| *** cloudtrainme has quit IRC | 00:12 | |
| *** sdake has quit IRC | 00:14 | |
| *** furlongm has joined #openstack-ansible | 00:17 | |
| *** abitha has quit IRC | 00:18 | |
| *** abitha has joined #openstack-ansible | 00:19 | |
| *** abitha has quit IRC | 00:24 | |
| *** markvoelker has quit IRC | 00:38 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-rabbitmq_server: [WIP] Multi-distro support for rabbitmq-server role https://review.openstack.org/286282 | 00:45 |
|---|---|---|
| *** weezS has quit IRC | 00:52 | |
| *** arif-ali has joined #openstack-ansible | 01:00 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 01:17 | |
| *** metral_zzz is now known as metral | 01:32 | |
| *** jthorne has joined #openstack-ansible | 01:37 | |
| *** markvoelker has joined #openstack-ansible | 01:38 | |
| *** jamielennox is now known as jamielennox|away | 01:45 | |
| *** markvoelker has quit IRC | 02:08 | |
| *** markvoelker has joined #openstack-ansible | 02:08 | |
| *** jmckind has quit IRC | 02:23 | |
| *** fawadkhaliq has quit IRC | 02:27 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: Added major version upgrade support (galera) https://review.openstack.org/271591 | 03:13 |
| *** kencjohnston has joined #openstack-ansible | 03:29 | |
| *** kencjohnston has quit IRC | 04:03 | |
| *** shausy has joined #openstack-ansible | 04:03 | |
| *** gfa_ is now known as gfa | 04:08 | |
| *** fawadkhaliq has joined #openstack-ansible | 04:18 | |
| *** daneyon_ has quit IRC | 04:26 | |
| *** jmckind has joined #openstack-ansible | 04:56 | |
| *** shausy has quit IRC | 05:04 | |
| *** kstepniewski has joined #openstack-ansible | 05:11 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: Retry tempest cirros image creation https://review.openstack.org/284057 | 05:24 |
| *** daneyon has joined #openstack-ansible | 05:25 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: Allow Horizon setup with external SSL termination https://review.openstack.org/214647 | 05:25 |
| openstackgerrit | Merged openstack/openstack-ansible: Docs: Ensure a horizontal rule prior to the navigation links https://review.openstack.org/285981 | 05:25 |
| *** jmckind has quit IRC | 05:33 | |
| *** javeriak has joined #openstack-ansible | 05:35 | |
| *** javeriak has quit IRC | 05:40 | |
| *** javeriak has joined #openstack-ansible | 05:40 | |
| *** fawadkhaliq has quit IRC | 05:42 | |
| *** javeriak_ has joined #openstack-ansible | 05:43 | |
| *** javeriak has quit IRC | 05:45 | |
| *** jthorne has quit IRC | 05:46 | |
| *** fawadkhaliq has joined #openstack-ansible | 05:58 | |
| *** sdake has joined #openstack-ansible | 06:11 | |
| *** rgogunskiy has joined #openstack-ansible | 06:21 | |
| *** daneyon has quit IRC | 06:23 | |
| *** sdake has quit IRC | 06:27 | |
| *** phiche has joined #openstack-ansible | 06:32 | |
| *** sdake has joined #openstack-ansible | 06:33 | |
| *** mariusv has quit IRC | 06:35 | |
| *** javeriak_ has quit IRC | 06:42 | |
| *** javeriak has joined #openstack-ansible | 06:54 | |
| *** lunarlamp has joined #openstack-ansible | 06:56 | |
| *** phiche has quit IRC | 06:57 | |
| *** daneyon has joined #openstack-ansible | 06:59 | |
| *** cemmason has joined #openstack-ansible | 06:59 | |
| *** phiche has joined #openstack-ansible | 07:03 | |
| *** daneyon has quit IRC | 07:04 | |
| *** sdake has quit IRC | 07:10 | |
| *** daneyon has joined #openstack-ansible | 07:10 | |
| *** sdake has joined #openstack-ansible | 07:12 | |
| *** mikelk has joined #openstack-ansible | 07:16 | |
| *** cemmason has quit IRC | 07:24 | |
| *** cemmason has joined #openstack-ansible | 07:24 | |
| *** cemmason has quit IRC | 07:45 | |
| *** sdake has quit IRC | 07:48 | |
| *** fawadkhaliq has quit IRC | 08:08 | |
| *** pcaruana has joined #openstack-ansible | 08:09 | |
| *** ccesario has quit IRC | 08:09 | |
| *** daneyon has quit IRC | 08:17 | |
| *** jwitko has quit IRC | 08:24 | |
| *** ccesario has joined #openstack-ansible | 08:38 | |
