| jra | well, this is only a dev cluster for me as well, so it's not huge and it's only running core services | 00:08 |
|---|---|---|
| jra | nonetheless, I expect things are likely fairly uniform at the -hosts level | 00:09 |
| cloudnull | they should be. | 00:09 |
| cloudnull | and most of the os_* roles are uniform | 00:10 |
| cloudnull | they all follow a similar pattern so they'll have a consistent deployment and override experience | 00:11 |
| goldenfri | hi can I trouble you for a sanity check. I have an AIO running and I want to add "external" floating IPs into neutron, instead of the built in "public" ones, to do that I need to add an additional network to the openstack_user_config right? | 00:13 |
| goldenfri | external being a flat non-vlan network that is connected by its self to a single nic | 00:14 |
| *** flemingo has quit IRC | 00:16 | |
| cloudnull | yes you can add floating ips to a flat network | 00:16 |
| goldenfri | - network: | 00:17 |
| goldenfri | container_bridge: "br-vlan" | 00:17 |
| goldenfri | container_type: "veth" | 00:17 |
| goldenfri | container_interface: "eth13" | 00:17 |
| goldenfri | host_bind_override: "enp5s0" | 00:17 |
| goldenfri | type: "flat" | 00:17 |
| goldenfri | #range: "1:1" | 00:17 |
| goldenfri | net_name: "external" | 00:17 |
| goldenfri | group_binds: | 00:17 |
| goldenfri | - neutron_linuxbridge_agent | 00:17 |
| goldenfri | ok so does something like this make sense? | 00:17 |
| goldenfri | not sure if I need the range or not | 00:17 |
| idlemind | goldenfri paste.openstack.org is your friend | 00:17 |
| goldenfri | oh yea duh sorry | 00:17 |
| cloudnull | idlemind: ++ | 00:18 |
| cloudnull | goldenfri: assuming enp5s0 is a free interface that will all work | 00:19 |
| idlemind | goldenfrie i'd have to check it against mine but it looks right as long enp5s0 exists in your system and connects to the network you expect that you've ranged out for "external" | 00:19 |
| cloudnull | ha! exactly what idlemind said :) | 00:19 |
| goldenfri | ok that is the int I am using for external_lb_vip_address is that going to break anything? | 00:19 |
| *** osnaya has joined #openstack-ansible | 00:19 | |
| cloudnull | when using a flat network I generally give neutron a free network interface | 00:22 |
| osnaya | @cloudnull @evrardjp Need clarification regarding....Openstack-ansible deploying on the Ubuntu Host (not containers) in lab…. If there is no explicit storage host but plan to use local storage on the computes…. Have question about /etc/openstack-deploy/openstack-user-config.yml setup values related to storage host…. Should the cinder storage host (storage hosts:…. storage1: ip:…section) be commented out or shoul | 00:23 |
| cloudnull | because I dont trust how neutron will handle it | 00:23 |
| goldenfri | yea, they are the same flat networks so I was hoping to get away with it if I can. | 00:23 |
| goldenfri | well one way to find out :) | 00:23 |
| *** chyka_ has joined #openstack-ansible | 00:23 | |
| idlemind | goldenfri i'd be nervous | 00:23 |
| cloudnull | you can, one way that I get around that is to create a bridge and hang a veth pair off of it | 00:24 |
| idlemind | goldenfri for the flat networks neutron "plugs" them into a bridge | 00:24 |
| *** sawblade_ has joined #openstack-ansible | 00:24 | |
| idlemind | so it might change how you've ip'd that external_lb_vip | 00:24 |
| goldenfri | ah ok hrm | 00:24 |
| idlemind | so ya veth pairing | 00:24 |
| idlemind | or qinq on-top of the existing br-vlan (i did qinq for myself but i have the necessary hardware for it) | 00:24 |
| cloudnull | goldenfri: https://github.com/openstack/openstack-ansible/blob/master/etc/network/interfaces.d/aio_interfaces.cfg#L44-L59 | 00:25 |
| *** sawblade6 has quit IRC | 00:25 | |
| osnaya | @cloudnull @evrardjp correcting typo above.... Should the cinder storage host (storage hosts:…. storage1: ip:…section) be commented out or should we specify the IP address for storage1: and iscsi_ip_address: value blank?) | 00:25 |
| goldenfri | ok, I'll have look into that, thank you for the help! | 00:26 |
| *** electrocucaracha has left #openstack-ansible | 00:26 | |
| cloudnull | that's what I use on a combined network setup where a flat network exists | 00:26 |
| goldenfri | one last question, how much do I have re-run when messing with the networks? host, infra or openstack? | 00:26 |
| cloudnull | in that case eth12 is the raw flat interface given to neutron | 00:27 |
| *** chyka has quit IRC | 00:27 | |
| cloudnull | osnaya: if you are using storage hosts you will need to fill in those sections | 00:27 |
| cloudnull | in general I set the ip address of the storage host and will have a storage netowrk and set the "iscsi_ip_address" to whatever address is on the storage network | 00:28 |
| *** chyka_ has quit IRC | 00:28 | |
| osnaya | @cloudnull In current setup, there is no separate storage host, plan to use local storage on compute... | 00:29 |
| *** jra has quit IRC | 00:29 | |
| cloudnull | then you'd set both option to the same thing | 00:30 |
| cloudnull | you can use the jinja variable but in general I set it to the address and leave nothing to chance | 00:30 |
| idlemind | osnaya the storage has to be stitched in via iscsi for 1 host or 12 hosts ... when you add you're second host you'll see it just connects up to first for all it's storage needs :) (unless you make that one have storage too, then both get disks and the cinder scheduler places volumes based on it's whims) | 00:32 |
| osnaya | @cloudnull when no storage host, do I comment out the storage host section in openstack-user-config.yml OR just keep the ip blank? | 00:32 |
| cloudnull | I'd just remove it if you have none | 00:33 |
| osnaya | @cloudnull is the above comment for me? remove the section from user-config.yml? or is that comment wrt another conversation? | 00:34 |
| cloudnull | yes for you, If you have no storage hosts you can just delete that section out of config | 00:34 |
| osnaya | @cloudnull thanks for the clarification.... will do that | 00:35 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_nova master: Convert role to use a common systemd init role https://review.openstack.org/466179 | 00:37 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_keystone master: Convert role to use a common systemd init role https://review.openstack.org/551028 | 00:38 |
| osnaya | @idlemind thanks for the pointer... for now, I don't have storage in my lab setup | 00:49 |
| idlemind | osnaya no worries; my lab uses lvm and i had to tweak some horizon timers to be able to use larger images or multi-image deploys ... let me know if you add in lvm and start hitting "timeout" style issues | 00:50 |
| idlemind | * not horizon; nova timers i think actually | 00:50 |
| idlemind | ... now i'm drooling over my freshly arrived 8tb 3.5" sata disks which will likely force me to start using ceph in the lab | 00:51 |
| idlemind | also; where has "night light" been this whole time. my old eyes are loving it | 00:53 |
| *** prometheanfire has quit IRC | 00:59 | |
| *** prometheanfire has joined #openstack-ansible | 01:04 | |
| *** hamzy has joined #openstack-ansible | 01:06 | |
| cloudnull | idlemind: night light? | 01:07 |
| idlemind | fedora/gnome feature | 01:07 |
| cloudnull | ah neat | 01:07 |
| idlemind | ya tons down blue light i guess | 01:07 |
| idlemind | * tones | 01:07 |
| cloudnull | an neat. | 01:08 |
| * cloudnull is looking if its available in kde | 01:08 | |
| *** ANKITA has joined #openstack-ansible | 01:18 | |
| ANKITA | HI | 01:18 |
| cloudnull | o/ | 01:22 |
| *** jwitko has quit IRC | 01:22 | |
| ANKITA | openstack-ansible setup-infrastructure.yml fails at https://pastebin.com/Y0bZX3Sv | 01:23 |
| cloudnull | ANKITA: looks like http://127.245.178.31:8181/os-releases/16.0.9/ubuntu-16.04-x86_64/requirements_absolute_requirements.txt is missing ? | 01:26 |
| cloudnull | is http://127.245.178.31:8181/os-releases/16.0.9 up ? | 01:26 |
| ANKITA | how to check it? | 01:28 |
| ANKITA | shall i download it? | 01:28 |
| ANKITA | osrelease for ubuntu? | 01:28 |
| cloudnull | mhayden: http://logs.openstack.org/79/466179/19/check/openstack-ansible-functional-centos-7/ce8f2db/job-output.txt.gz#_2018-03-09_00_44_41_547451 - looks like we're seeing package conflicts in cent ? | 01:29 |
| cloudnull | ANKITA: check your loadbalancer and see if the repo server is up | 01:29 |
| ANKITA | ok | 01:29 |
| ANKITA | ping works | 01:32 |
| ANKITA | so assume its acessible | 01:32 |
| cloudnull | ping but is port 8181 up ? | 01:33 |
| cloudnull | hatop -s /var/run/haproxy.stat | 01:33 |
| cloudnull | assuming you're using haproxy | 01:33 |
| ANKITA | yes i am | 01:34 |
| ANKITA | The program 'hatop' is currently not installed. shall i install itapt install hatop | 01:35 |
| cloudnull | it'll be installed on your haproxy node | 01:36 |
| ANKITA | telnet host 8181 telnet: could not resolve host/8181: No address associated with hostname | 01:37 |
| cloudnull | ? | 01:38 |
| ANKITA | hatop is down | 01:39 |
| ANKITA | shall i paste op? | 01:39 |
| cloudnull | so you see in the hatop interface that the repo server on port 8181 is down ? | 01:40 |
| ANKITA | yes | 01:40 |
| ANKITA | status is down | 01:40 |
| cloudnull | so that'll need to be something you'll have to resolve | 01:41 |
| cloudnull | check the repo server container a | 01:41 |
| cloudnull | and see if nginx is running | 01:41 |
| ANKITA | OK | 01:41 |
| ANKITA | OP WAS https://pastebin.com/Mu8n44xN | 01:41 |
| ANKITA | LET Me check nginx | 01:41 |
| *** sawblade_ has quit IRC | 01:41 | |
| cloudnull | that's the aodh server check the repo_server entries too | 01:42 |
| *** sawblade6 has joined #openstack-ansible | 01:42 | |
| cloudnull | but check nginx in the repo server containers | 01:42 |
| ANKITA | nginx inn repo container of infra is running | 01:44 |
| ANKITA | active status | 01:44 |
| cloudnull | port 8181 up? | 01:44 |
| ANKITA | how to check that hatop cmd doesn't work here | 01:45 |
| cloudnull | ss -ntlp | 01:45 |
| ANKITA | thanks | 01:45 |
| ANKITA | in o/p send q has values but nothing in receivedq column so is it down? | 01:47 |
| ANKITA | https://pastebin.com/ReaVTGPK | 01:47 |
| cloudnull | `ss -ntlp | grep 8181` | 01:48 |
| cloudnull | on your paste, it looks like 8181 is up | 01:48 |
| cloudnull | can you curl `http://localhost:8181` | 01:48 |
| ANKITA | LISTEN 0 128 *:8181 *:* users:(("nginx",pid=2021,fd=8),("nginx",pid=2020,fd=8),("nginx",pid=2019,fd=8),("nginx",pid=2018,fd=8),("nginx",pid=2017,fd=8),("nginx",pid=2016,fd=8),("nginx",pid=2015,fd=8),("nginx",pid=2014,fd=8),("nginx",pid=2013,fd=8),("nginx",pid=2012,fd=8),("nginx",pid=2011,fd=8),("nginx",pid=2010,fd=8),("nginx",pid=2009,fd=8)) | 01:48 |
| cloudnull | or the ip address in the container | 01:49 |
| ANKITA | curl connection refused | 01:49 |
| ANKITA | let me chek ip | 01:50 |
| cloudnull | something like `curl http://$(ip -o r g 1 | awk '{print $7}'):8181` | 01:50 |
| ANKITA | i can ping to ip of lxc-info --name infra1_repo_container-98365f98 | 01:53 |
| ANKITA | op/ of `curl http://$(ip -o r g 1 | awk '{print $7}'):8181` was 0000000 | 01:53 |
