*** gouthamr has joined #openstack-ansible | 00:12 | |
*** yifei has joined #openstack-ansible | 00:14 | |
cloudnull | looks like we're seeing """2017-08-15 23:29:58.387193 | failed: [aio1_utility_container-33872c18] (item={'disk': 1, 'vcpus': 1, 'id': 201, 'ram': 256, 'name': 'tempest1'}) => {"failed": true, "item": {"disk": 1, "id": 201, "name": "tempest1", "ram": 256, "vcpus": 1}, "msg": "Error fetching flavor list. (503) Server Error for url: http://172.29.236.100:8774/v2.1/flavors/detail?is_public=None Service Unavailable (Inner | 00:20 |
---|---|---|
cloudnull | Exception: Expecting value: line 1 column 1 (char 0))"}""" in the gate atm ? | 00:20 |
cloudnull | I see that error in several jobs | 00:20 |
cloudnull | which means something is not working with nova or something is busted in tempest | 00:21 |
openstackgerrit | Merged openstack/openstack-ansible-os_glance master: Update paste, policy and rootwrap configurations 2017-08-15 https://review.openstack.org/493800 | 00:21 |
openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_swift master: Add basic functional testing for swift3 https://review.openstack.org/490248 | 00:24 |
*** lucasxu has joined #openstack-ansible | 00:24 | |
*** schwicht has quit IRC | 00:26 | |
-cloudnull- we're seeing gate failure for tempest atm, if folks have a moment to take a look at what's going on it'd be greatly appreciated -- http://logs.openstack.org/87/493087/2/check/gate-openstack-ansible-openstack-ansible-aio-ubuntu-xenial/bab66a5/console.html#_2017-08-15_20_56_31_389607 | 00:26 | |
cloudnull | andymccr_: ^ | 00:27 |
cloudnull | will bbl | 00:27 |
openstackgerrit | Merged openstack/openstack-ansible-ops master: Remove neutron agent task fails to loop through multiple agents. https://review.openstack.org/493981 | 00:27 |
*** markvoelker has quit IRC | 00:30 | |
*** thorst has joined #openstack-ansible | 00:42 | |
*** lucasxu has quit IRC | 00:54 | |
*** hw_wutianwei has joined #openstack-ansible | 00:58 | |
openstackgerrit | Merged openstack/openstack-ansible-os_nova stable/ocata: Reorder the placement service setup https://review.openstack.org/493942 | 01:06 |
HarryHuang | hi guys, can bridges like br-mgmt, br-vlan, br-vxlan and br-storage be openvswitch bridge instead of linuxbridge? i saw there is a doc https://docs.openstack.org/developer/openstack-ansible-os_neutron/app-openvswitch.html but lxc-veth-wiring.sh in openstack-ansible-lxc_container_create only has command like brctl and shows no cover for ovs. so when can ovs support be finished ? or is there a issue in finishing this part ? | 01:08 |
openstackgerrit | Merged openstack/openstack-ansible-tests master: test-ansible-functional.sh: Only run yum-config-manager on CentOS https://review.openstack.org/493637 | 01:09 |
*** lucasxu has joined #openstack-ansible | 01:11 | |
*** gouthamr has quit IRC | 01:13 | |
*** lucasxu has quit IRC | 01:13 | |
cloudnull | HarryHuang: there is ovs support and a couple folks are using it already | 01:14 |
cloudnull | we've doc'd quite a bit of it | 01:14 |
cloudnull | but the best quick guide that I'm aware of is http://trumant.github.io/openstack-ansible-openvswitch.html | 01:14 |
cloudnull | which was developed by one of our core reviewers | 01:15 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/openstack-ansible-os_magnum master: Updated from OpenStack Ansible Tests https://review.openstack.org/490540 | 01:23 |
HarryHuang | cloudnull: thanks for reply, the diagram in quick guide seems refer to a implement of ovs bridge for provider bridge inside lxc. what i wanna know is whether osa support ovs bridge outside LXC (br-mgmt in the diagram for example) | 01:24 |
*** gkadam has quit IRC | 01:28 | |
*** gouthamr has joined #openstack-ansible | 01:30 | |
cloudnull | so you want to deploy ovs (neutron agents) without containers? | 01:31 |
cloudnull | so it's all running on the hosts? | 01:31 |
cloudnull | or something else? | 01:31 |
*** esberglu has quit IRC | 01:34 | |
*** esberglu has joined #openstack-ansible | 01:35 | |
*** esberglu has quit IRC | 01:39 | |
*** woodard has joined #openstack-ansible | 01:41 | |
*** esberglu has joined #openstack-ansible | 01:43 | |
*** esberglu has quit IRC | 01:43 | |
*** lbragstad has quit IRC | 01:44 | |
*** woodard has quit IRC | 01:48 | |
*** thorst has quit IRC | 01:59 | |
*** hybridpollo has quit IRC | 02:14 | |
*** hybridpollo has joined #openstack-ansible | 02:19 | |
*** markvoelker has joined #openstack-ansible | 02:31 | |
*** gouthamr has quit IRC | 02:34 | |
*** markvoelker has quit IRC | 03:04 | |
*** vishwana_ has joined #openstack-ansible | 03:18 | |
*** vishwanathj has quit IRC | 03:21 | |
*** udesale has joined #openstack-ansible | 03:44 | |
*** poopcat has quit IRC | 03:55 | |
*** markvoelker has joined #openstack-ansible | 04:01 | |
jmccrory | cloudnull andymccr_ : here's why tempest is failing in integrated gate, from nova-api-os-compute's nova-api-wsgi.log: http://paste.openstack.org/show/618488/ | 04:14 |
cloudnull | logan-: are you seeing those dynamic groups show up in the inventory and getting saved in the backup tar? | 04:16 |
cloudnull | it should not be appending to that tarball? at least I'd think it shouldn't | 04:16 |
jmccrory | also, not seeing openstack service/host logs for integrated gate checks, does anyone know what happened there? http://logs.openstack.org/87/493087/2/check/gate-openstack-ansible-openstack-ansible-aio-ubuntu-xenial/bab66a5/ | 04:17 |
cloudnull | jmccrory: yea I'm seeing the same | 04:18 |
cloudnull | when it fails | 04:18 |
cloudnull | when success it's collecting correctly | 04:19 |
cloudnull | I've not looked into it | 04:19 |
jmccrory | hmm yeah successes are there..weird | 04:20 |
cloudnull | what's "requestlog" ? | 04:21 |
cloudnull | is that something we're setting somewhere? | 04:21 |
cloudnull | ah https://github.com/openstack/openstack-ansible-os_nova/blob/master/templates/api-paste.ini.j2#L43 | 04:22 |
cloudnull | its upstream too | 04:23 |
cloudnull | https://github.com/openstack/nova/blob/master/etc/nova/api-paste.ini#L43 | 04:23 |
cloudnull | but it looks like it's making nova generally angry | 04:23 |
jmccrory | https://github.com/openstack/nova/commit/aa45a6f3ab7ae870f99365b14b276cceff51ec8d | 04:24 |
jmccrory | well, it's in the deployed api-paste | 04:25 |
cloudnull | and its for sure there | 04:26 |
cloudnull | https://github.com/openstack/nova/blob/master/nova/api/openstack/requestlog.py | 04:26 |
jmccrory | that's interesting. i tried 'import nova.api.openstack.requestlog' from inside the venv and got the same import error | 04:31 |
cloudnull | is the file found ? | 04:32 |
cloudnull | maybe we're pulling in a bad version within the integrated gate? | 04:32 |
cloudnull | we did just recently bump all of the policy files etc | 04:32 |
jmccrory | ah, nope file's not there | 04:32 |
cloudnull | at least i saw some fly by the other day | 04:33 |
* cloudnull doesn't remember if it was for nova | 04:33 | |
jmccrory | ohh sha bump? | 04:33 |
cloudnull | maybe we need one for nova? | 04:33 |
*** markvoelker has quit IRC | 04:34 | |
jmccrory | wow looks like it. we're just one commit behind the change needed . https://github.com/openstack/nova/commits/master?after=fefff942ef5f8b3b8b209be1554daba28865d73a+319 https://github.com/openstack/openstack-ansible/blob/master/playbooks/defaults/repo_packages/openstack_services.yml#L150 | 04:37 |
*** hybridpollo has quit IRC | 04:38 | |
cloudnull | typical ... :D | 04:38 |
jmccrory | heh | 04:40 |
*** weezS has joined #openstack-ansible | 04:43 | |
cloudnull | jmccrory: you making that change in the integrated gate? | 04:44 |