| *** admin0 has joined #openstack-ansible | 08:39 | |
| *** admin0 has quit IRC | 08:40 | |
| *** admin0 has joined #openstack-ansible | 08:43 | |
| *** openstackgerrit has quit IRC | 08:48 | |
| *** openstackgerrit has joined #openstack-ansible | 08:48 | |
| *** admin0 has quit IRC | 08:54 | |
| *** asettle has joined #openstack-ansible | 09:10 | |
| mattt | odyssey4me: is the relationship between https://review.openstack.org/#/c/284977 and https://review.openstack.org/#/c/284896/ wrong? should https://review.openstack.org/#/c/284896/ be the parent? | 09:14 |
| *** swati_ has joined #openstack-ansible | 09:19 | |
| *** mathlin has joined #openstack-ansible | 09:20 | |
| *** admin0 has joined #openstack-ansible | 09:24 | |
| swati_ | Hi,...I am trying to use designate on liberty....and all the services are up | 09:28 |
| *** keedya has quit IRC | 09:29 | |
| swati_ | but I get this error or "Unable to establish connection at endpoint http://controller:9001"... | 09:29 |
| swati_ | so I am not able to run designate commands | 09:29 |
| mattt | swati_: aren't you the person implementing designate? :) | 09:36 |
| swati_ | yes I am | 09:36 |
| mattt | then i think you're probably best to answer that question :) | 09:36 |
| swati_ | it ran fine on Kilo and I am rebasing it on Liberty | 09:36 |
| odyssey4me | if you can't reach the designate service, verify that it is up and that the load balancer can reach t | 09:36 |
| swati_ | all designate services are up | 09:37 |
| *** kstepniewski has quit IRC | 09:37 | |
| swati_ | my purpose of asking is may be somebody else might have encountered a similar issue for some other component in Ansible-Liberty | 09:38 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Added the pip packages for the repo-build process https://review.openstack.org/284896 | 09:38 |
| odyssey4me | swati_ is the designate service up in the containers you have it in? | 09:38 |
| mattt | swati_: can't say i have -- check the designate logs, try connecting locally from within the container, make sure that all works | 09:39 |
| odyssey4me | swati_ and have you verified that the load balancer is configured for it, and that it can reach it? | 09:39 |
| mattt | then troubleshoot through the LB | 09:39 |
| swati_ | Jesse...yes all designate services are up in the container and logs dont show an error as well | 09:40 |
| swati_ | the LB thing I need to verify | 09:41 |
| *** kstepniewski has joined #openstack-ansible | 09:41 | |
| swati_ | Mattt, any specific files that I need to check for LB configuration? | 09:41 |
| *** kstepniewski has quit IRC | 09:45 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: IRR - Implemented for os_keystone https://review.openstack.org/279710 | 09:45 |
| mattt | swati_: well can you connect to the service from within the container ? | 09:46 |
| mattt | swati_: start local and go from there | 09:46 |
| *** jthorne has joined #openstack-ansible | 09:46 | |
| swati_ | designate services are there within the container only | 09:47 |
| mattt | swati_: right but can you connect to them? | 09:47 |
| mattt | swati_: as far as LB, make sure there is a config for designate in /etc/haproxy/conf.d on the node where your LB runs | 09:47 |
| *** jthorne has quit IRC | 09:50 | |
| swati_ | ok mattt i will check /etc/haproxy/conf.d................what do you mean by connecting to designate service, if u can pls explain? | 09:51 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: IRR - Implemented for all os_* and dep roles https://review.openstack.org/284987 | 09:53 |
| mattt | swati_: well, you said you got "Unable to establish connection at endpoint http://controller:9001". where do you see that? | 09:56 |
| mattt | swati_: or put another way, how are you connecting to designate to receive that error? | 09:56 |
| swati_ | there is a designate container where all designate services are running | 09:56 |
| mattt | swati_: right, but how are you connecting to designate? keep in mind i've never used it | 09:57 |