| ANKITA | connection to port refused | 01:53 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_glance master: Convert role to use a common systemd init role https://review.openstack.org/551053 | 02:04 |
| cloudnull | if you're unable to run that command within the repo server container and it return an http response something is not right | 02:06 |
| cloudnull | you'll need to resolve whatever issue is making it so your repo server is not responding | 02:07 |
| ANKITA | am not in container now | 02:08 |
| *** vnogin has joined #openstack-ansible | 02:08 | |
| cloudnull | oh, that command will need to be run within the repo server container | 02:08 |
| ANKITA | am at infra | 02:08 |
| ANKITA | ok | 02:08 |
| cloudnull | or the haproxy node | 02:08 |
| ANKITA | srry | 02:08 |
| ANKITA | haproxy node ie equvalent to infra? | 02:09 |
| cloudnull | however i think we've estabilished that the haproxy server is not connecting you back to the repo server | 02:09 |
| cloudnull | it can be the same, that's all dependent on how you've setup the openstack_user_config | 02:10 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible master: Follow the new PTI for document build https://review.openstack.org/551054 | 02:11 |
| *** vnogin has quit IRC | 02:12 | |
| *** Leo_m has quit IRC | 02:12 | |
| *** Leo_m_ has joined #openstack-ansible | 02:13 | |
| osnaya | @cloudnull with NO separate storage host scenario, commented storage_host: section in openstack_user_config.yml and run openstack-ansible setup-hosts.yml --syntax-check ---> throws below error | 02:13 |
| osnaya | @cloudnull Variable files: "-e @/etc/openstack_deploy/user_secrets.yml -e @/etc/openstack_deploy/user_variables.yml " ERROR! Attempted to execute "/opt/openstack-ansible/playbooks/inventory/dynamic_inventory.py" as inventory script: Inventory script (/opt/openstack-ansible/playbooks/inventory/dynamic_inventory.py) had an execution error: Traceback (most recent call last): File "/opt/openstack-ansible/playbooks/inventory/ | 02:13 |
| ANKITA | in the container curl commond not found | 02:14 |
| cloudnull | osnaya: can you execute /opt/openstack-ansible/playbooks/inventory/dynamic_inventory.py | 02:14 |
| cloudnull | does that return ? | 02:14 |
| osnaya | @cloudnull output = generate.main(**all_args) File "/opt/openstack-ansible/osa_toolkit/generate.py", line 1141, in main user_defined_config File "/opt/openstack-ansible/osa_toolkit/generate.py", line 747, in container_skel_load address_prefix=p_net.get('address_prefix') File "/opt/openstack-ansible/osa_toolkit/generate.py", line 576, in _add_additional_networks address_prefix File "/opt/openstack-an | 02:15 |
| cloudnull | ANKITA: yes the containers are quite minimal, if you need debug utils you can install them | 02:15 |
| ANKITA | cloudnull curl commands not found in repo containers | 02:15 |
| osnaya | File "/opt/openstack-ansible/osa_toolkit/generate.py", line 650, in _add_additional_networks if user_config[cphg].get(container_host) is not None: KeyError: u'storage_hosts' | 02:15 |
| ANKITA | sure | 02:15 |
| ANKITA | i'll | 02:15 |
| cloudnull | osnaya: mind sharing your openstack_user_config.yml ? | 02:16 |
| cloudnull | pasting it in http://paste.openstack.org/ | 02:16 |
| cloudnull | If you can I'd be happy to test it out and see what's what | 02:17 |
| cloudnull | also osnaya what is the release you're using ? | 02:17 |
| osnaya | @cloudnull Pike | 02:17 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible master: Follow the new PTI for document build https://review.openstack.org/551054 | 02:24 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-apt_package_pinning master: Follow the new PTI for document build https://review.openstack.org/551057 | 02:34 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-apt_package_pinning master: Follow the new PTI for document build https://review.openstack.org/551057 | 02:34 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_glance master: Convert role to use a common systemd init role https://review.openstack.org/551053 | 02:36 |
| *** osnaya has quit IRC | 02:38 | |
| *** woodard has joined #openstack-ansible | 02:42 | |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_molteniron master: Follow the new PTI for document build https://review.openstack.org/551061 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-ceph_client master: Follow the new PTI for document build https://review.openstack.org/551062 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_monasca-agent master: Follow the new PTI for document build https://review.openstack.org/551063 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_zaqar master: Follow the new PTI for document build https://review.openstack.org/551064 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_monasca-ui master: Follow the new PTI for document build https://review.openstack.org/551066 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_designate master: Follow the new PTI for document build https://review.openstack.org/551067 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_cloudkitty master: Follow the new PTI for document build https://review.openstack.org/551065 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_ironic master: Follow the new PTI for document build https://review.openstack.org/551068 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_monasca master: Follow the new PTI for document build https://review.openstack.org/551070 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-pip_install master: Follow the new PTI for document build https://review.openstack.org/551069 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-ops master: Follow the new PTI for document build https://review.openstack.org/551071 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-repo_server master: Follow the new PTI for document build https://review.openstack.org/551072 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_trove master: Follow the new PTI for document build https://review.openstack.org/551073 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_almanach master: Follow the new PTI for document build https://review.openstack.org/551074 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_rally master: Follow the new PTI for document build https://review.openstack.org/551075 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-tests master: Follow the new PTI for document build https://review.openstack.org/551078 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_nova master: Follow the new PTI for document build https://review.openstack.org/551076 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_magnum master: Follow the new PTI for document build https://review.openstack.org/551077 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-rabbitmq_server master: Follow the new PTI for document build https://review.openstack.org/551079 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-lxc_container_create master: Follow the new PTI for document build https://review.openstack.org/551080 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_watcher master: Follow the new PTI for document build https://review.openstack.org/551099 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-repo_build master: Follow the new PTI for document build https://review.openstack.org/551102 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_octavia master: Follow the new PTI for document build https://review.openstack.org/551103 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_searchlight master: Follow the new PTI for document build https://review.openstack.org/551104 | 02:48 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-plugins master: Follow the new PTI for document build https://review.openstack.org/551105 | 02:48 |
| *** esberglu has joined #openstack-ansible | 03:02 | |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-apt_package_pinning master: Follow the new PTI for document build https://review.openstack.org/551057 | 03:05 |
| *** ansmith has quit IRC | 03:07 | |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible master: Follow the new PTI for document build https://review.openstack.org/551109 | 03:08 |
| *** hamzy has quit IRC | 03:14 | |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-apt_package_pinning master: Follow the new PTI for document build https://review.openstack.org/551057 | 03:29 |
| *** mburrows has joined #openstack-ansible | 03:44 | |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-apt_package_pinning master: Follow the new PTI for document build https://review.openstack.org/551057 | 03:45 |
| *** woodard_ has joined #openstack-ansible | 03:54 | |
| *** woodard has quit IRC | 03:54 | |
| *** gkadam has joined #openstack-ansible | 03:55 | |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-apt_package_pinning master: Follow the new PTI for document build https://review.openstack.org/551057 | 03:59 |
| *** esberglu has quit IRC | 04:10 | |
| *** dave-mccowan has quit IRC | 04:12 | |
| *** hamzy has joined #openstack-ansible | 04:23 | |
| *** ANKITA has quit IRC | 04:23 | |
| *** coolj has quit IRC | 04:25 | |
| *** armaan has quit IRC | 04:27 | |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_monasca master: Follow the new PTI for document build https://review.openstack.org/551120 | 04:27 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-ceph_client master: Follow the new PTI for document build https://review.openstack.org/551121 | 04:27 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_molteniron master: Follow the new PTI for document build https://review.openstack.org/551122 | 04:27 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_magnum master: Follow the new PTI for document build https://review.openstack.org/551123 | 04:27 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_rally master: Follow the new PTI for document build https://review.openstack.org/551124 | 04:27 |
| *** armaan has joined #openstack-ansible | 04:27 | |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_heat master: Follow the new PTI for document build https://review.openstack.org/551125 | 04:27 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_ceilometer master: Follow the new PTI for document build https://review.openstack.org/551126 | 04:27 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-lxc_hosts master: Follow the new PTI for document build https://review.openstack.org/551127 | 04:27 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_sahara master: Follow the new PTI for document build https://review.openstack.org/551128 | 04:27 |