cloudnull | I can smash it through if you like ? | 04:44 |
jmccrory | sure you can go for it, i'll +2 | 04:45 |
cloudnull | ++ | 04:45 |
openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible master: Move nova forward to unblock the gate https://review.openstack.org/494072 | 04:49 |
cloudnull | I pulled the head of stable pike | 04:49 |
cloudnull | which should have everything we need | 04:49 |
jmccrory | cool thanks | 04:52 |
*** chhavi has joined #openstack-ansible | 04:58 | |
cloudnull | thank you | 05:00 |
*** chhavi has quit IRC | 05:00 | |
*** drifterza has quit IRC | 05:00 | |
openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-os_sahara master: Add support for the openSUSE Leap distributions https://review.openstack.org/494076 | 05:15 |
*** tobberydberg has joined #openstack-ansible | 05:31 | |
*** markvoelker has joined #openstack-ansible | 05:32 | |
*** udesale__ has joined #openstack-ansible | 05:35 | |
*** udesale has quit IRC | 05:35 | |
*** pcaruana has joined #openstack-ansible | 05:45 | |
openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible master: Move nova forward to unblock the gate https://review.openstack.org/494072 | 05:55 |
*** weezS has quit IRC | 05:57 | |
*** xfei has joined #openstack-ansible | 06:01 | |
*** jbadiapa has joined #openstack-ansible | 06:04 | |
*** markvoelker has quit IRC | 06:05 | |
*** Oku_OS-away is now known as Oku_OS | 06:11 | |
*** xfei has quit IRC | 06:11 | |
*** jamesdenton has quit IRC | 06:16 | |
*** armaan has joined #openstack-ansible | 06:17 | |
*** jamesdenton has joined #openstack-ansible | 06:20 | |
*** udesale has joined #openstack-ansible | 06:23 | |
*** udesale__ has quit IRC | 06:23 | |
*** udesale has quit IRC | 06:35 | |
*** udesale has joined #openstack-ansible | 06:36 | |
openstackgerrit | Manuel Buil proposed openstack/openstack-ansible master: Add tacker installation https://review.openstack.org/485259 | 06:46 |
openstackgerrit | Merged openstack/openstack-ansible-os_keystone master: tasks: keystone_install: Fix virtualenv-tools issue on openSUSE https://review.openstack.org/493566 | 07:01 |
*** markvoelker has joined #openstack-ansible | 07:02 | |
*** markus_z has joined #openstack-ansible | 07:05 | |
*** markus_z has quit IRC | 07:09 | |
*** shardy_afk is now known as shardy | 07:10 | |
*** markus_z has joined #openstack-ansible | 07:11 | |
*** sxc731 has joined #openstack-ansible | 07:18 | |
odyssey4me | o/ | 07:18 |
*** zerda3 has joined #openstack-ansible | 07:23 | |
*** hw_wutianwei has quit IRC | 07:29 | |
*** hw_wutianwei has joined #openstack-ansible | 07:31 | |
*** shardy is now known as shardy_afk | 07:32 | |
*** markvoelker has quit IRC | 07:36 | |
*** mbuil has joined #openstack-ansible | 07:45 | |
openstackgerrit | Merged openstack/openstack-ansible-os_sahara master: Add support for the openSUSE Leap distributions https://review.openstack.org/494076 | 07:55 |
*** thorst has joined #openstack-ansible | 08:04 | |
hwoarang | mhayden: this debian failure is a mystery http://logs.openstack.org/89/488489/8/check/gate-ansible-hardening-ansible-func-debian-jessie/47fae70/console.html | 08:07 |
hwoarang | hello odyssey4me | 08:07 |
openstackgerrit | Andy McCrae proposed openstack/openstack-ansible master: Pin SHA's for stable/pike rc1 https://review.openstack.org/493816 | 08:07 |
andymccr_ | cloudnull: jmccrory will dig on that a bit more but we have a new sha pin patch for rc1 already ^ so that shouldve had an upto date enough nova? | 08:08 |
*** andymccr_ is now known as andymccr | 08:08 | |
*** thorst has quit IRC | 08:09 | |
*** jafeha has joined #openstack-ansible | 08:14 | |
odyssey4me | andymccr I think you might need https://review.openstack.org/493944 in for that pin patch to work | 08:15 |
odyssey4me | perhaps hwoarang or evrardjp could review? | 08:15 |
*** shardy_afk is now known as shardy | 08:17 | |
gokhan_ | hi folks, In step setup_infrastructure.yml in galera container we have failure on pip install . We have connection error | 08:18 |
gokhan_ | I tried python /opt/get-pip.py on galera container, it gives connection failure | 08:18 |
odyssey4me | gokhan_ so you have a connection issue ;) | 08:19 |
odyssey4me | something between your containers and the host is not right, you'll have to work through each connection to figure out what's going wrong - but it's likely a configuration error | 08:20 |
evrardjp | interesting | 08:22 |
gokhan_ | odyssey4me: this is failure log http://paste.openstack.org/show/618499/ but only it gives error on galera container | 08:22 |
*** drifterza has joined #openstack-ansible | 08:24 | |
gokhan_ | odyssey4me : this is my user config yaml file http://paste.openstack.org/show/618500/. I do not create haproxy . now I try to run osa for testing and then prapare for production | 08:26 |
evrardjp | so your LB is ready already? | 08:27 |
evrardjp | gokhan_: because no magic happens... | 08:27 |
gokhan_ | evrardjp: my LB is not ready . I want to go step to step. Firstly one cotroller and one compute so do not need LB I think. Do I have need LB? Sorry for asking simple questions. I am new to osa | 08:32 |
evrardjp | you need. | 08:32 |
evrardjp | Galera DB is load balaced, so are many other components, like your repo. | 08:33 |
*** markvoelker has joined #openstack-ansible | 08:33 | |
gokhan_ | evrardjp: ok got it. thanks for your help | 08:35 |
evrardjp | you could pre-configure your HW load balancer | 08:35 |
evrardjp | it doesn't hurt | 08:35 |
evrardjp | you have your inventory already | 08:35 |
evrardjp | if you think you won't change, that should be good enough | 08:35 |
evrardjp | at least not change the infra... | 08:36 |
gokhan_ | evrardjp: ok thanks but I do not have HW load balancer now. I will create haproxy | 08:39 |
evrardjp | yeah, sounds best. | 08:39 |
*** pbandark has joined #openstack-ansible | 08:42 | |
*** jbadiapa_ has joined #openstack-ansible | 08:50 | |
*** jbadiapa has quit IRC | 08:51 | |
mbuil | guys, are you aware of a gate problem when doing tempest? I see that it tries to fetch the flavor list but the service is unavailable | 09:01 |
*** thorst has joined #openstack-ansible | 09:05 | |
*** markvoelker has quit IRC | 09:06 | |
*** thorst has quit IRC | 09:09 | |
*** electrofelix has joined #openstack-ansible | 09:09 | |
admin0 | in one of my friends setup, i find kvm-compute_hosts and lxd-compute_hosts .. but doing a grep -ri on those returns no hits from playbooks or /etc/ansible/* .. those tags are also valid to specify hosts ? | 09:24 |
openstackgerrit | Merged openstack/openstack-ansible-os_cinder master: Fix reloads for uWSGI https://review.openstack.org/493944 | 09:25 |
openstackgerrit | Andy McCrae proposed openstack/openstack-ansible master: Pin SHA's for stable/pike rc1 https://review.openstack.org/493816 | 09:26 |
openstackgerrit | Merged openstack/openstack-ansible-os_swift master: Add dnf support https://review.openstack.org/490157 | 09:31 |
*** yifei has quit IRC | 09:34 | |
*** deadnull has joined #openstack-ansible | 09:36 | |
hwoarang | odyssey4me: i am trying a AIO build on opensuse and i get the segfaults in uwsgi again | 09:38 |
hwoarang | seems like the current fixes are not enough :/ | 09:38 |
*** deadnull has quit IRC | 09:41 | |
hwoarang | interestingly the AIO built cryptography 2.0 instead of 2.0.3 | 09:43 |
andymccr | hwoarang: do you have any logs for that? | 09:43 |
andymccr | hwoarang: and on which services? | 09:43 |
hwoarang | andymccr: it's is the aio1_keystone_container-7cfad04c container that failed | 09:44 |
hwoarang | log is identical to this https://bugs.launchpad.net/openstack-ansible/+bug/1705521 | 09:44 |