| swati_ | inside the same container, I am firing the designate commands | 09:57 |
| mattt | swati_: alright, so if the designate commands are using your /root/openrc file | 09:57 |
| mattt | then you also need to make sure that designate is registered as an endpoint and service in keystone | 09:58 |
| swati_ | openrc file, endpoint and designate service everything is already created | 09:59 |
| swati_ | You can view the openrc file also at https://github.com/sharmaswati6/designate_files/blob/master/openrc | 09:59 |
| mattt | swati_: ok | 10:01 |
| mattt | swati_: find the management IP of your designate container | 10:01 |
| mattt | swati_: and then try designate --os-endpoint http://<ip>:9001 or whatever the endpoint is | 10:02 |
| swati_ | yes I know | 10:02 |
| mattt | try to circumnavigate your LB | 10:02 |
| swati_ | ok i will try this | 10:03 |
| -openstackstatus- NOTICE: Gerrit is going to be restarted due to poor performance | 10:34 | |
| *** ChanServ changes topic to "Gerrit is going to be restarted due to poor performance" | 10:34 | |
| *** ChanServ changes topic to "Launchpad: https://launchpad.net/openstack-ansible || Weekly Meetings: https://wiki.openstack.org/wiki/Meetings/openstack-ansible || Review Dashboard: https://goo.gl/tTmdgs" | 10:40 | |
| -openstackstatus- NOTICE: gerrit finished restartign | 10:40 | |
| *** bsv has joined #openstack-ansible | 10:56 | |
| mattt | swati_: what did you find | 11:04 |
| bsv | So, succesfully deployed all playbooks and openstack is working, but no external internet access. Documentation does not contain any reference to the external networking part. | 11:05 |
| admin0 | bsv: yes . then you need to login to horizon as admin and start adding the networks | 11:07 |
| admin0 | set it to external, set the router and ips and you are done | 11:07 |
| bsv | I cannot see the external network anywhere in the documentation, what is it called? | 11:08 |
| bsv | The network implemented with the playbooks seem to match the "Classic with Linux Bridge" scenario. | 11:10 |
| mattt | bsv: the playbooks don't provision any actual provider/tenant networks tho | 11:10 |
| mattt | bsv: as admin0 said, you'll need to do that through horizon or through the neutron cli | 11:10 |
| bsv | mattt: i know, but I cannot see any reference about how/where to make the connection. | 11:11 |
| bsv | My guess, would be to map the "external" NIC into the container running the neutron L3 agent. | 11:11 |
| mattt | bsv: i see, let me see if i can find docs | 11:13 |
| bsv | mattt: thanks. | 11:13 |
| mattt | bsv: did you see http://docs.openstack.org/developer/openstack-ansible/install-guide/targethosts-network.html ? | 11:14 |
| bsv | yep | 11:14 |
| bsv | When I read that page, I see the setup for east-west traffic, nothing about north-south (external access to internet). | 11:15 |
| bsv | I did a search for 'external', and find info about containers. | 11:16 |
| mattt | bsv: search for 'provider' | 11:16 |
| bsv | mattt: br-vlan? | 11:17 |
| mattt | bsv: yeah that is the one you'll need to focus on | 11:17 |
| mattt | disclaimer: i'm pathetic when it comes to networking :) | 11:18 |
| bsv | So br-vlan is used to carry vlan-tagged traffic for tenants AND external traffic (also tagged) ? | 11:19 |
| mattt | bsv: you could use br-vxlan for tenant traffic | 11:19 |
| bsv | i do | 11:20 |
| bsv | but some applications (low latency) would prefer VLANs | 11:20 |
| bsv | So, just to refresh :) br-vlan on bond1 is actually used for 'external' traffic? | 11:21 |
| admin0 | bsv, i can attempt to explain .. you have 4 bridges right ? | 11:22 |
| admin0 | br-mgmt, br-storage, br-vxlan and br-vlan | 11:22 |
| bsv | correct | 11:22 |
| admin0 | 1 and 2 is obvious | 11:22 |
| bsv | yes | 11:22 |
| admin0 | br-vxlan is eas-west traiffc ( traffic between VMS ) | 11:22 |
| *** evrardjp has joined #openstack-ansible | 11:22 | |
| bsv | yes | 11:22 |
| admin0 | br-vlan is the uplink to the world (public ) | 11:22 |