| *** coolj has joined #openstack-ansible | 04:27 | |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-apt_package_pinning master: Follow the new PTI for document build https://review.openstack.org/551057 | 04:27 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_aodh master: Follow the new PTI for document build https://review.openstack.org/551130 | 04:27 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_zaqar master: Follow the new PTI for document build https://review.openstack.org/551129 | 04:27 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-pip_install master: Follow the new PTI for document build https://review.openstack.org/551131 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_swift master: Follow the new PTI for document build https://review.openstack.org/551132 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_octavia master: Follow the new PTI for document build https://review.openstack.org/551133 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_searchlight master: Follow the new PTI for document build https://review.openstack.org/551134 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_monasca-agent master: Follow the new PTI for document build https://review.openstack.org/551135 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-openstack_openrc master: Follow the new PTI for document build https://review.openstack.org/551136 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_watcher master: Follow the new PTI for document build https://review.openstack.org/551137 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_designate master: Follow the new PTI for document build https://review.openstack.org/551138 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_barbican master: Follow the new PTI for document build https://review.openstack.org/551139 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-galera_client master: Follow the new PTI for document build https://review.openstack.org/551140 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-haproxy_server master: Follow the new PTI for document build https://review.openstack.org/551142 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_almanach master: Follow the new PTI for document build https://review.openstack.org/551141 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-lxc_container_create master: Follow the new PTI for document build https://review.openstack.org/551143 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-galera_server master: Follow the new PTI for document build https://review.openstack.org/551144 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_monasca-ui master: Follow the new PTI for document build https://review.openstack.org/551145 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-repo_build master: Follow the new PTI for document build https://review.openstack.org/551146 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_gnocchi master: Follow the new PTI for document build https://review.openstack.org/551147 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-plugins master: Follow the new PTI for document build https://review.openstack.org/551148 | 04:28 |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible-os_keystone master: Follow the new PTI for document build https://review.openstack.org/551149 | 04:28 |
| *** chyka has joined #openstack-ansible | 04:30 | |
| *** chhagarw has joined #openstack-ansible | 04:31 | |
| *** chhavi__ has joined #openstack-ansible | 04:31 | |
| *** chyka has quit IRC | 04:34 | |
| openstackgerrit | zhongshengping proposed openstack/openstack-ansible master: Follow the new PTI for document build https://review.openstack.org/551109 | 04:37 |
| *** clone99 has quit IRC | 04:53 | |
| *** ivve has joined #openstack-ansible | 04:55 | |
| *** poopcat has quit IRC | 04:59 | |
| *** vnogin has joined #openstack-ansible | 05:08 | |
| *** vnogin has quit IRC | 05:13 | |
| *** mburrows has quit IRC | 05:20 | |
| *** udesale has joined #openstack-ansible | 05:27 | |
| *** hamza21 has joined #openstack-ansible | 05:33 | |
| *** radeks__ has joined #openstack-ansible | 06:04 | |
| *** radeks_ has joined #openstack-ansible | 06:06 | |
| *** radeks__ has quit IRC | 06:09 | |
| *** armaan has quit IRC | 06:18 | |
| *** armaan has joined #openstack-ansible | 06:18 | |
| *** mburrows has joined #openstack-ansible | 06:23 | |
| *** masber has quit IRC | 06:26 | |
| *** gkadam has quit IRC | 06:28 | |
| *** sar has joined #openstack-ansible | 06:43 | |
| *** Sha000000 has joined #openstack-ansible | 06:52 | |
| *** lvdombrkr has joined #openstack-ansible | 07:04 | |
| *** woodard_ has quit IRC | 07:09 | |
| *** lvdombrkr89 has joined #openstack-ansible | 07:12 | |
| *** lvdombrkr has quit IRC | 07:14 | |
| *** Sha000000 has quit IRC | 07:24 | |
| *** pcaruana has joined #openstack-ansible | 07:27 | |
| *** eumel8 has joined #openstack-ansible | 07:32 | |
| *** woodard has joined #openstack-ansible | 07:47 | |
| *** PTO has joined #openstack-ansible | 07:51 | |
| *** woodard has quit IRC | 07:52 | |
| gokhan_ | hi odyssey4me evrardjp ı wrote my role and see that venv of myservice is prepared with repo build. but when role tries to retrieve checksum for venv download, it gives http 404 error for download url. ı didn't find where I did wrong things. this is error log: http://paste.openstack.org/show/696017/ this is task:https://github.com/b3lab/openstack-ansible-b3lab_safirnotification/blob/master/tasks/safirnotification_install.yml#L62 what is your th | 07:58 |
| gokhan_ | oughts about this error? | 07:58 |
| *** lvdombrkr89 has quit IRC | 08:00 | |
| *** lvdombrkr89 has joined #openstack-ansible | 08:01 | |
| evrardjp | gokhan_: could you check your repo build logs? | 08:01 |
| evrardjp | I haven't tried with things like https://github.com/b3lab/openstack-ansible-b3lab_safirnotification/blob/master/defaults/main.yml#L72-L73 | 08:02 |
| *** epalper has joined #openstack-ansible | 08:08 | |
| *** MrWatson has quit IRC | 08:09 | |
| *** NostawRm has joined #openstack-ansible | 08:09 | |
| *** lvdombrkr has joined #openstack-ansible | 08:10 | |
| *** lvdombrkr89 has quit IRC | 08:10 | |
| gokhan_ | evrardjp, I have tried it before . it worked. https://github.com/b3lab/openstack-ansible-os_horizon/blob/safir/pike/defaults/main.yml#L303-L306 | 08:19 |
| gokhan_ | evrardjp, ı will run repo-build with -vvv | 08:22 |
| *** gkadam has joined #openstack-ansible | 08:25 | |
| *** armaan has quit IRC | 08:27 | |
| *** armaan has joined #openstack-ansible | 08:27 | |
| *** electrofelix has joined #openstack-ansible | 08:28 | |
| ThomasS | hi | 08:30 |
| *** lvdombrkr89 has joined #openstack-ansible | 08:41 | |
| *** lvdombrkr has quit IRC | 08:44 | |
| *** woodard has joined #openstack-ansible | 08:46 | |
| *** lvdombrkr has joined #openstack-ansible | 08:48 | |
| *** threestrands_ has quit IRC | 08:48 | |
| *** lvdombrkr89 has quit IRC | 08:49 | |
| *** woodard has quit IRC | 08:51 | |
| *** radeks__ has joined #openstack-ansible | 08:58 | |
| *** ygl has joined #openstack-ansible | 09:01 | |
| *** radeks_ has quit IRC | 09:01 | |
| ygl | hi all | 09:01 |
| ygl | when I do a AIO install the neutron role is failing | 09:01 |
| ygl | can someone help me please | 09:01 |
| ygl | cloudnull: hi, can you help me please | 09:02 |
| ygl | cloudnull: http://paste.openstack.org/show/696042/ | 09:03 |
| evrardjp | ygl: you need ovs ? | 09:06 |
| ygl | evrardjp: yes. I see this error in the neutron container | 09:06 |
| ygl | evrardjp: rmmod: ERROR: Module bridge is in use by: br_netfilter | 09:07 |
| ygl | evrardjp: aio1-neutron-server-container-f95665c5 systemd[1]: Failed to start Open vSwitch Forwarding Unit | 09:07 |
| evrardjp | yeah I guess | 09:07 |
| evrardjp | Probably some conditionals to add | 09:07 |
| *** hamza21 has quit IRC | 09:07 | |
| evrardjp | the ovs path is not exercised yet | 09:07 |
| ygl | evrardjp: where exactly ? | 09:07 |
| ygl | evrardjp: how to proceed npw ? | 09:08 |
| evrardjp | I mean ovs is not tested automatically | 09:08 |
| evrardjp | so there is room for improvement there | 09:08 |
| evrardjp | I heard ppl working on it yesterday | 09:08 |
| evrardjp | you're on master or queens I guess? | 09:08 |
| ygl | evrardjp: ovs-ctl[8164]: modprobe: ERROR: ../libkmod/libkmod.c:586 kmod_search_moddep() could not open moddep file '/lib/modules/4.4.0-116-generic/modules.dep.bin' | 09:08 |
| openstackgerrit | Taseer Ahmed proposed openstack/openstack-ansible master: [WIP] Integrate Blazar with OpenStack Ansible https://review.openstack.org/549956 | 09:08 |
| evrardjp | ygl: my question still stands: do you really need ovs ? | 09:08 |
| evrardjp | should I say: why do you need ovs? | 09:09 |
| ygl | evrardjp: I am just executing the ansible scripts in an ubuntu vm as per https://docs.openstack.org/openstack-ansible/queens/contributor/quickstart-aio.html | 09:09 |
| *** radeks__ has quit IRC | 09:09 | |
| evrardjp | the quickstart aio isn't doing ovs | 09:09 |
| ygl | evrardjp: so how should I proceed now ? | 09:10 |
| evrardjp | Sorry to focus on that but I think someone new to openstack-ansible should evaluate linux bridge first: more stable, more tested and works for the common use cases | 09:10 |
| ygl | evrardjp: is this bug in queens ansible ? | 09:10 |
| ygl | evrardjp: how to change it to linuc bridge ? | 09:10 |
| evrardjp | ygl: so you're following the deploy guide or the quickstart aio? | 09:11 |
| ygl | evrardjp: quickstart guide | 09:11 |
| evrardjp | because if you followed quickstart aio, nothing is doing ovs there, so you probably added a var | 09:11 |
| evrardjp | removing them should do the trick | 09:11 |
| ygl | evrardjp: how to remove an what to remove ? | 09:12 |
| evrardjp | ygl: don't get me wrong, it's possible to get ovs working, it just needs effort | 09:12 |
| evrardjp | ygl: what did you change vs the guide? | 09:12 |
| ygl | evrardjp: nothing. I am just following the quickstart aio as it is | 09:12 |
| evrardjp | that's not possible | 09:13 |
| evrardjp | let's evaluate what happened then | 09:13 |
| ygl | I am executing this "openstack-ansible os-neutron-install.yml" | 09:13 |
| evrardjp | could you show what's in your /etc/openstack_deploy/user_variables.yml ? | 09:14 |
| ygl | from the host | 09:14 |
| ygl | ok | 09:14 |
| evrardjp | ygl: the quickstart aio is an All In One machine, so the deploy node and the target node are the same | 09:14 |
| evrardjp | if you had simple linux configured, no fancy networking, the process should work just fine | 09:14 |
| ygl | evrardjp: http://paste.openstack.org/show/696048/ | 09:15 |
| evrardjp | cat /etc/network/interfaces ? | 09:15 |
| evrardjp | are you using ovs on the host already? | 09:15 |
| evrardjp | start with a clean ubuntu node | 09:16 |
| ygl | evrardjp: ok | 09:16 |
| ygl | evrardjp: I am not using ovs on the node | 09:16 |
| ygl | evrardjp: http://paste.openstack.org/show/696052/ | 09:17 |
| evrardjp | you have two source there, probably worth removing the first one | 09:17 |
| evrardjp | what about /etc/network/interfaces.d/*.cfg ? | 09:17 |
| ygl | evrardjp: u mean it should have only one interface on the host ? | 09:17 |
| evrardjp | no | 09:18 |
| evrardjp | I meant you have two lines that source /etc/network/interfaces.d/* and /etc/network/interfaces.d/*.cfg | 09:18 |
| evrardjp | that sounds bad | 09:18 |
| evrardjp | just having source /etc/network/interfaces.d/*.cfg is enough | 09:18 |
| evrardjp | and then I am wondering what's in those files | 09:18 |
| ygl | evrardjp: ok I will start a fresh installation | 09:19 |