openstack | Launchpad bug 1705521 in OpenStack Global Requirements "[os_keystone] uwsgi process segfaults on openSUSE" [Undecided,New] | 09:44 |
andymccr | hmm ok | 09:44 |
andymccr | so is it definitely the crypto bits? | 09:45 |
*** chyka has joined #openstack-ansible | 09:46 | |
hwoarang | something is fishy because the repo container has cryptography-2.0-cp27-cp27mu-manylinux1_x86_64.whl built | 09:46 |
andymccr | hmm | 09:46 |
*** hw_wutianwei has quit IRC | 09:47 | |
hwoarang | hmm and it's not built from source | 09:47 |
hwoarang | the md5sum is identical to cryptography-2.0-cp27-cp27mu-manylinux1_x86_64.whl in the pypi repo | 09:48 |
hwoarang | wth... | 09:48 |
hwoarang | libvirt-python is also 3.5.0 instead of 3.6.0 | 09:49 |
*** chyka has quit IRC | 09:50 | |
hwoarang | maybe my local dir is broken | 09:53 |
hwoarang | lets clean up stuff | 09:53 |
*** markvoelker has joined #openstack-ansible | 10:03 | |
*** thorst has joined #openstack-ansible | 10:05 | |
*** thorst has quit IRC | 10:11 | |
openstackgerrit | Andy McCrae proposed openstack/openstack-ansible master: Pin SHA's for stable/pike rc1 https://review.openstack.org/493816 | 10:12 |
openstackgerrit | Juan Vidal Allende proposed openstack/openstack-ansible master: Add support for OpenDaylight deployment in neutron https://review.openstack.org/480131 | 10:15 |
jvidal | hello | 10:16 |
jvidal | I'm having some issues verifying this patch: https://review.openstack.org/#/c/480128/ | 10:16 |
jvidal | it still fails due to SSL package not being installed on CentOS | 10:16 |
jvidal | I suspect that jenkins might be caching the artifacts (e.g. ansible-openstack-tests), and this is the reason why the CentOS fix is not being used | 10:17 |
jvidal | but I have no clue if that is the real reason, does anybody have some knowledge about jenkins job? | 10:18 |
jvidal | recheck does not seem to help :( | 10:18 |
*** pbandark has quit IRC | 10:22 | |
*** pbandark has joined #openstack-ansible | 10:23 | |
*** udesale has quit IRC | 10:29 | |
*** jbadiapa_ has quit IRC | 10:30 | |
*** jbadiapa_ has joined #openstack-ansible | 10:31 | |
*** markvoelker has quit IRC | 10:38 | |
*** yifei has joined #openstack-ansible | 10:50 | |
*** yifei has quit IRC | 10:54 | |
eoseq | hi | 11:00 |
eoseq | how to override ini file for a group? | 11:00 |
*** zerda3 has quit IRC | 11:00 | |
*** zerda3 has joined #openstack-ansible | 11:01 | |
*** thorst has joined #openstack-ansible | 11:04 | |
*** szaher has joined #openstack-ansible | 11:15 | |
*** schwicht has joined #openstack-ansible | 11:21 | |
*** pbandark has quit IRC | 11:21 | |
*** sxc731 has quit IRC | 11:27 | |
*** jbadiapa_ has quit IRC | 11:27 | |
openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-os_glance master: tasks: glance_install: Fix virtualenv-tools issue on openSUSE https://review.openstack.org/494175 | 11:34 |
*** markvoelker has joined #openstack-ansible | 11:35 | |
*** jvidal has quit IRC | 11:41 | |
*** pbandark has joined #openstack-ansible | 11:41 | |
*** jbadiapa_ has joined #openstack-ansible | 11:42 | |
*** schwicht has quit IRC | 11:45 | |
openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-os_swift master: tasks: swift_install: Fix virtualenv-tools issue on openSUSE https://review.openstack.org/493747 | 11:54 |
*** jvidal has joined #openstack-ansible | 11:58 | |
*** sxc731 has joined #openstack-ansible | 12:04 | |
*** MasterofJOKers has joined #openstack-ansible | 12:04 | |
*** jamesdenton has quit IRC | 12:07 | |
*** markvoelker has quit IRC | 12:08 | |
odyssey4me | eoseq you can use group_vars for that | 12:19 |
odyssey4me | just be aware that if you're using user_*.yml extra-vars you'll override any/all group_vars | 12:19 |
odyssey4me | evrardjp did a nice post about all that actually: https://evrard.me/group-and-host-variables-overriding-in-openstack-ansible.html | 12:20 |
*** woodard has joined #openstack-ansible | 12:27 | |
*** woodard has quit IRC | 12:27 | |
*** woodard has joined #openstack-ansible | 12:28 | |
eoseq | odyssey4me: thx | 12:30 |
openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-galera_server master: DNM - Testing experimental openSUSE jobs https://review.openstack.org/487498 | 12:40 |
andymccr | hwoarang: did you figure that crypto failure out? | 12:43 |
hwoarang | andymccr: probably something was stale in my workdir. I am running AIO on a clean env right now so we will see | 12:44 |
andymccr | ok cool - the master aio might be not in a working state, trying to figure out if my patches last night to reload services resolved everything or not | 12:45 |
*** hw_wutianwei has joined #openstack-ansible | 12:46 | |
hwoarang | ok | 12:46 |
andymccr | i think nova-api-os-compute was the only one causing issues though so if its that and tempest fails it may not be specific to suse | 12:47 |
odyssey4me | andymccr it may be a good idea to remove all the reload things and just do restarts instead | 12:51 |
odyssey4me | it's not perfect, and bad for rolling intra-series upgrade, but at least it's functional and buys us more time | 12:51 |
*** pwiltsey has joined #openstack-ansible | 12:52 | |
andymccr | yeah thats true, although reloads should work well now - the other services were all good for reloads after i tested them | 12:52 |
evrardjp | wow that's a good thing to hear. | 12:53 |
andymccr | plus with uwsgi we can reload everything if we do it right | 12:53 |
andymccr | so far only cinder/nova do reloads with uwsgi though | 12:53 |
odyssey4me | yeah, using reloads instead of restarts for config changes would be ideal | 12:53 |
odyssey4me | for software changes it obviously requires a restart though | 12:54 |
odyssey4me | although with uwsgi maybe not | 12:54 |
mhayden | morning folks | 12:59 |
*** markvoelker has joined #openstack-ansible | 12:59 | |
prometheanfire | moin | 12:59 |
sauloaislan | Morning! | 13:00 |
prometheanfire | uwsgi in emporer mode requires a full restart (reload does a restart too iirc) for adding / changing a service | 13:00 |
andymccr | not running in emperor mode :) | 13:03 |
prometheanfire | k | 13:04 |
andymccr | but yeah i think there is a lot of space to do cool things with uwsgi | 13:06 |
*** schwicht has joined #openstack-ansible | 13:08 | |
*** esberglu has joined #openstack-ansible | 13:08 | |
odyssey4me | mhayden hwoarang cloudnull we're doing different things on different platforms here, so I'm curious about why: https://github.com/openstack/openstack-ansible-lxc_hosts/blob/master/tasks/lxc_net.yml#L50-L83 | 13:09 |
odyssey4me | on centos/ubuntu we disable the lxc-net service, but on suse we let it run | 13:09 |
hwoarang | hmm | 13:10 |
hwoarang | centos actually has it enabled | 13:11 |
hwoarang | it's only debian that disables it right? | 13:11 |
odyssey4me | hwoarang nope, we mask it which stops it from starting | 13:11 |
hwoarang | where do you mask it on centos? | 13:11 |
hwoarang | because i only see that on debian | 13:12 |
odyssey4me | hwoarang heh, I see what you mean | 13:12 |
odyssey4me | I misread | 13:12 |
odyssey4me | I'm specifically curious about this comment which seems to have been there right from the get-go, so perhaps cloudnull is going to be the only one answering. https://github.com/openstack/openstack-ansible-lxc_hosts/blob/master/tasks/lxc_net.yml#L50-L51 | 13:13 |
odyssey4me | I've just hit an issue for Trusty where the LXC packages won't install unless I prevent lxc-net from starting, so I'm curious regarding why. | 13:14 |
odyssey4me | I see dnsmasq is installed and enabled, so it's likely started and conflicting on the port. | 13:14 |
odyssey4me | perhaps I should mask *it* from starting. | 13:14 |