| bsv | oh | 11:22 |
| bsv | admin0: thank you, that explains the "missing" link to the outside world. | 11:24 |
| admin0 | bsv, i am making a nice document for this missing link | 11:24 |
| mattt | admin0: yay! | 11:24 |
| mattt | thanks for filling in :) | 11:24 |
| admin0 | everyone stops on now openstack is done | 11:24 |
| admin0 | no one says .. this is what you need to do next | 11:24 |
| admin0 | that is where I play and play :D | 11:24 |
| bsv | There is some security issues with this setup, but I'll go into that when this PoC setup works :) | 11:25 |
| bsv | admin0: I only allow access to the external-vlan on the NICs used by the neutron L3 agent, right? Not every single compute-node? | 11:28 |
| bsv | in this case, the infrastructure hosts running containers with neutron-agents | 11:29 |
| bsv | admin0: or? :) | 11:33 |
| tiagogomes | and gerrit is down | 11:35 |
| *** tiagogomes has quit IRC | 11:39 | |
| admin0 | that is right | 11:39 |
| admin0 | bsv: | 11:39 |
| admin0 | there are 2 approaches here | 11:39 |
| admin0 | floating IP and flat DHCP | 11:39 |
| admin0 | if you allow flat DHCP, then you might need public access on all copute nodes since the instances get direct IP from the DHCP | 11:40 |
| *** neilus has joined #openstack-ansible | 11:40 | |
| bsv | how is loadbalancing handled then? | 11:40 |
| admin0 | which load balancing? | 11:41 |
| bsv | if a tenant want to use a LB in front of n+ services, for public access. | 11:42 |
| bsv | would it then 'hit' the LB-agent running inside the infrastructure containers? | 11:42 |
| admin0 | i use this setup .. public router -> public switch -> vyatta/pfsense -> openstack | 11:43 |
| admin0 | so everything I have is controlled | 11:43 |
| bsv | ah, I have routed a public /24 to the 'external-vlan'. | 11:44 |
| admin0 | basically, if you use direct piublic IPs, you are losing ips :D .. so I use the vyatta/pfsense so all my public ips live there, and from there i do a nat , port-forward to the required services | 11:44 |
| admin0 | so in my cluster, public is a simulated/assumed private range | 11:45 |
| bsv | ok | 11:45 |
| bsv | I need to have 1:1 NAT of public addresses directly into some instances. | 11:45 |
| bsv | sry | 11:46 |
| bsv | wrong | 11:46 |
| bsv | I need to assign public IP's to some instances. | 11:46 |
| admin0 | then you need to add the ips and then use floating ips | 11:46 |
| bsv | on the interface of the instance. | 11:46 |
| admin0 | direct public ? | 11:46 |
| bsv | yeah (voip is a bitch) | 11:46 |
| * admin0 worked in voip for 4 years before jumping to the cloud :D | 11:47 | |
| bsv | the traffic will be filtered before it reaches openstack, but still. :) | 11:47 |
| bsv | :D | 11:47 |
| bsv | i'll try to do that br-vlan mapping, see if it works :) | 11:47 |
| bsv | and take it from there. | 11:47 |
| bsv | thank you | 11:48 |
| admin0 | bsv, i have examples :) | 11:50 |
| admin0 | hold | 11:50 |
| admin0 | eth0 = br-vlan eth0.10 = br-mgmt eth0.20 = br-storage eth0.30 = br-vxlan — so eth0 = br-vlan = untagged there .. and in my router i have vlan 101, 102 , 103 and 201, 202, 203 .. | 11:52 |
| admin0 | so there are 2 ways a instance can connect to the intenret .. flaoting-ip and direct-ip | 11:52 |
| admin0 | so my vlan is 101 for floating ip and vlan 201 for direct internet | 11:53 |
| admin0 | so first, floating ip: neutron net-create floating-192-168-101 --shared --router:external true --provider:network_type vlan --provider:physical_network vlan --provider:segmentation_id 101 .. and then neutron subnet-create --allocation-pool start=192.168.101.5,end=192.168.101.250 --ip-version 4 --gateway 192.168.101.1 --disable-dhcp --name floating-192-168-101 --dns-nameserver 8.8.8.8 f0fe6db6-5299-49cf-9445-b5511255446e | 11:53 |
| admin0 | 192.168.101.0/24 | 11:53 |
| admin0 | now i need to be abel to have instances on direct internet as well | 11:53 |