| evrardjp | we can figure out the problem first if you like | 09:19 |
| evrardjp | but I think if you have many small misconfigurations like that, it can become a mess very quickly | 09:19 |
| ygl | evrardjp: anyway I am destroying that vm | 09:20 |
| ygl | evrardjp: I am installing a new vm ubuntu 16 | 09:23 |
| ygl | evrardjp: thanks for your help | 09:24 |
| *** masber has joined #openstack-ansible | 09:25 | |
| *** ygl has quit IRC | 09:27 | |
| hwoarang | good morning | 09:29 |
| hwoarang | is anyone looking at the ceph-ansible link failures that affect all aio jobs http://logs.openstack.org/70/550870/1/check/openstack-ansible-linters/065d3f3/job-output.txt.gz#_2018-03-08_15_40_42_296213 ? | 09:29 |
| evrardjp | hey hwoarang | 09:30 |
| evrardjp | https://review.openstack.org/#/c/550542/ | 09:32 |
| *** holser__ has joined #openstack-ansible | 09:32 | |
| evrardjp | the follow up would be https://review.openstack.org/#/c/550749/3 | 09:33 |
| evrardjp | but we have things to iron out | 09:33 |
| hwoarang | ok thanks evrardjp | 09:36 |
| hwoarang | hmm. ubuntu times out. maybe we should move that to aio_basekit due to node slowness? | 09:37 |
| hwoarang | so we can get things moving a bit? | 09:37 |
| hwoarang | iirc in ptg they said that gates are slower due to meltdown/spectre mess | 09:38 |
| evrardjp | yeah | 09:38 |
| evrardjp | we discussed that with odyssey4me yesterday | 09:39 |
| *** mbuil has joined #openstack-ansible | 09:39 | |
| evrardjp | I think right now we should merge things, and get queens deadline out of the door | 09:40 |
| openstackgerrit | Adrien Cunin proposed openstack/openstack-ansible-lxc_hosts master: Also copy /etc/apt/sources.list.d/ into containers https://review.openstack.org/551229 | 09:44 |
| *** holser__ has quit IRC | 09:45 | |
| *** SmearedBeard has quit IRC | 09:45 | |
| *** holser__ has joined #openstack-ansible | 09:47 | |
| *** lvdombrkr has quit IRC | 09:48 | |
| evrardjp | hwoarang: we can discuss it together if you like | 09:49 |
| *** woodard has joined #openstack-ansible | 09:51 | |
| *** woodard has quit IRC | 09:55 | |
| hwoarang | we can discuss | 09:55 |
| hwoarang | i presume the timeouts affect queens so that would help merging things faster right? | 09:56 |
| *** chyka has joined #openstack-ansible | 09:57 | |
| *** SmearedBeard has joined #openstack-ansible | 09:57 | |
| *** chyka has quit IRC | 10:01 | |
| *** vnogin has joined #openstack-ansible | 10:02 | |
| *** Smeared_Beard has joined #openstack-ansible | 10:02 | |
| *** SmearedBeard has quit IRC | 10:04 | |
| *** armaan has quit IRC | 10:06 | |
| *** armaan has joined #openstack-ansible | 10:06 | |
| jrosser | should this work? https://www.irccloud.com/pastebin/XHwwooVa/ | 10:10 |
| *** armaan has quit IRC | 10:11 | |
| evrardjp | hwoarang: the idea is to refactor the jobs | 10:23 |
| hwoarang | ok yep but would aio_basekit be a temp solution until we get queens out? | 10:27 |
| hwoarang | though we seem to have all sort of different issues. timeout is just one of them | 10:27 |
| hwoarang | hmm seems like centos mirror are sick again | 10:27 |
| *** udesale_ has joined #openstack-ansible | 10:29 | |
| openstackgerrit | git-harry proposed openstack/openstack-ansible-tests master: Add tests for APTRepositoryCacheUpdateRule https://review.openstack.org/551246 | 10:30 |
| *** vnogin has quit IRC | 10:30 | |
| *** udesale has quit IRC | 10:32 | |
| *** udesale has joined #openstack-ansible | 10:33 | |
| *** udesale_ has quit IRC | 10:33 | |
| *** udesale has quit IRC | 10:40 | |
| *** woodard has joined #openstack-ansible | 10:44 | |
| *** vnogin has joined #openstack-ansible | 10:46 | |
| *** woodard has quit IRC | 10:48 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/queens: Clean-up gate code to use mirror_info and pin ceph-ansible SHA https://review.openstack.org/551251 | 10:54 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/queens: Clean-up gate code to use mirror_info https://review.openstack.org/551251 | 10:55 |
| odyssey4me | gokhan_ the download URL for the venv/checksum is provided in the group_vars - for example https://github.com/openstack/openstack-ansible/blob/master/inventory/group_vars/horizon_all.yml#L41-L42 - you would need to have similar vars in an appropriate group_vars file in /etc/openstack_deploy/group_vars | 10:57 |
| *** mburrows has quit IRC | 11:04 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/pike: Clean-up gate code to use mirror_info https://review.openstack.org/551258 | 11:11 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/ocata: Clean-up gate code to use mirror_info https://review.openstack.org/551259 | 11:19 |
| *** hw_wutianwei has quit IRC | 11:22 | |
| *** hw_wutianwei has joined #openstack-ansible | 11:22 | |
| *** SmearedBeard has joined #openstack-ansible | 11:23 | |
| *** Smeared_Beard has quit IRC | 11:24 | |
| *** vnogin has quit IRC | 11:26 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/newton: Clean-up gate code to use mirror_info https://review.openstack.org/551260 | 11:27 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/pike: Clean-up gate code to use mirror_info https://review.openstack.org/551258 | 11:28 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/ocata: Clean-up gate code to use mirror_info https://review.openstack.org/551259 | 11:31 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/ocata: Clean-up gate code to use mirror_info https://review.openstack.org/551259 | 11:33 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/newton: Clean-up gate code to use mirror_info https://review.openstack.org/551260 | 11:33 |
| *** vnogin has joined #openstack-ansible | 11:38 | |
| xdfil | anyone going to Vancouver? | 11:46 |
| *** vnogin has quit IRC | 11:50 | |
| *** woodard has joined #openstack-ansible | 11:51 | |
| *** vnogin has joined #openstack-ansible | 11:52 | |
| *** woodard has quit IRC | 11:55 | |
| *** SerenaFeng has joined #openstack-ansible | 12:00 | |
| jamespo | Hello, getting this error on pike, trying to open network_topology page from horizon: [Wed Mar 07 15:14:27.334311 2018] [wsgi:error] [pid 4418:tid 140360685680384] UserWarning: Using keystoneclient sessions has been deprecated | 12:18 |
| jamespo | . Please update your software to use keystoneauth1.Recoverable error: Unknown Error (HTTP 504) | 12:18 |
| jamespo | Is this our config or mismatch between component versions? | 12:18 |
| jamespo | ( from infra2_horizon_container-19edb659/horizon/horizon-error.log) | 12:19 |
| *** dave-mccowan has joined #openstack-ansible | 12:20 | |
| ioni | i have a question, i take care of a cloud with 5 regions, i have a primary keystone on the first region in which all other 4 connect to | 12:20 |
| ioni | i want to upgrade from newton to ocata | 12:20 |
| ioni | do you happen to know if keystone is on ocata and the other 4 on newton, it will work? | 12:21 |
| gokhan_ | odyssey4me I added it appropriate group_vars file and it worked :) thanks alot :) | 12:22 |
| *** chyka has joined #openstack-ansible | 12:24 | |
| *** chyka has quit IRC | 12:29 | |
| evrardjp | jamespo: mmm we should be okay on that side, because we are officially using the version provided by requirements | 12:32 |
| evrardjp | jamespo: is that an upgrade, or a fresh install? | 12:32 |
| *** stuartgr has joined #openstack-ansible | 12:32 | |
| sar | Am I correct in assuming that the monitoring user that gets created during the rabbitmq playbook is intended to be used for logging in to the rabbitmq management gui at the :15672 port? | 12:33 |
| evrardjp | ioni: good question | 12:33 |
| jamespo | evrardjp: It is a fresh install | 12:33 |
| evrardjp | sar: I'd say yes on the first part of the thing | 12:33 |
| evrardjp | sar: the GUI not so sure, but at least sending administrative commands | 12:33 |
| evrardjp | jamespo: fresh installs should be fine | 12:34 |
| evrardjp | you should probably file a bug with your current sha version, your user_*.yml variables | 12:34 |
| evrardjp | maybe showing the version of keystoneclient/keystoneauth1 in the venv. | 12:34 |
| evrardjp | jamespo: but that looks like an unconstrained version of keystone | 12:35 |
| sar | Ok. I'm getting these error messages in the rabbitmq log: "HTTP access denied: user 'monitoring' - Not management user". Just wondering if that was what is was intended for | 12:35 |
| evrardjp | mmm | 12:35 |
| evrardjp | let me double check sar | 12:35 |
| jamespo | OK thanks, will try and reproduce a few more times | 12:35 |
| evrardjp | https://github.com/openstack/openstack-ansible/blob/85ce5a0fadc22e81fcd1267afa0be0f462edd953/playbooks/rabbitmq-install.yml#L70-L75 | 12:36 |
| evrardjp | sar ^ | 12:36 |
| *** udesale has joined #openstack-ansible | 12:36 | |
| evrardjp | wondering why you see the errors | 12:36 |
| sar | Yes, i've found that same playbook. But what is the monitoring gui? The one at port 15672, right? | 12:36 |
| evrardjp | how are you monitoring the rabbitmq? | 12:37 |
| sar | I'm just trying to access it using a web browser. It shows a login form, but i get that error when i try to log in | 12:37 |
| evrardjp | oh | 12:39 |
| evrardjp | I see what you mean now | 12:39 |
| evrardjp | that should be the admin user | 12:39 |
| evrardjp | the monitoring is just for monitoring purposes | 12:39 |
| evrardjp | iirc | 12:40 |
| *** SerenaFeng has quit IRC | 12:40 | |
| sar | Ah, ok. Well' there is no admin user. Only the users created by OSA | 12:40 |
| evrardjp | oh no it should do the trick | 12:40 |
| evrardjp | let me double check | 12:41 |
| evrardjp | :) | 12:41 |
| sar | Please. And thank you very much for the help :-) | 12:41 |
| evrardjp | ok, so to access the web interface, we have the plugin that's enabled by default. Good. | 12:42 |
| evrardjp | Then you need an admistrative root user | 12:42 |
| evrardjp | we don't have any other user by default than monitoring | 12:42 |
| *** gkadam_ has joined #openstack-ansible | 12:43 | |
| evrardjp | it was added long ago, maybe before the time we vhosted rabbitmq | 12:44 |
| sar | Ok. I can always add a user myself, or modify the monitoring user. I was just wondering if the monitoring user was meant to be used like this | 12:45 |
| *** gkadam has quit IRC | 12:46 | |
| evrardjp | maybe we should add something like vhost: / | 12:46 |
| evrardjp | configure_priv: .* | 12:47 |
| evrardjp | read_priv: .* | 12:47 |
| evrardjp | I'dd add a separate user | 12:47 |
| evrardjp | keep the monitoring user for other reasons | 12:47 |
| sar | Yes, I agree | 12:47 |
| evrardjp | but it seems our monitoring user, as is, is quite meaning less | 12:47 |
| jamespo | OK, looks like things are trying to connect to external address for keystone which they shouldn't be | 12:47 |
| evrardjp | jamespo: yes but why | 12:47 |
| evrardjp | jamespo: not sure what you're doing and what version of keystoneauth1 is used | 12:48 |
| jamespo | I think it may have been misleading, I can't get that same error now, just 504 errors, which I have followed back to nova log | 12:49 |
| jamespo | which is trying to access external side of haproxy | 12:49 |