*** gouthamr has joined #openstack-ansible | 13:15 | |
hwoarang | no clue but does 'trusty' matter? | 13:17 |
odyssey4me | hwoarang it matters to RPC :) | 13:17 |
odyssey4me | and yes, this actually appears like it's doing the right sort of things versus what was running before | 13:18 |
odyssey4me | I wonder if this is potentially the source of the issues where networking doesn't come up right after a reboot | 13:18 |
hwoarang | hmm | 13:19 |
odyssey4me | heh, ok I think this actually is the source of all the trouble | 13:22 |
*** jwitko_ has joined #openstack-ansible | 13:23 | |
odyssey4me | lemme push a patch up as this pertains to xenial/master too | 13:23 |
*** kysse has quit IRC | 13:26 | |
cloudnull | mornings | 13:27 |
cloudnull | odyssey4me: did you get the lxc-net things situated? | 13:28 |
MasterofJOKers | hi. I'm having a hard time trying figuring out if openstack-ansible can be used on Debian or just on Ubuntu. Given that debian packages for newer openstack releases seem very unlikely, we might want to switch to directly deploying via ansible. Could anyone clarify this, please? | 13:29 |
odyssey4me | cloudnull ok, so I hit an issue on trusty with the latest lxc packages where they won't install in our standard deployment | 13:29 |
cloudnull | oh? | 13:29 |
cloudnull | whats the hang up ? | 13:29 |
odyssey4me | cloudnull so I took a peek and discovered that lxc-net wouldn't start because dnsmasq was already started, and lxc-net tries to start when installing the package | 13:29 |
odyssey4me | so I figured ok, I'll just mask it so it won't try - and that worked | 13:29 |
cloudnull | cool! | 13:30 |
odyssey4me | then I took a closer look and wondered why dnsmasq was starting | 13:30 |
cloudnull | we used to have that marked as disabled in trusty ? | 13:30 |
cloudnull | is it not anymore? | 13:30 |
odyssey4me | because the only things that use it are LXC and neutron things | 13:30 |
andymccr | MasterofJOKers: we haven't tested at all, and you would run into some issues around var names ( we specify vars for ubuntu-16.04) but assuming they operate pretty similarly, which they should - it might not be too much of a stretch to get them working (but it would require some work) | 13:30 |
odyssey4me | and none of those require dnsmasq to start on boot because they do the starting thing themselves | 13:30 |
cloudnull | odyssey4me: yes, dnsmasq is stared along with the lxcbr0 interface | 13:30 |
odyssey4me | exactly | 13:30 |
cloudnull | same with the iptable rules and such | 13:31 |
MasterofJOKers | andymccr, that sounds like what I was thinking. we might give it a try, thanks. | 13:31 |
odyssey4me | so I instead masked dnsmasq and retried installing, and that worked too - and lxc-net started with all the right configs too | 13:31 |
odyssey4me | as it turns out, there is the 'dnsmasq' package, and the 'dnsmasq-base' package | 13:31 |
andymccr | MasterofJOKers: sure no problem, i think the biggest trick is going to be that in places we have hardcoded options based on xenial (we used to support both xenial and trusty during the newton cycle so we needed to make a distinction) | 13:31 |
andymccr | so its not straight "debian" | 13:31 |
cloudnull | does one drop an init and other no? | 13:32 |
odyssey4me | the first sets up dnsmasq as a local forwarder and all that, whereas the second just has the binaries - and lxc actually has a dep on the second | 13:32 |
odyssey4me | cloudnull yep | 13:32 |
cloudnull | and we install the former im guessing | 13:32 |
odyssey4me | we only need the second, not the first - so our package list has been wrong all along and we've been working around the stuff that didn't have to be there | 13:32 |
cloudnull | nice! | 13:32 |
cloudnull | smashing bugs | 13:32 |
cloudnull | i had no idea about that package differenc e | 13:33 |
odyssey4me | I'm going to push a patch just to switch the package in the list, but I wonder if you couldn't revisit the lxc-net stuff and see if there's not more things we could rip out there now because of this finding | 13:33 |
hwoarang | odyssey4me: cloudnull interestingly enough lxc-net is disabled on suse and not running at all | 13:33 |
hwoarang | and things still work... | 13:33 |
cloudnull | it should be disabled. | 13:33 |
odyssey4me | hwoarang yes, but that's because we do all sorts of other stuff to make it the way we want to | 13:33 |
hwoarang | ahh o kthen | 13:33 |
hwoarang | *ok then | 13:33 |
odyssey4me | I'm not so sure that it *needs* to be disabled any more - we might be able to remove a bunch of workaround things | 13:34 |
odyssey4me | ie we could perhaps actually use it | 13:34 |
cloudnull | if we can set the ip range and subnet mask for the service then it could be used | 13:34 |
odyssey4me | is that our only requirement? | 13:34 |
cloudnull | last i remember it was hard-coded at 10.0.3.0/24 | 13:34 |
cloudnull | also we needed the ability to restart the lxcbr0 interface | 13:35 |
cloudnull | well ifup/down | 13:35 |
cloudnull | which that service did not provide | 13:35 |
*** seba has joined #openstack-ansible | 13:36 | |
cloudnull | i remember it not restarting correctly and leaving lxc networking in a funky state should the inteface need to be ifdown/up for some reason | 13:36 |
odyssey4me | ok, let me get part one done which is to change the packahe list | 13:37 |
odyssey4me | that at least saves some trouble | 13:37 |
cloudnull | and, idk if anyone is actually using this functionality, but there was a request long ago to get the lxcbr0 interface plugged into a physical network instead of nat'ing | 13:37 |
odyssey4me | wow, that sounds like a bad idea | 13:38 |
cloudnull | it makes total sense. | 13:38 |
*** pbandark1 has joined #openstack-ansible | 13:38 | |
odyssey4me | unless we want to transition that bridge to the management network | 13:38 |
cloudnull | lxcbr0 is replaced with br-mgmt | 13:38 |
cloudnull | or something similar | 13:38 |
cloudnull | maybe corp.* had their own dhcp servers and didn't want dnsmasq running on all of the hosts for some reason | 13:39 |
*** sxc731 has quit IRC | 13:39 | |
cloudnull | I can't totally remember tbh | 13:39 |
cloudnull | I just remember it being an ask we did long ago | 13:39 |
*** pbandark has quit IRC | 13:39 | |
*** pbandark1 is now known as pbandark | 13:39 | |
*** schwicht has quit IRC | 13:40 | |
cloudnull | did we get the tempest gate blocker squared away ? | 13:42 |
*** BjoernT has joined #openstack-ansible | 13:45 | |
* mhayden hopes the ansible-hardening gate will unblock today! | 13:45 | |
mhayden | project-config changes merged | 13:45 |
cloudnull | we had the issues with tempest | 13:47 |
cloudnull | well issues with nova and tempest was failing ? | 13:47 |
*** vakuznet has joined #openstack-ansible | 13:48 | |
cloudnull | andymccr: form the latest sha bump http://logs.openstack.org/16/493816/5/check/gate-openstack-ansible-openstack-ansible-ceph-ubuntu-xenial/9899474/console.html | 13:49 |
*** BjoernT has quit IRC | 13:49 | |
cloudnull | specifically http://logs.openstack.org/16/493816/5/check/gate-openstack-ansible-openstack-ansible-ceph-ubuntu-xenial/9899474/console.html#_2017-08-16_13_09_33_769973 | 13:49 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-lxc_hosts master: Remove dnsmasq from package list https://review.openstack.org/494212 | 13:50 |
odyssey4me | cloudnull hwoarang ^ your thoughts on that? | 13:51 |
cloudnull | http://logs.openstack.org/87/493087/2/check/gate-openstack-ansible-openstack-ansible-ceph-ubuntu-xenial/4305a3b/console.html#_2017-08-16_02_04_54_826945 | 13:51 |
hwoarang | mhayden: for ansible-hardening to unblock we need to get https://review.openstack.org/#/c/493908/ asap | 13:51 |
cloudnull | that is the current gate blocker | 13:51 |