| admin0 | neutron net-create dhcp-192-168-201 --shared --router:external false --provider:network_type vlan --provider:physical_network vlan --provider:segmentation_id 201 , and then neutron subnet-create --allocation-pool start=192.168.201.5,end=192.168.201.250 --ip-version 4 --no-gateway --host-route destination=0.0.0.0/0,nexthop=192.168.201.1 --enable-dhcp --name dhcp-192-168-201 --dns-nameserver 8.8.8.8 023394b8-1b9e-47a8-be50-9a0fd9b842f | 11:54 |
| admin0 | 192.168.201.0/24 | 11:54 |
| admin0 | with this, you create a network and then assign floating ip from the floating-192-168-101 range | 11:54 |
| admin0 | or just create instance on the dhcp-192-168-201 and it will get IPs directly | 11:54 |
| admin0 | replace the 101 and 102 and the ips with proper public range | 11:54 |
| admin0 | the id after the nameserver is the ID you get after having the first command | 11:55 |
| admin0 | bsv: did that help ? | 12:01 |
| *** johnmilton has joined #openstack-ansible | 12:02 | |
| *** admin0 has left #openstack-ansible | 12:03 | |
| *** admin0 has joined #openstack-ansible | 12:04 | |
| *** bjornar__ has quit IRC | 12:07 | |
| *** andrei_ has quit IRC | 12:07 | |
| *** stian__ has quit IRC | 12:07 | |
| *** andrei_ has joined #openstack-ansible | 12:08 | |
| *** stian__ has joined #openstack-ansible | 12:09 | |
| *** bjornar__ has joined #openstack-ansible | 12:09 | |
| *** javeriak has quit IRC | 12:09 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server: Use slurp to collect the repo_server ssh keys https://review.openstack.org/286042 | 12:14 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server: Removed sshd_config file https://review.openstack.org/284410 | 12:15 |
| bsv | admin0: sry, was afk.. lunch | 12:20 |
| bsv | admin0: reading.... :) | 12:22 |
| *** ccesario has quit IRC | 12:24 | |
| *** tiagogomes has joined #openstack-ansible | 12:24 | |
| bsv | admin0: yes, much better. | 12:25 |
| bsv | While reading the docs, I did not get the impression that br-vlan was for "public" traffic. | 12:26 |
| admin0 | well, its just an interface name :D | 12:26 |
| admin0 | make it br-bsv :D | 12:26 |
| bsv | :) | 12:26 |
| bsv | I would use a dedicated interface for external traffic. | 12:28 |
| admin0 | well, in my example its a vlan | 12:28 |
| admin0 | you can have anything | 12:28 |
| admin0 | all ansible/openstack cares is that it sees br-vlan ( or br-bsv there ) | 12:28 |
| *** swati_ has quit IRC | 12:28 | |
| admin0 | and in physical it can be a dedi interface, vlan, bond | 12:29 |
| bsv | yes, but naming it br-vlan, indicated (to me), that it was for east-west traffic using vlans instead of vxlans. | 12:29 |
| bsv | I was looking for something like br-provider/br-external :) | 12:29 |
| admin0 | usually public ip ranges are separated using vlans .. so i think it was their way of saying br-vlan = this is where you specicy your vlans | 12:29 |
| bsv | yeah | 12:30 |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible: Add ability to set up Swift container-sync realms https://review.openstack.org/286545 | 12:32 |
| admin0 | yesterday i managed to create a new my_awesome_hosts and run some services in there :D | 12:33 |
| admin0 | on top of *-infra_hosts | 12:33 |
| *** retreved has joined #openstack-ansible | 12:49 | |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible-os_swift: Add ability to set up Swift container-sync realms https://review.openstack.org/286568 | 12:57 |
| *** neilus1 has joined #openstack-ansible | 13:00 | |
| *** woodard has joined #openstack-ansible | 13:01 | |
| *** neilus has quit IRC | 13:04 | |
| *** mgoddard_ has joined #openstack-ansible | 13:04 | |
| *** neilus1 has quit IRC | 13:04 | |
| *** tiagogomes has quit IRC | 13:06 | |
| *** neilus has joined #openstack-ansible | 13:06 | |
| *** mgoddard has quit IRC | 13:07 | |
| *** tiagogomes has joined #openstack-ansible | 13:07 | |
| *** neilus has quit IRC | 13:11 | |
| *** neilus has joined #openstack-ansible | 13:11 | |
| *** cloudtrainme has joined #openstack-ansible | 13:12 | |