| *** woodard has joined #openstack-ansible | 12:49 | |
| evrardjp | that sounds bad too | 12:49 |
| evrardjp | what's your user configuration? | 12:49 |
| evrardjp | there is something wrong on your configuration | 12:50 |
| evrardjp | nova should speak internally | 12:50 |
| jamespo | Yes I think we must have a bit of config wrong somewhere, sorry to bug you | 12:51 |
| *** hw_wutianwei has quit IRC | 12:51 | |
| *** hw_wutianwei has joined #openstack-ansible | 12:52 | |
| evrardjp | jamespo: no worries | 12:52 |
| evrardjp | andymccr: fyi everything is blocked until https://review.openstack.org/#/c/550542/ passes | 12:53 |
| *** woodard has quit IRC | 12:54 | |
| evrardjp | we're gonna move to a different testing, it's too hard to make things pass now. | 12:54 |
| evrardjp | thanks spectre | 12:54 |
| evrardjp | mhayden: if you could have a look at why https://review.openstack.org/#/c/503086/ is failing at centos that would be nice | 12:56 |
| evrardjp | gcc is your friend right? :p | 12:56 |
| *** vnogin_ has joined #openstack-ansible | 12:58 | |
| *** vnogin has quit IRC | 12:58 | |
| *** woodard has joined #openstack-ansible | 13:05 | |
| *** woodard has quit IRC | 13:05 | |
| *** woodard has joined #openstack-ansible | 13:06 | |
| openstackgerrit | Jean-Philippe Evrard proposed openstack/openstack-ansible stable/newton: Clean-up gate code to use mirror_info https://review.openstack.org/551260 | 13:07 |
| *** hamza21 has joined #openstack-ansible | 13:09 | |
| *** ansmith has joined #openstack-ansible | 13:11 | |
| evrardjp | hwoarang: could you make this happen: https://review.openstack.org/#/c/551251/2 ? | 13:11 |
| *** vnogin_ has quit IRC | 13:12 | |
| evrardjp | mhayden: openstask! | 13:12 |
| jamespo | evrardjp: checked and all endpoints look right, think this could be a bug potentially | 13:13 |
| jamespo | almost like auth address is being passed all the way from the browser | 13:13 |
| *** Sha0000 has joined #openstack-ansible | 13:14 | |
| *** vnogin has joined #openstack-ansible | 13:16 | |
| mhayden | morning! | 13:16 |
| *** armaan has joined #openstack-ansible | 13:17 | |
| *** idlemind has quit IRC | 13:19 | |
| *** armaan has quit IRC | 13:19 | |
| *** armaan has joined #openstack-ansible | 13:20 | |
| evrardjp | cloudnull: could you check my comment on https://review.openstack.org/#/c/551032/ ? | 13:20 |
| evrardjp | mhayden: good morning! | 13:20 |
| evrardjp | mhayden: every centos is borked | 13:20 |
| evrardjp | :) | 13:21 |
| evrardjp | that's a nice way to say hello, isn't it? | 13:21 |
| evrardjp | asking infra if it's in images or not | 13:22 |
| *** armaan has quit IRC | 13:24 | |
| jrosser | evrardjp: i think this is another tag related problem | 13:26 |
| jrosser | https://www.irccloud.com/pastebin/XHwwooVa/ | 13:27 |
| *** jsheeren has joined #openstack-ansible | 13:27 | |
| mhayden | oh | 13:28 |
| mhayden | what's wrong? | 13:28 |
| * mhayden is looking | 13:29 | |
| mhayden | egads -- dependency conflict | 13:29 |
| * mhayden builds a test instance to look | 13:30 | |
| mhayden | i wonder if it has something to do with what infra has in their default image | 13:31 |
| *** vnogin has quit IRC | 13:34 | |
| *** SmearedBeard has quit IRC | 13:34 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-tests master: Upgrade CentOS packages before testing https://review.openstack.org/551279 | 13:36 |
| mhayden | testing ^^ | 13:36 |
| evrardjp | jrosser: that or some register missing :) | 13:36 |
| evrardjp | jrosser: we have a tagfilter strategy | 13:37 |
| evrardjp | Maybe we should use it in our testing. | 13:37 |
| *** stuartgr has quit IRC | 13:38 | |
| evrardjp | mhayden: yeah dependencies on gcc :/ | 13:38 |
| mhayden | i'm wondering if the current infra image has a bad package set | 13:39 |
| * mhayden couldn't reproduce on a recent centos cloud image | 13:39 | |
| *** vnogin has joined #openstack-ansible | 13:39 | |
| jrosser | ahhhh right looks like i cant use pike at all :( https://bugs.launchpad.net/nova/+bug/1751349 | 13:41 |
| openstack | Launchpad bug 1751349 in OpenStack Compute (nova) pike "Keystone auth parameters cannot be configured in [keystone] section" [Undecided,Invalid] | 13:42 |
| *** Sha0000 has quit IRC | 13:43 | |
| *** SmearedBeard has joined #openstack-ansible | 13:44 | |
| *** radeks_ has joined #openstack-ansible | 13:44 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-tests master: Upgrade CentOS packages before testing https://review.openstack.org/551279 | 13:44 |
| *** hw_wutianwei has quit IRC | 13:46 | |
| evrardjp | mhayden: yeah I was wondering too | 13:47 |
| evrardjp | jrosser: :( | 13:47 |
| jrosser | evrardjp: yeah thats really pants | 13:48 |
| *** hamza21 has quit IRC | 13:48 | |
| jrosser | we're just wondering if we should hack logan's rejected patch all over the compute nodes, or just burn it and put queens on it | 13:48 |
| jrosser | advice welcome there about the state of queens right now | 13:49 |
| evrardjp | jrosser: gates are borked, but that's because infra has changed variables. I can do my own build very quick to tell you the real state for deployers if you like | 13:49 |
| jrosser | that would be super helpful | 13:50 |
| *** vnogin has quit IRC | 13:50 | |
| jrosser | we'll take a look at bodging the nova code in the meantime | 13:50 |
| *** vnogin has joined #openstack-ansible | 13:50 | |
| *** holser__ has quit IRC | 13:51 | |
| *** holser___ has joined #openstack-ansible | 13:51 | |
| *** hamza21 has joined #openstack-ansible | 13:51 | |
| evrardjp | jrosser: don't forget it's not hard to have your own fork of the code :) | 13:52 |
| evrardjp | that's what it was also built for :D | 13:52 |
| evrardjp | but yeah, using queens is maybe a better thing | 13:52 |
| jrosser | yeah, i know what to do for the osa roles in that respect, but for nova itself? | 13:52 |
| evrardjp | yeah yeah | 13:52 |
| evrardjp | we built the toolkit FOR that | 13:52 |
| evrardjp | https://github.com/openstack/openstack-ansible/blob/stable/pike/playbooks/defaults/repo_packages/openstack_services.yml | 13:53 |
| evrardjp | jrosser: ^ point to whatever your like | 13:53 |
| evrardjp | your fork will be used when repo-build-ing | 13:53 |
| jrosser | excellent | 13:53 |
| evrardjp | but you're on your own after that, we can't support anything that's your doing! :p | 13:54 |
| *** jsheeren has quit IRC | 13:54 | |
| evrardjp | (I mean: editing nova is a can full of ... surprises!) | 13:54 |
| *** vnogin has quit IRC | 13:55 | |
| evrardjp | https://i.imgur.com/Zvet0.gif | 13:59 |
| *** Sha0000 has joined #openstack-ansible | 14:01 | |
| *** gkadam__ has joined #openstack-ansible | 14:01 | |
| *** vnogin has joined #openstack-ansible | 14:03 | |
| jamespo | Thanks evrardjp, just forking it now (nova) and patching | 14:03 |
| *** gkadam_ has quit IRC | 14:03 | |
| *** hw_wutianwei has joined #openstack-ansible | 14:05 | |
| evrardjp | jamespo: ? | 14:21 |
| evrardjp | you should try to avoid that :p | 14:21 |
| jamespo | https://review.openstack.org/#/c/547654/ + https://github.com/openstack/openstack-ansible/blob/stable/pike/playbooks/defaults/repo_packages/openstack_services.yml | 14:23 |
| *** throwsb1 has joined #openstack-ansible | 14:25 | |
| evrardjp | jamespo: so you're also victim of that bug? I thought that was only for specific use cases | 14:26 |
| jamespo | It is the cause of what I was asking earlier, I worth with jrosser | 14:29 |
| jamespo | we need everything to talk over internal networks, only end users should be going to the public endpoint | 14:29 |
| jamespo | And I think that's a sensible use case really, not sure why anything would try to go to the external/public endpoint from inside | 14:30 |
| jamespo | work with * | 14:32 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-tests master: Allow yum to skip broken package dependencies https://review.openstack.org/551279 | 14:34 |
| *** sawblade6 has quit IRC | 14:35 | |
| *** sawblade6 has joined #openstack-ansible | 14:35 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_zaqar master: Follow the new PTI for document build https://review.openstack.org/551129 | 14:39 |
| evrardjp | jamespo: yes that's definitely a sensible use case | 14:39 |
| evrardjp | but I am not sure why this started to happen | 14:39 |
| evrardjp | jamespo: I'd use Queens. | 14:40 |
| jamespo | It looks like it was always happening and just that nobody found the problem, we might move to queens but very close with this deployment so was hoping to get the last few things working | 14:41 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-tests master: Allow yum to skip broken package dependencies https://review.openstack.org/551279 | 14:46 |
| evrardjp | jamespo: you mean you have the same kind of overrides as logan, right? Because I am concerned gates didn't catch that | 14:51 |
| evrardjp | probably worth doing one more scenario for it | 14:51 |
| jamespo | I'm not sure exactly what Logan was up to, but I assumed/hoped that everything would talk internally | 14:58 |
| jamespo | from a security point of view at least | 14:58 |
| jamespo | Seems like it would be sensible to always talk within one cloud/region, then if something needs to talk cross region then use the external endpoint for that | 15:00 |
| evrardjp | yes, taht's what's by default: Always talk within one region | 15:04 |
| *** kstev has joined #openstack-ansible | 15:06 | |
| jamespo | Apart from nova :) | 15:11 |
| *** Sha0000 has quit IRC | 15:11 | |
| jamespo | or when I say within the region, I meant internally too | 15:11 |
| *** epalper has quit IRC | 15:19 | |
| *** osn1 has joined #openstack-ansible | 15:19 | |
| *** rickflare2 has quit IRC | 15:22 | |
| jamespo | evrardjp: I added "nova_git_repo" to a user_variables.yml file but it won't take, do I need to do something else to make it load ing? | 15:22 |
| jamespo | I assume this role is setting up the repo container with the right version https://github.com/openstack/openstack-ansible-repo_build | 15:23 |
| *** SerenaFeng has joined #openstack-ansible | 15:23 | |
| *** vnogin has quit IRC | 15:27 | |
| openstackgerrit | Merged openstack/openstack-ansible stable/queens: Clean-up gate code to use mirror_info https://review.openstack.org/551251 | 15:28 |
| *** vnogin has joined #openstack-ansible | 15:28 | |
| *** hamza21 has quit IRC | 15:31 | |
| *** SerenaFeng has quit IRC | 15:32 | |
| *** vnogin has quit IRC | 15:33 | |
| *** Sha000000 has joined #openstack-ansible | 15:33 | |
| jamespo | sorry, it does look like it's working but it still pulled in the old/normal nova repo too | 15:34 |
| openstackgerrit | Amy Marrich (spotz) proposed openstack/openstack-ansible master: [WIP][DOC] Update for Newton EOL https://review.openstack.org/551316 | 15:34 |
| *** vnogin has joined #openstack-ansible | 15:34 | |
| spotz | Prep for when Newton goes EOL | 15:34 |
| *** esberglu has joined #openstack-ansible | 15:37 | |
| *** vnogin has quit IRC | 15:39 | |
| *** udesale has quit IRC | 15:41 | |
| evrardjp | thanks spotz | 15:49 |
| spotz | evrardjp: Knowing you you'd tag it while I was asleep or riding:) | 15:50 |
| evrardjp | haha | 15:50 |