*** schwicht has joined #openstack-ansible | 13:57 | |
*** hw_wutianwei has quit IRC | 13:59 | |
mhayden | oh yeah | 14:00 |
mhayden | i wonder if we are going to need to merge the bindep stuff into it | 14:00 |
mhayden | probably will | 14:00 |
*** udesale has joined #openstack-ansible | 14:01 | |
*** kysse has joined #openstack-ansible | 14:01 | |
openstackgerrit | Major Hayden proposed openstack/ansible-hardening master: Manually check apparmor_status https://review.openstack.org/493908 | 14:03 |
*** woodard has quit IRC | 14:03 | |
mhayden | hwoarang / cloudnull: could y'all look at https://review.openstack.org/#/c/493908/ ? i merged in the updates to bindep/run_tests to unblock the gate | 14:03 |
hwoarang | mhayden: patchset #4 was also good | 14:04 |
cloudnull | done. | 14:04 |
openstackgerrit | Major Hayden proposed openstack/ansible-hardening master: Updated from global requirements https://review.openstack.org/487982 | 14:04 |
hwoarang | all gates were green | 14:05 |
openstackgerrit | Major Hayden proposed openstack/ansible-hardening master: Skip sysctl configs when enabled: no https://review.openstack.org/493580 | 14:05 |
openstackgerrit | Major Hayden proposed openstack/ansible-hardening master: Change default prohibit root sshd password auth https://review.openstack.org/493657 | 14:05 |
mhayden | hwoarang: but it depended on that change to bindep | 14:05 |
mhayden | but bindep change couldn't go in without apparmor fix | 14:05 |
hwoarang | ah i see | 14:05 |
hwoarang | ok | 14:05 |
mhayden | so i had to put them together :/ | 14:05 |
*** arbrandes has joined #openstack-ansible | 14:06 | |
*** schwicht has quit IRC | 14:08 | |
*** arbrandes1 has quit IRC | 14:09 | |
openstackgerrit | Markos Chandras (hwoarang) proposed openstack/openstack-ansible-os_cinder master: tasks: cinder_install: Fix virtualenv-tools issue on openSUSE https://review.openstack.org/493578 | 14:09 |
errr_ | is there a link one can use in a blog or something that would point to https://docs.openstack.org/openstack-ansible/latest/contributor/quickstart-aio.html and it not change very often? Back in July it was https://docs.openstack.org/developer/openstack-ansible/developer-docs/quickstart-aio.html but now that link doesnt work because of the /latest/ addition | 14:09 |
sauloaislan | I have new/old erro http://paste.openstack.org/show/618430/ | 14:10 |
*** schwicht has joined #openstack-ansible | 14:11 | |
*** lbragstad has joined #openstack-ansible | 14:13 | |
odyssey4me | logan- here you go, good suggestion there: https://bugs.launchpad.net/openstack-ansible/+bug/1711140 | 14:14 |
openstack | Launchpad bug 1711140 in openstack-ansible "newton/trusty lxc package install failure" [Undecided,New] | 14:14 |
hwoarang | odyssey4me: did i misunderstood your comment on https://review.openstack.org/#/c/494116/ ? | 14:15 |
hwoarang | afaics the job is properly filtered | 14:15 |
hwoarang | *misunderstand | 14:17 |
odyssey4me | hwoarang I didn't see that one. I'm blind today it seems. | 14:19 |
hwoarang | heh | 14:19 |
*** weezS has joined #openstack-ansible | 14:19 | |
mattt | cloudnull: nice1 on https://review.openstack.org/#/c/449372/, always wished this was possible | 14:24 |
cloudnull | ++ its close | 14:25 |
*** lucasxu has joined #openstack-ansible | 14:25 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-lxc_hosts master: Ubuntu: Use dnsmasq-base, not dnsmasq https://review.openstack.org/494212 | 14:27 |
odyssey4me | hwoarang ^ removed the changes to anything except ubuntu, as that's where the issue I discovered lies | 14:29 |
odyssey4me | logan- also added a bug reference and upgrade path | 14:29 |
*** schwicht has quit IRC | 14:30 | |
*** jvidal has quit IRC | 14:34 | |
cloudnull | errr_: https://docs.openstack.org/openstack-ansible/latest/contributor/quickstart-aio.html | 14:34 |
cloudnull | that should be fairly static | 14:35 |
*** schwicht has joined #openstack-ansible | 14:35 | |
*** BjoernT has joined #openstack-ansible | 14:36 | |
*** pwiltsey has quit IRC | 14:36 | |
odyssey4me | errr_ that's the first time it's changes in many years :) | 14:36 |
errr_ | oh ok | 14:36 |
odyssey4me | well, I guess it's the first time since liberty I think | 14:37 |
errr_ | thanks | 14:37 |
odyssey4me | maybe kilo | 14:37 |
odyssey4me | I can't remember when we first created it | 14:37 |
odyssey4me | and I think if there's another docs shuffle there will end up being a revolt | 14:37 |
errr_ | I was getting ready to turn in my mandtory optional blog and was checkign to make sure it all still worked and found that link had changed | 14:37 |
cloudnull | yea the docs got shifted around in the last couple of cycles w/ the deploy guide | 14:39 |
cloudnull | but i suspect/hope that the pathing is stable now | 14:41 |
mgariepy | http 301 ftw | 14:42 |
mgariepy | (next time maybe) | 14:42 |
*** gkadam has joined #openstack-ansible | 14:45 | |
sauloaislan | Someone with problems in os_swift: Get swift rings, problem with authorization? | 14:45 |
cloudnull | andymccr: ^ | 14:46 |
*** jvidal has joined #openstack-ansible | 14:48 | |
andymccr | sauloaislan: whats the issue? can you provide a bit of logs/detail? | 14:48 |
*** pwiltsey has joined #openstack-ansible | 14:49 | |
andymccr | btw i figured out the gate issues now so master should be up just working on a patch | 14:53 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-lxc_hosts master: Ubuntu: Use dnsmasq-base, not dnsmasq https://review.openstack.org/494212 | 14:53 |
*** pwiltsey has quit IRC | 14:53 | |
armaan | folks, I am going to upgrade Mitaka to Newton. I was wondering whether you recommend using the upgrade script or doing manual upgrade as described here https://docs.openstack.org/openstack-ansible/newton/upgrade-guide/index.html | 14:55 |
sauloaislan | andymccr: http://paste.openstack.org/show/618534/ | 14:55 |
andymccr | sauloaislan: guessing it cut off there, it looks like nothing failed in that specific bit :) | 14:56 |
*** pwiltsey has joined #openstack-ansible | 14:58 | |
*** Oku_OS is now known as Oku_OS-away | 15:07 | |
*** woodard has joined #openstack-ansible | 15:10 | |
sauloaislan | andymccr: http://paste.openstack.org/show/618538/ | 15:11 |
sauloaislan | sorry | 15:11 |
*** tobberyd_ has joined #openstack-ansible | 15:14 | |
andymccr | sauloaislan: what are you running? e.g. which version and what commands | 15:14 |
mgariepy | armaan, i personally use the manual steps | 15:16 |
*** tobberydberg has quit IRC | 15:17 | |
sliver | Anyone around that is familiar with the repo-build play? Not able to get multiple major release versions of ubuntu to actually build the wheels. It only builds the xenial 16.04 files and skips trusty... | 15:17 |
armaan | mgariepy: cool ... I was also thinking of using manual steps. :) | 15:18 |
*** schwicht has quit IRC | 15:18 | |
*** tobberyd_ has quit IRC | 15:19 | |
sauloaislan | andymccr: master "openstack-ansible setup-openstack.yml" | 15:19 |
*** marst has joined #openstack-ansible | 15:19 | |
mgariepy | sliver, which tag are you using ? | 15:19 |
*** markus_z has quit IRC | 15:20 | |
mgariepy | sliver, you might need those : https://review.openstack.org/#/c/485266/ https://review.openstack.org/#/c/485632/ | 15:20 |
openstackgerrit | Andy McCrae proposed openstack/openstack-ansible master: Pin SHA's for stable/pike rc1 https://review.openstack.org/493816 | 15:20 |
andymccr | ^ should fix the master gate and pin sha's for rc1 | 15:21 |
andymccr | sauloaislan: can you try just openstack-ansible os-swift-install.yml | 15:23 |
*** mgariepy has quit IRC | 15:23 | |
sliver | 14.2.6 | 15:23 |
sliver | tried 14.2.7, same result | 15:23 |