| *** tiagogomes has quit IRC | 13:14 | |
| *** sdake has joined #openstack-ansible | 13:16 | |
| *** neilus has quit IRC | 13:16 | |
| *** furlongm has quit IRC | 13:16 | |
| *** sdake has quit IRC | 13:21 | |
| *** sdake has joined #openstack-ansible | 13:21 | |
| *** furlongm has joined #openstack-ansible | 13:21 | |
| *** evrardjp has quit IRC | 13:22 | |
| *** winggundamth has joined #openstack-ansible | 13:22 | |
| winggundamth | hi | 13:26 |
| *** mgoddard_ has quit IRC | 13:26 | |
| winggundamth | I have a question about flat, vlan and vxlan configuration. | 13:26 |
| admin0 | ok | 13:27 |
| *** mgoddard_ has joined #openstack-ansible | 13:27 | |
| admin0 | winggundamth: i replied to a similar one an hour back :D | 13:27 |
| winggundamth | do I need it all? what is the difference between them? | 13:27 |
| winggundamth | lol | 13:27 |
| winggundamth | so do you have history somewhere? | 13:27 |
| admin0 | yes :D | 13:27 |
| admin0 | everything in the internet is logged | 13:27 |
| winggundamth | haha | 13:27 |
| winggundamth | so what is the url for this channel irc log? | 13:28 |
| odyssey4me | openstack has eavesdrop logs of all official channels :) | 13:28 |
| odyssey4me | http://eavesdrop.openstack.org/irclogs/%23openstack-ansible/ | 13:28 |
| *** neilus has joined #openstack-ansible | 13:28 | |
| odyssey4me | winggundamth admin0 ^ | 13:28 |
| winggundamth | got it | 13:29 |
| admin0 | http://eavesdrop.openstack.org/irclogs/%23openstack-ansible/%23openstack-ansible.2016-03-01.log.html — search for br-vlan / admin0 | 13:29 |
| odyssey4me | mattt andymccr hughsaunders cloudnull automagically jmccrory d34dh0r53 stevelle we need all hands on deck to resolve the current master branch breakage which includes the role breakout work - please join https://etherpad.openstack.org/p/openstack-ansible-role-breakout to help identify the work to be done and to assign work to yourself | 13:30 |
| mhayden | Apsu: you may want to add yourself onto this review -> https://review.openstack.org/#/c/275440/7 | 13:30 |
| * mhayden looks at odyssey4me's list and gulps | 13:31 | |
| *** neilus1 has joined #openstack-ansible | 13:31 | |
| *** neilus1 has quit IRC | 13:31 | |
| *** neilus1 has joined #openstack-ansible | 13:31 | |
| winggundamth | reading | 13:32 |
| *** tiagogomes has joined #openstack-ansible | 13:32 | |
| *** neilus has quit IRC | 13:34 | |
| * tiagogomes wonders how difficult would be having keystone and horizon sharing the same container/bare metal machine | 13:34 | |
| mattt | tiagogomes: it doesn't work at the moment | 13:34 |
| tiagogomes | yes i know | 13:34 |
| *** cloudtrainme has quit IRC | 13:34 | |
| mattt | tiagogomes: i wouldn't say hard, but a bit of work is required to get apache to jive between the two | 13:35 |
| mattt | so did you just sign up for that work? :) | 13:35 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server: Implement verbose logging for functional tests https://review.openstack.org/286591 | 13:35 |
| admin0 | tiagogomes: easy if its 2 different containers | 13:35 |
| admin0 | on the same machine | 13:35 |
| tiagogomes | I am finding containers unreliable. Sometimes deployment fails with no SSH connection to some container | 13:36 |
| odyssey4me | tiagogomes a question that certainly stands is why you'd want to :) but yeah, changes could be implement to improve colocation of services that use apache | 13:36 |
| odyssey4me | tiagogomes oh? we do many, many gate tests daily and don't see that | 13:36 |
| tiagogomes | Really, last time I asked here about that, other people experienced the same | 13:37 |
| odyssey4me | tiagogomes we used to see that, but generally found that with the default ubuntu sshd_config as long as you keep forks under 10 you should be fine | 13:37 |
| *** furlongm_ has joined #openstack-ansible | 13:38 | |
| winggundamth | admin0: so the flat mean flat-dhcp and vlan mean floating-ip and I have to choose one right? | 13:39 |
| *** furlongm has quit IRC | 13:40 | |