| evrardjp | For now things are broken, so it's not yet ready | 15:50 |
| spotz | I'll blame mhayden cause I can't for much longer:) | 15:51 |
| evrardjp | oh :( | 15:52 |
| evrardjp | yeah | 15:52 |
| evrardjp | if I could get some votes on https://review.openstack.org/#/c/551260/3 that would be great | 15:52 |
| spotz | I'm working on it.. mhayden and review:) | 15:52 |
| evrardjp | spotz: centos is broken, and because he brought all of that to us, it's kinda his fault? | 15:52 |
| evrardjp | :) | 15:52 |
| spotz | evrardjp: It's just always his fault!:) | 15:52 |
| evrardjp | What will we do when we can't blame him anymore? We won't progress as fast too! | 15:53 |
| evrardjp | doom and despair! | 15:53 |
| spotz | We'll still blame him, he just may not be logged in to see | 15:53 |
| spotz | odyssey4me evrardjp - fixing the releasenote | 15:55 |
| spotz | actually its mnaser and it's the commit message:) | 15:56 |
| spotz | reading is hard yo | 15:56 |
| mnaser | whatd i do | 15:56 |
| osn1 | @cloudnull @evrardjp here is the openstack_user_config.yml for your review.... (deployment on 1 controller, 2comp) - No containers, No storage host...http://paste.openstack.org/show/696392/ | 15:56 |
| spotz | mnaser: Typo in the commit, I've fixed it | 15:56 |
| mnaser | oh ok cool :D | 15:57 |
| mnaser | thank you, tho most of that patch was odyssey4me | 15:57 |
| spotz | Usually I ask first if it's someone I've never fixed before but for some reason I thought it was Jesse's patch | 15:57 |
| openstackgerrit | Amy Marrich (spotz) proposed openstack/openstack-ansible stable/newton: Clean-up gate code to use mirror_info https://review.openstack.org/551260 | 15:57 |
| evrardjp | mnaser: you fixed stuff with odyssey4me :) | 15:57 |
| evrardjp | what... it was almost ready! | 15:57 |
| mnaser | oh yay | 15:58 |
| mnaser | horizon running out of open files | 15:58 |
| evrardjp | yeah | 15:58 |
| evrardjp | love those kinds | 15:58 |
| SamYaple | openstackgerrit: did you get a chance to look at the ekystone bindep patch? https://review.openstack.org/#/c/549223/ | 15:58 |
| mnaser | OperationalError: (2004, 'Can\\'t create TCP/IP socket (24 "Too many open files")') | 15:58 |
| SamYaple | ooops odyssey4me ^ | 15:58 |
| mnaser | would we be okay with proposing a patch upstream to fix that isuse | 15:58 |
| spotz | evrardjp andymccr - one of you re +2 and workflow please | 15:58 |
| mnaser | by fix, it's increase the limit of open files because i am not about to dig into horizons code | 15:58 |
| andymccr | got it! | 15:59 |
| *** stuartgr has joined #openstack-ansible | 15:59 | |
| spotz | andymccr: thanks | 16:00 |
| osn1 | @cloudnull @evrardjp when commented storage section from the config, it errors out on --syntax-check .... please advise (there is no separate storage in my setup), plan to use just local storage | 16:01 |
| *** pcaruana has quit IRC | 16:02 | |
| *** kstev1 has joined #openstack-ansible | 16:03 | |
| *** kstev has quit IRC | 16:04 | |
| mnaser | may i ask the decision behind using mysql with horizon for session rather than memcache? | 16:05 |
| SamYaple | mnaser: persistenance | 16:06 |
| SamYaple | restarting memcache could lose someones active session, backed by mysql tahts not the case | 16:06 |
| SamYaple | its actually backed by mysql, cached by memcache | 16:07 |
| osn1 | @cloudnull @evrardjp will join back around 9am pst .. will ping you then | 16:07 |
| *** osn1 has quit IRC | 16:07 | |
| *** jamesdenton has quit IRC | 16:08 | |
| *** vnogin has joined #openstack-ansible | 16:10 | |
| *** stuartgr has quit IRC | 16:10 | |
| cloudnull | mornings | 16:13 |
| *** ansmith has quit IRC | 16:16 | |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible-tests master: Allow yum to skip broken package dependencies https://review.openstack.org/551279 | 16:16 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_glance master: Convert role to use a common systemd init role https://review.openstack.org/551053 | 16:17 |
| openstackgerrit | Merged openstack/openstack-ansible-ops master: Follow the new PTI for document build https://review.openstack.org/551151 | 16:18 |
| *** shardy has quit IRC | 16:19 | |
| *** radeks_ has quit IRC | 16:20 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_molteniron master: Follow the new PTI for document build https://review.openstack.org/551122 | 16:21 |
| *** Sha000000 has quit IRC | 16:27 | |
| *** holser___ is now known as holser__ | 16:27 | |
| mnaser | is there a way to 'omit' config values in OSA right now (barring sending a patch which has an if foo is set) | 16:30 |
| mnaser | example: single az setup where the az name is not 'nova'. default_schedule_zone=nova is configured by default, i rather have it omitted (defaulting to none) | 16:31 |
| openstackgerrit | Merged openstack/openstack-ansible-os_almanach master: Follow the new PTI for document build https://review.openstack.org/551141 | 16:32 |
| cloudnull | mnaser: we use config_template to add config but dont have the same to remove default config. | 16:33 |
| cloudnull | we could effectively remove that option, or add a capability to pull options out using config_template ? | 16:34 |
| *** ansmith has joined #openstack-ansible | 16:34 | |
| openstackgerrit | Merged openstack/openstack-ansible-apt_package_pinning master: Follow the new PTI for document build https://review.openstack.org/551057 | 16:34 |
| cloudnull | but that's not something we have today | 16:34 |
| openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-os_nova master: Omit default_schedule_zone if its not configured https://review.openstack.org/551324 | 16:34 |
| mnaser | cloudnull: yeah that seems like a fairly difficult thing... anyways, ^ | 16:34 |
| logan- | i think you can do it with config_template | 16:36 |
| mnaser | oh? that'd be nice | 16:36 |
| openstackgerrit | Merged openstack/openstack-ansible-os_watcher master: Follow the new PTI for document build https://review.openstack.org/551137 | 16:36 |
| *** hamza21 has joined #openstack-ansible | 16:38 | |
| logan- | default_schedule_zone: '' | 16:38 |
| cloudnull | oh that might work for that use case | 16:39 |
| mnaser | will that result in 'default_schedule_zone=' or simply it would omit it | 16:39 |
| logan- | the former | 16:39 |
| cloudnull | maybe we could use a special word to remove config options, something like omit? | 16:40 |
| openstackgerrit | Merged openstack/openstack-ansible-os_rally master: Follow the new PTI for document build https://review.openstack.org/551124 | 16:40 |
| mnaser | logan-: well in that case, i can just set the ansible variable to '' and that might do the same thing, but im not sure if oslo.config considers that as none/default value | 16:40 |
| openstackgerrit | Merged openstack/openstack-ansible-os_ceilometer master: Follow the new PTI for document build https://review.openstack.org/551126 | 16:42 |
| *** gkadam__ has quit IRC | 16:42 | |
| mhayden | so i think the repo for centos in infra is in bad shape | 16:44 |
| mhayden | i can't figure out a workaround unless we abandon openstack repos | 16:44 |
| openstackgerrit | Merged openstack/openstack-ansible-os_cloudkitty master: Follow the new PTI for document build https://review.openstack.org/551164 | 16:45 |
| *** chyka has joined #openstack-ansible | 16:47 | |
| odyssey4me | mnaser if a default we set is not required, then generally we should be removing it - config_template allows anything to be set... we didn't always have that facility, so there are a lot of vars which could be removed as they have nothing to do with the 'glue', but instead just pre-date the config_template facility | 16:48 |
| *** masber has quit IRC | 16:54 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_ironic master: Follow the new PTI for document build https://review.openstack.org/551161 | 17:00 |
| *** Faster-Fanboi has quit IRC | 17:00 | |
| mnaser | odyssey4me: i think so too | 17:02 |
| mnaser | also, anyone ever seen this? | 17:02 |
| mnaser | [Fri Mar 9 16:57:35 2018] POST /v3//auth/tokens => generated 11881 bytes in 144442 msecs (HTTP/1.1 201) 6 headers in 386 bytes (1 switches on core 0) | 17:02 |
| openstackgerrit | Merged openstack/openstack-ansible-os_swift master: Follow the new PTI for document build https://review.openstack.org/551132 | 17:05 |
| *** SerenaFeng has joined #openstack-ansible | 17:12 | |
| *** ivve has quit IRC | 17:14 | |
| openstackgerrit | Merged openstack/openstack-ansible stable/pike: Clean-up gate code to use mirror_info https://review.openstack.org/551258 | 17:17 |
| *** mbuil has quit IRC | 17:18 | |
| *** osnaya has joined #openstack-ansible | 17:20 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_octavia master: Follow the new PTI for document build https://review.openstack.org/551133 | 17:21 |
| *** osnaya has quit IRC | 17:21 | |
| *** electrofelix has quit IRC | 17:22 | |
| *** osnaya has joined #openstack-ansible | 17:23 | |
| osnaya | @cloudnull Morning.... | 17:23 |
| osnaya | @cloudnull Any ideas on the error on storage_hosts during inventory generation.... from the user_config.yml from this morning? | 17:24 |
| osnaya | @cloudnull per your suggestion from yesterday, it is there on paste | 17:25 |
| *** vnogin has quit IRC | 17:26 | |
| *** hachi has joined #openstack-ansible | 17:26 | |
| *** holser__ has quit IRC | 17:28 | |
| odyssey4me | cloudnull not sure if you have a gap to do some analysis, but it would seem that somehow we've gained around 30 mins on the integrated gate for master somehow... queens seems to be holding a similar time, but master has gotten quite a lot worse | 17:28 |
| *** Leo_m_ has quit IRC | 17:28 | |
| *** radeks_ has joined #openstack-ansible | 17:28 | |
| *** hamza21 has quit IRC | 17:30 | |
| odyssey4me | It's plausible that in some of the master changes for the connection plugins, or the networkd configs that we've lost time. It might be due to a retry timeout being quite long, or some other reason, but we need to try and isolate it as it's killing our ability to merge things on the integrated gate. | 17:30 |
| odyssey4me | We can work around it by switching to the basekit scenario, but then we lose coverage. | 17:31 |
| openstackgerrit | Merged openstack/openstack-ansible-os_magnum master: Follow the new PTI for document build https://review.openstack.org/551123 | 17:32 |
| openstackgerrit | Merged openstack/openstack-ansible-galera_client stable/pike: Allow Galera package downloads over HTTP https://review.openstack.org/550857 | 17:32 |
| *** Leo_m has joined #openstack-ansible | 17:33 | |
| openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-plugins master: Check for keys ending with git_install_branch https://review.openstack.org/550640 | 17:34 |
| cloudnull | odyssey4me: i recently put this pr up which I hope should resolve some of the connection plugin issues https://review.openstack.org/#/c/550331/ | 17:37 |
| cloudnull | however if we keep seeing the libc_fork issues I'll have to revert the lxc-attach changes | 17:37 |
| cloudnull | I know that issue is causing us problems and may be elongating the gate | 17:38 |