*** craigs__ has joined #openstack-ansible | 15:23 | |
sliver | have 3 14.04 repos(old), added one new xenial repo for getting the bits in place to upgrade | 15:24 |
sliver | it selects the xenial repo as the master in the repo-build play | 15:24 |
sliver | my other region is only 16.04 so didn't face this there :/ (also 14.2.6) | 15:24 |
*** gkadam is now known as gkadam-afk | 15:24 | |
sliver | the next best thing would be a decent procedure to prune the 16.04 repo box now since the bits are there and let 14.04 do its thing | 15:25 |
*** mgariepy has joined #openstack-ansible | 15:25 | |
sauloaislan | andymccr: Ok I'll try | 15:25 |
sliver | maybe inventory_manage script would suffice in cleaning up inventory enough to attempt | 15:25 |
*** udesale has quit IRC | 15:25 | |
sliver | armaan, manual :), control the downtime of the component groups | 15:26 |
sliver | armaan, and make sure you get the latest rabbit bits that pin erlang 19.x !! | 15:26 |
sliver | you may have to manually do it as I don't think it's back ported still | 15:26 |
armaan | sliver: Are you also upgrading your environment? | 15:27 |
armaan | sliver: from mitaka to newton? | 15:27 |
sliver | did 4 weeks back | 15:27 |
sliver | yessir | 15:27 |
*** shardy is now known as shardy_afk | 15:27 | |
sliver | 13.1.x to 14.2.6 | 15:27 |
armaan | sliver: cool, any show stopper with trusty to xenial upgrade? | 15:27 |
sliver | Ignoring hardware compat, only this repo thing. | 15:28 |
sliver | and the rabbit/erlang mix but that was resolved with patches | 15:28 |
sliver | like I said, just manual still in newton branch I believe so be very aware of that | 15:29 |
armaan | sliver: I am not sure i understand how you fixed the rabbit issue? You applied patches to the infra playbooks? | 15:29 |
sliver | Followed the pin guidance/patch on the rabbit nodes | 15:29 |
armaan | sliver: Is it possible that you could share the link for that? | 15:30 |
sliver | https://bugs.launchpad.net/openstack-ansible/+bug/1706122 | 15:30 |
openstack | Launchpad bug 1706122 in openstack-ansible "openstack-ansible rabbitmq installs latest release of erlang which is not supported" [Undecided,Incomplete] | 15:30 |
armaan | sliver: thanks a ton:) | 15:30 |
*** gkadam-afk has quit IRC | 15:31 | |
*** woodard has quit IRC | 15:31 | |
sliver | np | 15:31 |
*** mgariepy has quit IRC | 15:31 | |
sliver | I am that submitter fwiw | 15:31 |
*** chyka has joined #openstack-ansible | 15:31 | |
*** woodard has joined #openstack-ansible | 15:32 | |
*** gkadam has joined #openstack-ansible | 15:32 | |
*** craigs__ has quit IRC | 15:32 | |
sliver | mgariepy, ahh just now saw your second reply, checking those patches | 15:32 |
*** schwicht has joined #openstack-ansible | 15:33 | |
sliver | I think I need to apply 485266 as it hasn't been released to a tag yet. did 485632 which did fix the rsync.. thanks | 15:33 |
*** mgariepy has joined #openstack-ansible | 15:34 | |
*** yolanda has quit IRC | 15:44 | |
*** yolanda has joined #openstack-ansible | 15:44 | |
*** tobberydberg has joined #openstack-ansible | 15:49 | |
sauloaislan | andymccr: same erro | 15:51 |
andymccr | sauloaislan: can you paste the logs, it's weird because it isn't running the key sync job on the swift storage host which it should be. | 15:51 |
*** tobberydberg has quit IRC | 15:56 | |
*** pcaruana has quit IRC | 15:57 | |
sauloaislan | andymccr: http://paste.openstack.org/show/618550/ | 15:57 |
sauloaislan | andymccr: http://paste.openstack.org/show/618551/ | 16:00 |
openstackgerrit | Merged openstack/ansible-hardening master: Manually check apparmor_status https://review.openstack.org/493908 | 16:01 |
*** tobberydberg has joined #openstack-ansible | 16:02 | |
armaan | sliver: quick question, since Newton is supported for both trusty and xenial. Did you upgrade the openstack first and then upgraded the host OS? | 16:03 |
mgariepy | armaan, you will need to upgrade to newton first. then the OS> | 16:04 |
armaan | mgariepy: gotcha! thanks | 16:04 |
mgariepy | armaan, sliver :: https://etherpad.openstack.org/p/osa-newton-xenial-upgrade | 16:05 |
mgariepy | some notes taken by logan- on upgrading to xenial. | 16:06 |
armaan | cool, thanks for sharing this :) | 16:08 |
mhayden | nice! | 16:09 |
*** tobberydberg has quit IRC | 16:10 | |
odyssey4me | looking for another core review for https://review.openstack.org/494212 | 16:11 |
*** germs has joined #openstack-ansible | 16:14 | |
*** germs1 has joined #openstack-ansible | 16:19 | |
mgariepy | odyssey4me, done. | 16:19 |
odyssey4me | thanks mgariepy | 16:19 |
*** germs has quit IRC | 16:19 | |
*** germs1 has quit IRC | 16:21 | |
*** tobberydberg has joined #openstack-ansible | 16:23 | |
*** germs has joined #openstack-ansible | 16:23 | |
mhayden | yay ansible hardening gate is unblocked | 16:27 |
openstackgerrit | Merged openstack/openstack-ansible-os_glance master: tasks: glance_install: Fix virtualenv-tools issue on openSUSE https://review.openstack.org/494175 | 16:27 |
*** germs1 has joined #openstack-ansible | 16:27 | |
*** germs1 has left #openstack-ansible | 16:27 | |
*** tobberydberg has quit IRC | 16:27 | |
openstackgerrit | Merged openstack/openstack-ansible-lxc_hosts master: Ubuntu: Use dnsmasq-base, not dnsmasq https://review.openstack.org/494212 | 16:31 |
*** germs2 has joined #openstack-ansible | 16:31 | |
*** mbuil has quit IRC | 16:32 | |
*** TxGirlGeek has joined #openstack-ansible | 16:32 | |
*** germs has quit IRC | 16:33 | |
armaan | mgariepy sliver: I am getting this error http://paste.openstack.org/show/618559/ on running the deploy-config-changes.yml. have you seen this issue before? | 16:34 |
*** os_user has joined #openstack-ansible | 16:34 | |
armaan | there is a "openstack_deploy.MITAKA" instead of "openstack_deploy.NEWTON" dir. | 16:36 |
os_user | Has anyone run into this error trying to deploy and AIO from master? “TASK [resolvconf : Gather variables for each operating system] **** fatal: [aio1_repo_container-ac82321a]: FAILED! => {"failed": true, "msg": "No file was found when using with_first_found. Use the 'skip: true' option to allow this task to be skipped if no files are found”}” | 16:40 |
*** drifterza has quit IRC | 16:42 | |
sauloaislan | os_user: https://review.openstack.org/#/c/493739/ | 16:43 |
os_user | thanks | 16:43 |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-lxc_hosts stable/ocata: Ubuntu: Use dnsmasq-base, not dnsmasq https://review.openstack.org/494266 | 16:44 |
*** pwiltsey has quit IRC | 16:46 | |
openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-lxc_hosts stable/newton: Ubuntu: Use dnsmasq-base, not dnsmasq https://review.openstack.org/494267 | 16:47 |
*** tobberydberg has joined #openstack-ansible | 16:47 | |
armaan | This task https://github.com/openstack/openstack-ansible/blob/master/scripts/upgrade-utilities/playbooks/deploy-config-changes.yml#L60 is failing with this error http://paste.openstack.org/show/618559/ | 16:47 |
armaan | I suppose this is not a bug? | 16:47 |
*** jvidal has quit IRC | 16:47 | |
armaan | odyssey4me andymccr ^^ | 16:47 |
*** germs has joined #openstack-ansible | 16:48 | |
*** weezS has quit IRC | 16:48 | |
*** tobberydberg has quit IRC | 16:52 | |
*** tobberydberg has joined #openstack-ansible | 16:58 | |
*** pwiltsey has joined #openstack-ansible | 16:59 | |
*** strigazi_OFF is now known as strigazi | 17:00 | |
*** larry1085 has joined #openstack-ansible | 17:00 | |
*** pwiltsey has quit IRC | 17:01 | |
*** tobberydberg has quit IRC | 17:02 | |
*** shardy_afk is now known as shardy | 17:03 | |
*** pwiltsey has joined #openstack-ansible | 17:04 | |
odyssey4me | armaan could be - the mitaka->newton upgrades have no tests, so humans have to test | 17:04 |