| tiagogomes | Another reason two use baremetal, sometimes the controller machine is an embedded device constrained in terms of resources and you don't want to hungry apache2 processes running there | 13:41 |
| tiagogomes | s/two/to | 13:41 |
| admin0 | winggundamth: i have not used flat . i use br-vlan for both floating-ip as well as flat | 13:42 |
| admin0 | flat-dhcp | 13:42 |
| automagically | odyssey4me: Checking the etherpad now | 13:43 |
| winggundamth | admin0: ok. I confuse about type is vlan, container_bridge is br-vlan. how about the range? do I need to config to match with switch? | 13:44 |
| winggundamth | for example if I have 1.2.3.4 - 1.2.3.10 as my floating-ip and my core switch assign it to vlan 10. so I have to config range 10:10? | 13:45 |
| admin0 | no | 13:46 |
| admin0 | you just add it (like I showed in that example ) | 13:46 |
| *** neilus has joined #openstack-ansible | 13:49 | |
| winggundamth | admin0: I'm not sure I understand it. so what is the meaning of range in provider_networks configuration? | 13:50 |
| *** sdake_ has joined #openstack-ansible | 13:50 | |
| admin0 | for east-west traffic, if you use vlan , you specify the range there, if you use vxlan, you specify the vxlan range there | 13:50 |
| winggundamth | oh ic | 13:51 |
| odyssey4me | thanks automagically | 13:51 |
| admin0 | winggundamth: but yyes i think, on br-vlan, you still need to whitelist that ip range | 13:51 |
| *** jguy_ has joined #openstack-ansible | 13:51 | |
| odyssey4me | automagically in theory, the os_keystone role should be able to functionally pass already with a few bits added to make it work in openstack-ci | 13:52 |
| *** neilus1 has quit IRC | 13:52 | |
| admin0 | so to be safe, add say 10:100 in br-vlan , and you can use those vlans are floating/dhcp range when you are adding the network | 13:53 |
| *** neilus has quit IRC | 13:54 | |
| *** sdake has quit IRC | 13:54 | |
| winggundamth | admin0: so from here http://docs.openstack.org/developer/openstack-ansible/install-guide/configure-networking.html from step 7. this configure means east-west traffic or does it mean floating-ip for type vlan? | 13:55 |
| admin0 | winggundamth: if you read my earlier example, of how i am using vlans, this is what i have br-vxlan range: "1:1000”, br-vlan: range: "100:300" | 13:55 |
| openstackgerrit | Andy McCrae proposed openstack/openstack-ansible-os_swift: Add ability to set up Swift container-sync realms https://review.openstack.org/286568 | 13:55 |
| odyssey4me | andymccr perhaps you can pick up the swift role work in https://etherpad.openstack.org/p/openstack-ansible-role-breakout ? | 13:56 |
| jguy_ | Hi everyone. I finally got my OSAD setup up and running. I have a bit of a theory question though. I want to test the L2gateway package, and the documentation is rather vague. It says to install parts on the controller, and network node. However, in OSAD there is not a specific container for controller or network node. | 13:56 |
| admin0 | so i add the 101, 102, 103 and 201, 202, 203 as flaoting and direct-dhcp range | 13:56 |
| odyssey4me | jguy_ the network node is, by default, the neutron_agents container | 13:56 |
| jguy_ | I assume the controller is the neutron server container? | 13:57 |
| winggundamth | admin0: so what you said about router. it is physical router or virtual router in openstack that have vlan 101 102....? | 13:57 |
| odyssey4me | jguy_ the controller is the set of hosts with all the containers on it :) | 13:57 |
| odyssey4me | jguy_ ie the host which contains all the controlling backplane | 13:58 |
| jguy_ | Sure, just trying to learn the mapping from a “classical” openstack install guide terminology, to the separated containers model. | 13:59 |
| jguy_ | typically the basic install has controller, networtk, and compute | 13:59 |
| jguy_ | I will also be testing Ironic, and some other packages. So I want to make sure I understand this at a deeper lever. | 14:00 |