| odyssey4me | cloudnull interestingly, I'venot seen that issue in the integrated gate - only role tests | 17:38 |
| *** Leo_m has quit IRC | 17:38 | |
| cloudnull | in the integrated gate it would result in a connection retry | 17:39 |
| cloudnull | which has a cool down, and we have SSH retry enabled | 17:39 |
| cloudnull | I | 17:39 |
| cloudnull | I've also noticed in master that the repo build is taking considerabily longer. | 17:40 |
| odyssey4me | hmm, interesting | 17:40 |
| odyssey4me | the only new change in https://review.openstack.org/550542 is to make it use the infra wheel mirror, which should make it *faster* not *slower* | 17:40 |
| odyssey4me | I wonder if perhaps that pypi reverse proxy needs some way of handling retries too | 17:41 |
| cloudnull | the repo build length may just be due to the wheels not being in the openstack mirrors | 17:42 |
| cloudnull | so it could just because the development cycle is opening up | 17:42 |
| odyssey4me | well, perhaps adding the extra URL is causing a slow down due to the extra URL query being implemented | 17:43 |
| odyssey4me | instead of just checking pypi, it's now checking two URLs | 17:43 |
| cloudnull | that could be | 17:43 |
| odyssey4me | queens seems to have worked though - not sure if the same slow down is there in the same process | 17:44 |
| odyssey4me | looks like the infra wheel addition is actually helping though: http://logs.openstack.org/42/550542/12/check/openstack-ansible-deploy-aio_lxc-ubuntu-xenial/fc46bd1/logs/openstack/aio1_repo_container-97b04d7a/repo/wheel_build.log.txt.gz | 17:45 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance master: [WIP] Use a common python build role for source builds https://review.openstack.org/551344 | 17:56 |
| cloudnull | Im spinning a master integrated AIO. | 17:56 |
| cloudnull | something is just not quite right | 17:56 |
| odyssey4me | cloudnull thanks - I'm feeling quite rotten... I guess Dublin's snow has caught up to me and I'm now getting worked over by jmccrory's 'flu. :/ | 17:57 |
| cloudnull | ha! | 17:57 |
| cloudnull | go feel better | 17:57 |
| cloudnull | the jmccrory flu aint nothing to mess with | 17:58 |
| jmccrory | oh god sorry! hope you get better soon :/ | 17:58 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_glance master: [WIP] Use a common python build role for source builds https://review.openstack.org/551344 | 17:59 |
| cloudnull | nice! ^ | 18:01 |
| odyssey4me | :) I guess I just need some rest and recuperation. Such is life. | 18:01 |
| *** Bico_Fino has joined #openstack-ansible | 18:01 | |
| odyssey4me | ^ that patch is a start for using a common role for venv building and the source-based builds... there's a way to go, but it's a start. | 18:02 |
| odyssey4me | It'll help with nmaser's quest for binary builds to separate things out. | 18:02 |
| *** vnogin has joined #openstack-ansible | 18:04 | |
| *** aedan9902 has joined #openstack-ansible | 18:08 | |
| odyssey4me | heh, that should have been mnaser :p | 18:12 |
| * odyssey4me goes to !computer | 18:12 | |
| aedan9902 | Hello. I am attempting to install openstack and place the ssl certificate on the keystone container. Instead of placing it on the haproxy lb. I still need to use the haproxy lb, but don't want to have unencrypted communication between the 2. Online documentation says this is possible, but does not tell how to do it. I have tried to do this for about 3 weeks now. I have spoken with the PTL and he cannot find documentation | 18:13 |
| aedan9902 | either. He believes as do I that it may not work and that I will need to submit a bug reporting that it does not work. However, before doing so. I wanted to know if anyone here may have gotten it to work. If so, could you provide documentation. | 18:13 |
| *** vnogin has quit IRC | 18:14 | |
| *** dmsimard is now known as dmsimard|afk | 18:25 | |
| *** jbadiapa has quit IRC | 18:27 | |
| osnaya | @cloudnull ... did you get a chance to look at the config for (No container, No storage) scenario? Please advise | 18:31 |
| openstackgerrit | Merged openstack/openstack-ansible-galera_server stable/pike: Allow Galera package downloads over HTTP https://review.openstack.org/550856 | 18:40 |
| *** Leo_m has joined #openstack-ansible | 18:49 | |
| jrosser | evrardjp: we convinced nova (via logans patch) to use the internal rather than external endpoint. | 18:54 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-os_glance master: Convert role to use a common systemd init role https://review.openstack.org/551053 | 18:55 |
| jrosser | you mentioned adding a scenario to test this, somehow blackholing the external vip in the right places would do it, but by the looks of things pike and earlier suffers from this | 18:56 |
| cloudnull | osnaya: sorry, i dont have the config, mind sharing it again ? | 18:56 |
| logan- | jrosser: funny you ran into that so soon after I did | 18:56 |
| jrosser | logan-: this week weve properly isolated the mgmt network | 18:57 |
| cloudnull | anyone around mind giving https://review.openstack.org/#/c/550161/ a push ? | 18:57 |
| jrosser | i think there may have been a crafty route round in our prevoius build | 18:57 |
| osnaya | @cloudnull please check http://paste.openstack.org/show/696392 | 19:00 |
| *** chhavi__ has quit IRC | 19:02 | |
| *** chhagarw has quit IRC | 19:02 | |
| cloudnull | osnaya: and that's on pike ? | 19:02 |
| cloudnull | osnaya: http://paste.openstack.org/show/696772/ | 19:06 |
| cloudnull | so there were two issues | 19:06 |
| cloudnull | one the cidr_networks and used_ips sections had bad indenting | 19:06 |
| cloudnull | and the host under log hosts was named log1 instead of infra1 | 19:07 |
| cloudnull | these were the errors http://paste.openstack.org/show/696780/ | 19:08 |
| cloudnull | the yaml parsing error was just due to the indentation | 19:08 |
| cloudnull | and the host name error was due to the fact there were two hosts listed with the same IP, which the inventory script tries to protect against | 19:09 |
| *** Leo_m_ has joined #openstack-ansible | 19:09 | |
| *** Leo_m has quit IRC | 19:09 | |
| cloudnull | so osnaya with the revised user config file which i pasted before you should be on the right track | 19:09 |
| *** radeks_ has quit IRC | 19:10 | |
| *** radeks has joined #openstack-ansible | 19:12 | |
| osnaya | @cloudnull will check it out | 19:17 |
| *** flemingo has joined #openstack-ansible | 19:36 | |
| *** flemingo_ has joined #openstack-ansible | 19:38 | |
| *** eumel8_ has joined #openstack-ansible | 19:39 | |
| *** flemingo has quit IRC | 19:40 | |
| *** eumel8 has quit IRC | 19:42 | |
| spotz | aedan9902: I could have sworn there's a section on SSL certs. Give me a few to see if it's still about | 19:48 |
| spotz | aedan9902: The page I was thinking of is user/security/ssl-certificates. There's a section with a rabbitmq example which might be worth adapting and trying | 19:51 |
| *** throwsb1 has quit IRC | 19:52 | |
| aedan9902 | Hi Spotz. I have tried that. It places the certificates in the container, but does not enable them. Basically, if that section is the documentation. It is incomplete at best. | 19:53 |
| spotz | aedan9902: That was the only hit I got when grepping the main repo docs | 19:54 |
| aedan9902 | Yeah. Me too. Basically the what seems to get me close is doing this: | 19:55 |
| aedan9902 | keystone_ssl: True | 19:55 |
| aedan9902 | ssl_protocol: 'SSLv2 SSLv3' | 19:55 |
| aedan9902 | keystone_service_proto: https | 19:55 |
| aedan9902 | keystone_user_ssl_cert: /etc/letsencrypt/live/jake-aio.jakelab.info/cert.pem | 19:55 |
| aedan9902 | keystone_user_ssl_ca_cert: /etc/letsencrypt/live/jake-aio.jakelab.info/chain.pem | 19:55 |
| aedan9902 | keystone_user_ssl_key: /etc/letsencrypt/live/jake-aio.jakelab.info/privkey.pem | 19:55 |
| aedan9902 | Then changing the mode for keystone in the group_vars/all/haproxy.yml file to tcp | 19:56 |
| spotz | But that doesn't actually implement it? | 19:56 |
| osnaya | @cloudnull thanks... the 2 errors are gone now... but the original error surfaced again for storage hosts (in NO separate storage host case) - File "/opt/openstack-ansible/osa_toolkit/generate.py", line 650, in _add_additional_networks if user_config[cphg].get(container_host) is not None: KeyError: u'storage_hosts' | 19:56 |
| aedan9902 | no | 19:57 |
| spotz | you there | 19:58 |
| spotz | bah lost half of that | 19:59 |
| spotz | aedan9902: I'd say bug it and include the steps that got you there | 19:59 |
| aedan9902 | ok. Will do. Thanks | 19:59 |
| *** ansmith has quit IRC | 20:00 | |
| osnaya | @cloudnull http://paste.openstack.org/show/696867/ with above keyError: u'storage_hosts | 20:01 |
| spotz | aedan9902: My concern there is keystone_ssl: true but it didn't use it | 20:01 |
| aedan9902 | Well. With all that enabled. It updates nginx and what not. It looks like it may be configured(well close to configured) However, openrc does not get updated correctly and glance is not able to connect to keystone once the changes are implemented. That is where the playbooks fail. I assume all other services would also fail. | 20:04 |
| openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-os_nova master: Omit default_schedule_zone if its not configured https://review.openstack.org/551324 | 20:05 |
| cloudnull | osnaya: testing that config I'm not seeing the same thign | 20:05 |
| cloudnull | **thing | 20:05 |
| cloudnull | running ./inventory/dynamic_inventory.py works without issue | 20:06 |
| cloudnull | can you make sure you've not got some extra config getting in the way under /etc/openstack_deploy/{env.d,conf.d} ? | 20:07 |
| cloudnull | these are the files I have when testing | 20:08 |
| cloudnull | http://paste.openstack.org/show/696888/ | 20:08 |
| mnaser | could i borrow eyes on https://review.openstack.org/#/c/550189/ :) | 20:09 |
| cloudnull | mnaser: lgtm | 20:10 |
| *** vnogin has joined #openstack-ansible | 20:19 | |
| *** threestrands has joined #openstack-ansible | 20:22 | |
| *** threestrands has quit IRC | 20:22 | |
| *** threestrands has joined #openstack-ansible | 20:22 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/openstack-ansible-plugins master: Resolve nsenter fork exception https://review.openstack.org/550331 | 20:22 |
| cloudnull | odyssey4me: logan-: This should help the general gate ^ | 20:22 |
| cloudnull | restored the lxc-attach command when working with lxc containers | 20:22 |
| cloudnull | it seems they've worked around the libc issues internally | 20:22 |
| cloudnull | while lxc-attach will hang from time to time it wont crash which is better. | 20:23 |
| cloudnull | in the future we can circle back on removing that command again once all supported distros are using libc 2.25 + | 20:23 |
| logan- | :( | 20:25 |
| cloudnull | I was seeing the issue even with the retry | 20:26 |
| cloudnull | the no-fork option | 20:27 |
| *** vnogin has quit IRC | 20:27 | |
| cloudnull | and several kernel tunings. lxc-attach has some code within it that works around the problem | 20:27 |
| cloudnull | so back to lxc-attach for now | 20:27 |
| *** woodard_ has joined #openstack-ansible | 20:30 | |
