4me | so yeah - please review what's up - anything from here on will rely on what's up being ok for everyone | 17:39 |
| odyssey4me | time to w/end! :) | 17:39 |
| *** Adri2000 has joined #openstack-ansible | 17:39 | |
| *** jmccrory has joined #openstack-ansible | 17:39 | |
| mnaser | odyssey4me: have a good evening | 17:43 |
| *** chhagarw has quit IRC | 17:48 | |
| *** dcdamien has quit IRC | 17:55 | |
| mnaser | do we have a pattern for checking if a service is deployed | 18:01 |
| mnaser | i.e. i want to have magnum set `magnum_cert_manager_type` to `barbican` if barbican is deployed | 18:02 |
| logan- | there should be an example in most of the core services using local facts | 18:03 |
| *** jawad_axd has quit IRC | 18:06 | |
| *** rmart04 has quit IRC | 18:09 | |
| *** rmart04 has joined #openstack-ansible | 18:10 | |
| *** rmart04 has quit IRC | 18:27 | |
| *** jawad_axd has joined #openstack-ansible | 18:31 | |
| noonedeadpunk | probably oftopic question, but I've just noticed, that neutron dnsmasq replies on DNS recursive requests from worldwide on Rocky/ Does this problem occur only for me (due to some bug in neutron itself), or it's a common issue? | 18:31 |
| *** hamzaachi has joined #openstack-ansible | 18:49 | |
| *** jawad_axd has quit IRC | 18:59 | |
| *** dcdamien has joined #openstack-ansible | 19:00 | |
| *** DanyC has joined #openstack-ansible | 19:06 | |
| *** electrofelix has quit IRC | 19:06 | |
| *** DanyC has quit IRC | 19:10 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Remove unnecessary octavia scenario AIO bootstrap https://review.openstack.org/619792 | 19:18 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Remove unnecessary octavia scenario AIO bootstrap https://review.openstack.org/619792 | 19:18 |
| jamesdenton | noonedeadpunk - https://bugs.launchpad.net/neutron/+bug/1501206 | 19:19 |
| openstack | Launchpad bug 1501206 in neutron "router:dhcp ports are open resolvers" [High,In progress] - Assigned to Dr. Jens Harbott (j-harbott) | 19:19 |
| noonedeadpunk | Thanks jamesdenton... But it seems, that patch stuck somewhere on it's way, and it's not going to change in the nearest future... | 19:21 |
| jamesdenton | yeah it's been around a while | 19:21 |
| noonedeadpunk | Is there some normal workaround this stuff? As by far the only way, which I found during quick research, is to set no-resolv in dnsmasq-neutron.conf, and dnsmasq_local_resolv=True in dhcp_agent.ini (with commenting out dnsmasq_dns_servers), but I think I've broke smth for customers without DNS in their networks... | 19:25 |
| noonedeadpunk | So dnsmasq sends just empty reply to attaker, which is not really usefull for them... | 19:26 |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible stable/rocky: move ceph-ansible to stable-3.2 instead of stable-3.1 https://review.openstack.org/619796 | 19:36 |
| openstackgerrit | Marc Gariépy (mgariepy) proposed openstack/openstack-ansible stable/rocky: Add openstackclient bash completion https://review.openstack.org/619354 | 19:39 |
| *** dw has quit IRC | 19:48 | |
| *** emine has quit IRC | 19:58 | |
| *** cshen has quit IRC | 20:03 | |
| *** cshen has joined #openstack-ansible | 20:19 | |
| mgariepy | nice. now i have the same error on ceph-ubuntu-bionic, xenial and ceph opensuse ! :D | 20:34 |
| *** cshen has quit IRC | 20:34 | |
| *** cshen has joined #openstack-ansible | 20:36 | |
| *** noonedeadpunk has quit IRC | 20:37 | |
| mgariepy | anyone seens this ? http://logs.openstack.org/54/619354/2/check/openstack-ansible-deploy-ceph-ubuntu-xenial/f0f7eff/job-output.txt.gz#_2018-11-23_20_29_25_157979 | 20:42 |
| mgariepy | ERROR! The requested handler 'restart ceph mons' was not found in either the main handlers list nor in the listening handlers list | 20:42 |
| *** dave-mccowan has joined #openstack-ansible | 20:56 | |
| *** dave-mccowan has quit IRC | 21:11 | |
| guilhermesp | ppl, in order to test so_placement role, would I need to create the task like "test-install-placement.yml" in here https://github.com/openstack/openstack-ansible-tests ? | 21:16 |
| openstackgerrit | Merged openstack/openstack-ansible stable/pike: Fix ansible-sshd version https://review.openstack.org/618995 | 21:19 |
| *** fimaz has joined #openstack-ansible | 21:22 | |
| openstackgerrit | Logan V proposed openstack/openstack-ansible stable/queens: Update etcd role to v2.0 https://review.openstack.org/616408 | 21:24 |
| *** fimaz has quit IRC | 21:29 | |
| *** fimaz has joined #openstack-ansible | 21:30 | |
| *** jawad_axd has joined #openstack-ansible | 21:34 | |
| *** hamzaachi has quit IRC | 21:44 | |
| *** hamzaachi has joined #openstack-ansible | 21:45 | |
| *** pcaruana has quit IRC | 21:46 | |
| *** hamzaachi has quit IRC | 21:46 | |
| *** hamzaachi has joined #openstack-ansible | 21:47 | |
| openstackgerrit | git-harry proposed openstack/openstack-ansible stable/ocata: Fix ansible-sshd version https://review.openstack.org/619800 | 21:49 |
| *** fimaz has quit IRC | 21:49 | |
| *** fimaz has joined #openstack-ansible | 21:49 | |
| *** hamzaachi has quit IRC | 21:57 | |
| *** hamzaachi has joined #openstack-ansible | 21:58 | |
| *** hamzaachi has quit IRC | 21:59 | |
| *** hamzaachi has joined #openstack-ansible | 21:59 | |
| *** fimaz has quit IRC | 22:01 | |
| *** hamzaachi has quit IRC | 22:04 | |
| *** hamzaachi has joined #openstack-ansible | 22:05 | |
| *** hamzaachi has quit IRC | 22:06 | |
| *** hamzaachi has joined #openstack-ansible | 22:06 | |
| *** fimaz has joined #openstack-ansible | 22:11 | |
| *** fimaz has quit IRC | 22:12 | |
| *** lbragstad has joined #openstack-ansible | 22:53 | |
| *** dcdamien has quit IRC | 22:53 | |
| *** hamzaachi has quit IRC | 23:11 | |
| *** hamzaachi has joined #openstack-ansible | 23:12 | |
| *** hamzaachi has quit IRC | 23:13 | |
| *** hamzaachi has joined #openstack-ansible | 23:14 | |
| *** jawad_axd has quit IRC | 23:30 | |
| *** hamzaachi has quit IRC | 23:42 | |
| *** cshen has quit IRC | 23:43 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!