odyssey4me | we have tests now for newton->ocata->pike | 17:04 |
*** germs has left #openstack-ansible | 17:04 | |
armaan | odyssey4me: ohhh, any tip how can i fix it. | 17:05 |
odyssey4me | armaan not off-hand, I'm not in a position to dig into the cause and therefore figure out a solution | 17:05 |
odyssey4me | maybe the solution is in a later branch? | 17:05 |
armaan | odyssey4me: I can understand, you have more pressing matters! | 17:05 |
odyssey4me | the upgrade mechanism hasn't changed all that much in later versions, so perhaps there's a fix already in a later branch | 17:06 |
odyssey4me | but I'd suggest registering a bug regardless | 17:06 |
odyssey4me | we have a bug smash scheduled tomorrow, so get your bug in now ;) | 17:06 |
armaan | odyssey4me: thanks, I will file the bug then :) | 17:07 |
*** weezS has joined #openstack-ansible | 17:07 | |
mgariepy | need one more review please :D https://review.openstack.org/#/c/493657/ | 17:13 |
*** shardy has quit IRC | 17:23 | |
*** drifterza has joined #openstack-ansible | 17:25 | |
*** gkadam has quit IRC | 17:25 | |
*** weezS has quit IRC | 17:29 | |
*** tobberydberg has joined #openstack-ansible | 17:29 | |
sauloaislan | odyssey4me: http://paste.openstack.org/show/618565/ repo_build failed in ocata, can you help me? | 17:32 |
*** tobberydberg has quit IRC | 17:33 | |
openstackgerrit | Merged openstack/ansible-hardening master: Change default prohibit root sshd password auth https://review.openstack.org/493657 | 17:38 |
*** armaan_ has joined #openstack-ansible | 17:42 | |
*** armaan has quit IRC | 17:45 | |
*** armaan has joined #openstack-ansible | 17:46 | |
odyssey4me | sauloaislan can you re-run the repo-build playbook with -vvv and pastebin the entire result please? | 17:47 |
*** armaan_ has quit IRC | 17:49 | |
sauloaislan | odyssey4me: Apparently is a proxy problem I will try again! | 17:50 |
sauloaislan | odyssey4me: thanks! | 17:51 |
odyssey4me | sauloaislan ah yes, proxies are a bit of a pain | 17:51 |
*** admin0 has left #openstack-ansible | 17:51 | |
*** admin0 has joined #openstack-ansible | 17:51 | |
sauloaislan | odyssey4me: Hey I have a old problem with swift, can you help me ? | 17:54 |
odyssey4me | sauloaislan sorry, andymccr is the swift guy - I'm not all that familiar with it | 17:54 |
*** strigazi is now known as strigazi_OFF | 17:55 | |
sauloaislan | odyssey4me: ah ok! thanks! | 17:55 |
*** poopcat has joined #openstack-ansible | 17:56 | |
*** electrofelix has quit IRC | 18:00 | |
sauloaislan | os_user: Your OSA it worked ? | 18:01 |
*** pwiltsey has quit IRC | 18:01 | |
*** albertcard1 has joined #openstack-ansible | 18:01 | |
*** poopcat has quit IRC | 18:02 | |
*** h1nch has joined #openstack-ansible | 18:08 | |
os_user | sauloaisian: it got farther, failed on “Could not get lock /var/lib/apt/lists/lock” in the aio1_memcached_contained container. Running it again. | 18:08 |
*** sxc731 has joined #openstack-ansible | 18:08 | |
*** armaan has quit IRC | 18:09 | |
*** pcaruana has joined #openstack-ansible | 18:09 | |
*** pwiltsey has joined #openstack-ansible | 18:11 | |
*** homeski has joined #openstack-ansible | 18:12 | |
*** albertcard1 is now known as albertcard | 18:14 | |
*** albertcard is now known as poopcat | 18:14 | |
*** tobberydberg has joined #openstack-ansible | 18:18 | |
*** tobberydberg has quit IRC | 18:19 | |
*** tobberydberg has joined #openstack-ansible | 18:20 | |
*** tobberydberg has quit IRC | 18:22 | |
*** tobberydberg has joined #openstack-ansible | 18:22 | |
odyssey4me | if any cores are around, I have two backports looking for attention which I'd like to get merged ASAP: https://review.openstack.org/#/q/I0b0a8dec3943651f379479ea1a3b8038736b9219,n,z | 18:23 |
odyssey4me | thanks jmccrory :) | 18:25 |
odyssey4me | how're you doing by the way? what's going on over there? | 18:25 |
jmccrory | np, good catch on that | 18:25 |
sauloaislan | os_user: ok! I have an problem with swift | 18:25 |
jmccrory | not bad, settling into another reorg | 18:26 |
*** tobberydberg has quit IRC | 18:26 | |
odyssey4me | ah, reorgs - I suspect that sometimes it's just a way to keep people busy | 18:27 |
jmccrory | heh pretty much | 18:28 |
openstackgerrit | Merged openstack/openstack-ansible-lxc_hosts stable/ocata: Ubuntu: Use dnsmasq-base, not dnsmasq https://review.openstack.org/494266 | 18:32 |
openstackgerrit | Merged openstack/openstack-ansible-lxc_hosts stable/newton: Ubuntu: Use dnsmasq-base, not dnsmasq https://review.openstack.org/494267 | 18:32 |
*** marst_ has joined #openstack-ansible | 18:34 | |
*** poopcat has quit IRC | 18:35 | |
*** marst has quit IRC | 18:37 | |
*** weezS has joined #openstack-ansible | 18:44 | |
*** poopcat has joined #openstack-ansible | 18:45 | |
*** jwitko__ has joined #openstack-ansible | 18:46 | |
*** tobberydberg has joined #openstack-ansible | 18:46 | |
*** xdfil has joined #openstack-ansible | 18:47 | |
*** tobberydberg has quit IRC | 18:51 | |
*** armaan has joined #openstack-ansible | 19:03 | |
xdfil | is there a general option to deal with package conflicts when running books? Like for example I'm getting mariadb package conflicts | 19:03 |
*** pwiltsey has quit IRC | 19:11 | |
*** pwiltsey has joined #openstack-ansible | 19:15 | |
*** homeski has quit IRC | 19:16 | |
*** hybridpollo has joined #openstack-ansible | 19:16 | |
*** armaan has quit IRC | 19:18 | |
*** pbandark has quit IRC | 19:18 | |
*** armaan has joined #openstack-ansible | 19:18 | |
*** sauloaislan has quit IRC | 19:21 | |
*** hybridpollo has quit IRC | 19:23 | |
*** hybridpollo has joined #openstack-ansible | 19:23 | |
*** hybridpollo is now known as ants-in-my-eyes- | 19:24 | |
*** aludwar has quit IRC | 19:30 | |
*** aludwar has joined #openstack-ansible | 19:31 | |
*** sxc731 has quit IRC | 19:36 | |
*** firebat has joined #openstack-ansible | 19:46 | |
firebat | Hey can someone explain to me all the connections that get made for containers on the infra hosts? For example what does this mean: 22940463_eno12@if88. Additionally I have my neutron_agent lxc container running against an if85 that doesn’t seem to exist on the host it’s on should that matter? | 19:47 |
*** pwiltsey has quit IRC | 19:55 | |
xdfil | firebat: 22940463_eno12@if88 the first part is the cointainer ID right? | 19:55 |
*** pwiltsey has joined #openstack-ansible | 19:56 | |
*** bmchugh has joined #openstack-ansible | 19:56 | |
firebat | Yeah | 19:56 |
xdfil | idk you seem to understand it pretty well | 19:57 |
firebat | lol | 19:57 |
xdfil | that neutron container has no other if? | 19:57 |
firebat | It has a bunch of extras… I’ve definitely messed something up | 19:58 |
firebat | I’m like 99.9% complete and they want a working demo tomorrow so I’m in a bit of a freak out mode | 19:58 |
firebat | Essentially my agent host can’t ping the gateway of the external provider network, but it has an interface and address on that network (or so it seems) | 19:59 |
xdfil | hmm, did you mess with the invenotry.json file at all? that seems odd | 19:59 |
firebat | No but this morning I was cleaning and deploying a bunch | 19:59 |
firebat | and redeploying on top of existing stuff because I was in a hurry | 20:00 |
firebat | Should I destroy the containers and do a redeploy to clean up those routes? | 20:00 |
xdfil | I think you should start clean yes | 20:00 |
xdfil | re-deploying on existing stuff get tricky sometimes | 20:01 |
firebat | Yeah I told my boss that… he doesn’t like how long it takes to deply, but this is on old machiens with spinning disks | 20:02 |
xdfil | eh, bosses. You can't live with them.... | 20:02 |