| *** woodard has quit IRC | 20:30 | |
| *** hw_wutianwei has quit IRC | 20:31 | |
| osnaya | @cloudnull I do have some entries in env.d and conf.d folders (for deploying on Ubuntu, we need the file aio_metal.yml in in env.d, right? | 20:32 |
| *** hw_wutianwei has joined #openstack-ansible | 20:32 | |
| osnaya | @cloudnull http://paste.openstack.org/show/696918/ | 20:32 |
| cloudnull | osnaya: yes you need that file to deploy without containers | 20:41 |
| cloudnull | let me try adding that file and see where it goes | 20:42 |
| cloudnull | os http://paste.openstack.org/show/696944/ | 20:45 |
| cloudnull | still works with that all is metal file too | 20:45 |
| *** woodard_ has quit IRC | 20:45 | |
| cloudnull | im testing by executing `./inventory/dynamic_inventory.py` | 20:45 |
| *** woodard has joined #openstack-ansible | 20:46 | |
| *** hachi has quit IRC | 20:47 | |
| osnaya | @cloudnull ... will try and dig in more, will let u know | 20:52 |
| osnaya | @quit | 20:53 |
| *** osnaya has quit IRC | 20:54 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible master: [TEST] Update Ansible to 2.5.0rc2 https://review.openstack.org/501814 | 21:10 |
| *** radeks has quit IRC | 21:10 | |
| *** SerenaFeng has quit IRC | 21:13 | |
| jmccrory | logan- not sure if https://review.openstack.org/#/c/550665/ breaks assumptions in your environments, shouldn't effect anything less than ansible 2.5 though | 21:20 |
| *** threestrands has quit IRC | 21:22 | |
| *** eumel8_ has quit IRC | 21:24 | |
| *** vnogin has joined #openstack-ansible | 21:26 | |
| *** kstev1 has quit IRC | 21:27 | |
| *** sep has quit IRC | 21:36 | |
| cloudnull | osnaya did you make it go ? | 21:36 |
| *** sep has joined #openstack-ansible | 21:37 | |
| *** armaan has joined #openstack-ansible | 21:37 | |
| *** armaan has quit IRC | 21:38 | |
| *** armaan has joined #openstack-ansible | 21:38 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova master: Added vendor_data.json support https://review.openstack.org/550189 | 21:43 |
| *** armaan has quit IRC | 21:43 | |
| *** sar has quit IRC | 21:46 | |
| *** evin has joined #openstack-ansible | 21:46 | |
| *** hamza21 has joined #openstack-ansible | 21:50 | |
| *** evin has quit IRC | 21:51 | |
| *** sar has joined #openstack-ansible | 21:51 | |
| *** mburrows has joined #openstack-ansible | 21:52 | |
| *** kstev has joined #openstack-ansible | 21:52 | |
| *** sar has quit IRC | 21:52 | |
| *** sar has joined #openstack-ansible | 21:53 | |
| *** kstev1 has joined #openstack-ansible | 21:55 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_gnocchi master: Follow the new PTI for document build https://review.openstack.org/551147 | 21:56 |
| *** sar has quit IRC | 21:56 | |
| *** evin has joined #openstack-ansible | 21:56 | |
| *** kstev has quit IRC | 21:56 | |
| *** Leo_m_ has quit IRC | 21:57 | |
| *** Leo_m has joined #openstack-ansible | 21:57 | |
| openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-os_nova stable/queens: Added vendor_data.json support https://review.openstack.org/551396 | 22:03 |
| openstackgerrit | Mohammed Naser proposed openstack/openstack-ansible-os_nova stable/pike: Added vendor_data.json support https://review.openstack.org/551397 | 22:03 |
| mnaser | i usually stop backporting up until where i need things | 22:03 |
| mnaser | do we usually try to just backport up until oldest | 22:03 |
| mnaser | or is it really 'as needed' | 22:03 |
| openstackgerrit | Merged openstack/openstack-ansible-os_aodh master: Follow the new PTI for document build https://review.openstack.org/551130 | 22:10 |
| *** hamza21 has quit IRC | 22:12 | |
| *** osnaya has joined #openstack-ansible | 22:27 | |
| *** jamespage has quit IRC | 22:27 | |
| *** AndyWojo has quit IRC | 22:28 | |
| *** neith has quit IRC | 22:28 | |
| osnaya | @cloudnull ... tested running ./inventory/dynamic_inventory.py resulting same error " File "/opt/openstack-ansible/osa_toolkit/generate.py", line 650, in _add_additional_networks if user_config[cphg].get(container_host) is not None: KeyError: u'storage_hosts' | 22:28 |
| *** jamespage has joined #openstack-ansible | 22:28 | |
| *** AndyWojo has joined #openstack-ansible | 22:28 | |
| osnaya | @cloudnull you mentioned about env.d and conf.d earlier so looking closely into that... this is what I have | 22:29 |
| *** neith has joined #openstack-ansible | 22:29 | |
| osnaya | @cloudnull please check http://paste.openstack.org/show/697091/ on entries from env.d and conf.d | 22:36 |
| osnaya | @cloudnull error: File "/opt/openstack-ansible/osa_toolkit/generate.py", line 650, in _add_additional_networks if user_config[cphg].get(container_host) is not None: KeyError: u'storage_hosts' | 22:39 |
| cloudnull | mnaser: I generally do the same. | 22:39 |
| cloudnull | backport to the place where I need it. | 22:40 |
| cloudnull | if folks ask for more / further backports I normally am happy to do it so long as it's not a total pain | 22:40 |
| *** Leo_m has quit IRC | 22:40 | |
| *** Leo_m has joined #openstack-ansible | 22:40 | |
| cloudnull | it looks like storage_hosts is defined somewhere in config | 22:42 |
| cloudnull | or in the existing openstack_inventory.json file | 22:42 |
| cloudnull | if this is a new deployment you might need to just delete the inventory file and have it regenerated | 22:43 |
| cloudnull | but it's hard to say, something is seeing storage_hosts | 22:43 |
| mnaser | FAILED! => {"failed": true, "msg": "the field 'args' has an invalid value, which appears to include a variable that is undefined. The error was: 'dict object' has no attribute 'ansible_pkg_mgr'\n\nThe error appears to have been in '/opt/openstack-ansible/playbooks/common-tasks/os-lxc-container-setup.yml': line 98, column 3, but may\nbe elsewhere in the file depending on the exact syntax problem.\n\nThe | 22:44 |
| mnaser | offending line appears to be:\n\n# default on openSUSE.\n- name: Lxc container restart\n ^ here\n"} | 22:44 |
| mnaser | this has to be a bug | 22:44 |
| mnaser | how has no one ran into this? | 22:44 |
| spotz | mnaser: I run in to things no one else has all the time, the question is can you repeat it on different days and machines. That's my deciding factor | 22:44 |
| mnaser | but this happens eeeverytime | 22:44 |
| mnaser | and even looking at it | 22:45 |
| mnaser | https://github.com/openstack/openstack-ansible/commit/a955c0b5070d06933af1fe679390ab4b04662032 | 22:46 |
| mnaser | hostvars[physical_host]['ansible_pkg_mgr'] fails because we never collected the facts for the physical host! | 22:47 |
| spotz | mnaser: Which release? | 22:52 |
| mnaser | stable/pike | 22:52 |
| mnaser | i am doing some wild things like manually running the os nova install playbooks | 22:53 |
| mnaser | openstack-ansible common-playbooks/nova.yml -e'nova_hosts=nova_scheduler:nova_consoleauth:!nova_conductor' -enova_serial="['1','100%']" | 22:53 |
| mnaser | nothing magical tho | 22:53 |
| *** esberglu has quit IRC | 22:53 | |
| cloudnull | mnaser: fact collection would happen in the os-nova-install.yml playbook | 22:53 |
| mnaser | https://github.com/openstack/openstack-ansible/blob/master/playbooks/os-nova-install.yml | 22:54 |
| cloudnull | however you can force it with `ansible -m setup nova_all` | 22:54 |
| mnaser | but it doesnt seem like it does | 22:54 |
| osnaya | @cloudnull so I delete /etc/openstack_deploy/openstack_inventory.son file and re-run, it will generate i fresh again, right? | 22:54 |
| cloudnull | osnaya: yes | 22:54 |
| mnaser | hosts: nova_conductor stuff has gather_facts: no | 22:54 |
| mnaser | and then after that its just including a playbook (which i ddo manually) | 22:54 |
| cloudnull | interesting... that's a bummer, fact collection used to happen in all of the top level playbooks but i guess that was refactored. | 22:55 |
| cloudnull | my bad... | 22:55 |
| spotz | mnaser: I'm about to do a pike AIO but nothing fancy | 22:57 |
| mnaser | spotz: i think this is an issue if you manually run the os-<service>-install.yml | 22:57 |
| mnaser | i guess if you use setup-openstack.yml .. some other role would collcet that fact at some point | 22:57 |
| cloudnull | it would, but that without smart fact gathering that could lead to a poor experience | 22:58 |
| osnaya | @cloudnull wow.... deleting inv.json and re-run - now it run successfully.... thanks for the pointer, something for me to keep in mind.... | 22:58 |
| cloudnull | normally you wont want to delete that file as it will contain the inventory and the listings for all of yoour containers | 22:59 |
| cloudnull | but in your case there will be no containers so it really doesn't matter | 23:00 |
| *** ThomasS has quit IRC | 23:03 | |
| *** dave-mccowan has quit IRC | 23:09 | |
| mnaser | ok uh | 23:11 |
| osnaya | @cloudnull ... luckily I just moved it to another place as .old for reference, but with your comment above poses a question for me... later when I use containers, after re-run wouldn't it create the inv file againa? | 23:11 |
| mnaser | https://github.com/openstack/openstack-ansible/blob/master/playbooks/os-nova-install.yml#L89-L93 | 23:11 |
| mnaser | note nova_hosts = "nova_scheduler:nova_consoleauth:!nova_conductor" | 23:12 |
| mnaser | yet | 23:12 |
| mnaser | https://github.com/openstack/openstack-ansible/blob/master/inventory/env.d/nova.yml | 23:12 |
| mnaser | it says... nova_console there? | 23:12 |
| cloudnull | osnaya: it would amend the inventory file with the containers | 23:12 |
| cloudnull | however if something were to go wrong theres a tar ball for the old inventories | 23:13 |
| cloudnull | mnaser: its just filtering the available hosts in stages | 23:13 |
| *** flemingo_ has quit IRC | 23:13 | |
| mnaser | cloudnull: but the filter is a typo? | 23:14 |
| mnaser | nova_consoleauth but the group is nova_console? | 23:14 |
| osnaya | @cloudnull are you referring to backup_openstack_inventory.tar? | 23:14 |
| mnaser | nova_consoleauth doesnt even exist in env.d/nova.yaml | 23:14 |
| cloudnull | ah fair enough | 23:14 |
| cloudnull | yes that should be just nova_console | 23:14 |
| mnaser | i wonder how that every deployed :X | 23:15 |
| mnaser | lol | 23:15 |
| osnaya | @cloudnull will script re-run overwrite the backup_openstack_inventory.tar? | 23:16 |
| cloudnull | consoleauth used to be a group, I wonder if this was something that was missed in queens when we did the group reduction work | 23:17 |
| cloudnull | osnaya: no, the backups are timestamped | 23:17 |
| osnaya | @cloudnull ok, thanks for the clarification | 23:18 |
| mnaser | cloudnull: https://github.com/openstack/openstack-ansible/blob/master/playbooks/os-nova-install.yml#L105-L109 | 23:18 |
| mnaser | it is installed later | 23:18 |
| *** masber has joined #openstack-ansible | 23:27 | |
| *** masber has quit IRC | 23:30 | |
| *** masber has joined #openstack-ansible | 23:30 | |
| *** flemingo has joined #openstack-ansible | 23:40 | |
| *** flemingo has quit IRC | 23:40 | |
| *** flemingo has joined #openstack-ansible | 23:41 | |
| *** osnaya has quit IRC | 23:50 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