xdfil | Once you iron out your config though the deployments are pretty fast considering all the stuff that is happening | 20:03 |
xdfil | Also point out that OSA architecture is better than any commercial/paid option at the moment | 20:04 |
firebat | One thing I haven’t had time to think about due to being so thick in the weeds… what’s my path to upgrading since deploying on top of a deploy can be tricky? | 20:04 |
openstackgerrit | Major Hayden proposed openstack/openstack-ansible-rabbitmq_server master: [WIP] Streamline RabbitMQ install https://review.openstack.org/493923 | 20:04 |
xdfil | Mirantis k8s looks cool, but from what I hear its not production ready | 20:04 |
firebat | Oh yeah that’s why we’re going with this, because we experienced VMWare VRA | 20:04 |
mhayden | hwoarang: pardon my lack of knowledge on zypper... is it possible to use addlock for a package spec w/version? like zypper addlock erlang*19.3* ? | 20:05 |
xdfil | to do minor version upgrades you just run the playbooks again the same as you ran them before to install | 20:05 |
xdfil | major versions require a bit more attention though | 20:06 |
firebat | Any good information on having DR with a somewhat basic osa cloud? I.e. we’ve got two DCs and we plan on (maybe) running two separate clouds… would DR need to be application based completely or could we configure things to double-up so to say. | 20:10 |
xdfil | DR as in automatic failover to the other DC? | 20:11 |
xdfil | if so, you'd have to build it yourself somewhat. | 20:12 |
xdfil | but that was my initial draw to OpenStack was that it enables you to build those sort of solutions | 20:13 |
*** poopcat has quit IRC | 20:14 | |
xdfil | Cells might be of use? IDK I havent fully understood or looked at how Cells work yet | 20:14 |
*** poopcat has joined #openstack-ansible | 20:14 | |
*** bmchugh has quit IRC | 20:15 | |
openstackgerrit | Major Hayden proposed openstack/openstack-ansible-lxc_hosts master: Compile SELinux module via async https://review.openstack.org/494316 | 20:21 |
*** pwiltsey_ has joined #openstack-ansible | 20:24 | |
*** pwiltsey has quit IRC | 20:25 | |
*** pwiltsey_ has quit IRC | 20:25 | |
firebat | Hmm ok yeah I’ve done something wrong here.. got an error on the redeploy.. mind if I paste here, and get some advice? (Having to do with my vxlan network/veth pairs) | 20:28 |
*** pwiltsey has joined #openstack-ansible | 20:28 | |
firebat | xdfil: Yeah that’s what I meant was automatic failover… the management team seems to have big dreams, but don’t realize we’ll have to work for it lol. | 20:29 |
firebat | Sorry for not getting back sooner I’m in deep as they say | 20:29 |
mhayden | odyssey4me / andymccr: patches for mariadb yum + erlang-solutions reverse proxies are pending | 20:30 |
firebat | https://pastebin.com/2zyzmXX1 | 20:31 |
*** ants-in-my-eyes- is now known as hybridpollo | 20:32 | |
*** pcaruana has quit IRC | 20:41 | |
*** os_user has quit IRC | 20:43 | |
openstackgerrit | Merged openstack/openstack-ansible master: Pin SHA's for stable/pike rc1 https://review.openstack.org/493816 | 20:45 |
*** pwiltsey has quit IRC | 20:46 | |
*** gouthamr has quit IRC | 20:46 | |
*** esberglu has quit IRC | 20:50 | |
mhayden | cloudnull: when you did the erlang stuff, was there a reason not to use esl-erlang? (the one that has all pkgs built in?) | 20:50 |
*** esberglu has joined #openstack-ansible | 20:50 | |
*** poopcat has quit IRC | 20:53 | |
*** poopcat has joined #openstack-ansible | 20:54 | |
*** armaan_ has joined #openstack-ansible | 20:54 | |
*** esberglu has quit IRC | 20:55 | |
*** armaan has quit IRC | 20:57 | |
openstackgerrit | Major Hayden proposed openstack/openstack-ansible-rabbitmq_server master: [WIP] Try the single esl-erlang package https://review.openstack.org/494324 | 20:57 |
*** lucasxu has quit IRC | 20:57 | |
*** xdfil has quit IRC | 20:59 | |
*** jwitko_ has quit IRC | 21:00 | |
*** craigs has joined #openstack-ansible | 21:01 | |
hwoarang | mhayden: i don't think you can lock based on package version. I will ask, but the alternative would be to install the version you want and then lock it | 21:02 |
hwoarang | i mean add a package lock which will prevent further changes to this package | 21:02 |
*** vishwana_ has quit IRC | 21:03 | |
*** esberglu has joined #openstack-ansible | 21:03 | |
*** vishwanathj has joined #openstack-ansible | 21:04 | |
*** jwitko__ has quit IRC | 21:04 | |
openstackgerrit | Major Hayden proposed openstack/openstack-ansible-lxc_hosts master: Compile SELinux module via async https://review.openstack.org/494316 | 21:05 |
mhayden | ah i see | 21:05 |
mhayden | i'm looking at possibly installing that big esl-erlang pkg to speed up the gate some | 21:05 |
mhayden | installing all of those individual erlang packages is slowing down centos/xenial | 21:05 |
*** woodard has quit IRC | 21:11 | |
openstackgerrit | Major Hayden proposed openstack/openstack-ansible-lxc_hosts master: Compile SELinux module via async https://review.openstack.org/494316 | 21:18 |
firebat | Hey guys even after I overwrite the host_bind_override my container keeps trying to join the br-vxlan at eth10 instead of what I’ve set it to… are there strict rules i.e. eth10 should be br-vxlan eth11 should be br-vlan, etc? | 21:18 |
*** vakuznet has quit IRC | 21:19 | |
*** thorst has quit IRC | 21:19 | |
*** aludwar has quit IRC | 21:43 | |
*** aludwar has joined #openstack-ansible | 21:44 | |
LiterateHawk | You might have to delete the facts? | 21:46 |
firebat | Thanks great idea | 21:49 |
*** thorst has joined #openstack-ansible | 21:53 | |
*** thorst has quit IRC | 21:53 | |
*** vishwana_ has joined #openstack-ansible | 21:54 | |
*** woodard has joined #openstack-ansible | 21:55 | |
*** vishwanathj has quit IRC | 21:58 | |
*** schwicht has quit IRC | 22:00 | |
*** germs has joined #openstack-ansible | 22:06 | |
*** armaan_ has quit IRC | 22:08 | |
cloudnull | mhayden: no reason that I can remember | 22:09 |
cloudnull | if that's faster we should investigate that | 22:09 |
openstackgerrit | Merged openstack/openstack-ansible-os_horizon master: tasks: horizon_install: Fix virtualenv-tools issue on openSUSE https://review.openstack.org/493743 | 22:18 |
*** drifterza has quit IRC | 22:19 | |
*** germs has quit IRC | 22:19 | |
*** esberglu has quit IRC | 22:22 | |
*** thorst has joined #openstack-ansible | 22:24 | |
openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible master: Use correct dnf.conf path for proxy https://review.openstack.org/493087 | 22:24 |
*** thorst has quit IRC | 22:29 | |
*** schwicht has joined #openstack-ansible | 22:29 | |
*** germs has joined #openstack-ansible | 22:33 | |
*** germs has left #openstack-ansible | 22:41 | |
*** weezS has quit IRC | 22:49 | |
*** BjoernT has quit IRC | 22:54 | |
*** itsgottabered has joined #openstack-ansible | 22:56 | |
itsgottabered | anyone around to help clear up some openvswitch confusion? | 22:56 |
*** schwicht has quit IRC | 22:58 | |
*** esberglu has joined #openstack-ansible | 22:59 | |
*** esberglu has quit IRC | 23:03 | |
*** marst_ has quit IRC | 23:06 | |
*** thorst has joined #openstack-ansible | 23:11 | |
*** thorst has quit IRC | 23:13 | |
*** schwicht has joined #openstack-ansible | 23:18 | |
*** gouthamr has joined #openstack-ansible | 23:27 | |
*** chyka has quit IRC | 23:32 | |
*** chyka has joined #openstack-ansible | 23:33 | |
*** firebat has quit IRC | 23:39 | |
*** itsgottabered has quit IRC | 23:42 | |
*** chyka has quit IRC | 23:45 | |
*** aludwar has quit IRC | 23:53 | |
*** aludwar has joined #openstack-ansible | 23:57 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!