| odyssey4me | jguy_ I'm pretty sure that our install guide could use with some adjustments to help with the mapping to the openstack documentation, and patches to do so would be appreciated! | 14:00 |
| odyssey4me | jguy_ if you could make notes in an etherpad, then we can help by answering questions in the etherpad and the notes can form the basis of some documentation updates :) | 14:01 |
| jguy_ | I would be more than happy to do that…but I need to understand it first. | 14:01 |
| *** XuChao918 has joined #openstack-ansible | 14:02 | |
| jguy_ | I agree the network node and the compute nodes mapping is rather obvious. | 14:02 |
| winggundamth | admin0: so the range in configuration is like limit of segmentation_id that I can create right? | 14:02 |
| cloudnull | morning | 14:03 |
| jguy_ | thanks odyssey4me. I will do a little more digging, and see what I can do to clarify things. | 14:05 |
| cloudnull | odyssey4me: whats the current breakage ? | 14:05 |
| *** furlongm_ is now known as furlongm | 14:09 | |
| *** neilus has joined #openstack-ansible | 14:12 | |
| *** XuChao918 has quit IRC | 14:12 | |
| admin0 | winggundamth: today i plan to redo www.openstackfaq.com with my setup where i go into detail on explaining this :) | 14:14 |
| *** XuChao918 has joined #openstack-ansible | 14:14 | |
| *** XuChao918 has quit IRC | 14:15 | |
| *** XuChao918 has joined #openstack-ansible | 14:15 | |
| *** rgogunskiy has quit IRC | 14:16 | |
| winggundamth | admin: that's nice. I'm looking forward to it | 14:16 |
| *** Xu_Chao has joined #openstack-ansible | 14:18 | |
| *** Xu_Chao has quit IRC | 14:19 | |
| *** XuChao918 has quit IRC | 14:19 | |
| *** sdake_ is now known as sdake | 14:19 | |
| *** Xu_Chao has joined #openstack-ansible | 14:19 | |
| *** winggundamth has quit IRC | 14:24 | |
| jguy_ | Does OSAD support adding openvswitch mechanism driver? Or is this something I have to manually add to the setup? | 14:25 |
| *** furlongm_ has joined #openstack-ansible | 14:25 | |
| *** furlongm has quit IRC | 14:26 | |
| *** ccesario has joined #openstack-ansible | 14:26 | |
| Xu_Chao | openstack-ansible project support the deployment of OpenStack HA architecture? | 14:28 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-openstack_openrc: [WIP] IRR-ify https://review.openstack.org/286623 | 14:29 |
| *** jroll has quit IRC | 14:29 | |
| *** jroll has joined #openstack-ansible | 14:30 | |
| *** jroll has quit IRC | 14:30 | |
| *** jroll has joined #openstack-ansible | 14:30 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-repo_build: Updated base repo for IRR https://review.openstack.org/286625 | 14:31 |
| cloudnull | as you all are working on basic IRR bits I have https://github.com/os-cloud/openstack-ansible-TEMPLATE which is a simple skeleton of an IRR repo which can be used to rapidly update the new roles. | 14:33 |
| cloudnull | just copy the bits in the repo your working on and run the find / replace command to give everything the new name find . -type f -exec sed -i 's/TEMPLATE/CHANGE_ME_PLEASE/g' {} \; | 14:34 |
| mattt | cloudnull: that tox.ini looks a bit out of date | 14:35 |
| cloudnull | it is | 14:35 |
| cloudnull | last updated that repo a month or so ago | 14:35 |
| cloudnull | I can push a new version. | 14:35 |
| *** winggundamth has joined #openstack-ansible | 14:35 | |
| winggundamth | got answer from here https://developer.rackspace.com/blog/neutron-networking-vlan-provider-networks/ | 14:35 |
| *** weshay has joined #openstack-ansible | 14:36 | |
| mattt | cloudnull: any idea why linters is failing here? https://gist.githubusercontent.com/mattt416/9b1b98301f3ee350d7d2/raw/277d30e9c9353b228e98a0e94b7ab0cd24543299/gistfile1.txt | 14:36 |
| cloudnull | mattt: pushed a new version of the tox file to the TEMPLATE repo | 14:36 |
| mattt | cloudnull: sweet! | 14:36 |
| cloudnull | mattt: ansible-lint is not installed | 14:37 |
| mattt | cloudnull: if i source .tox/linters/bin/activate, it's there | 14:37 |
| Xu_Chao | Ansible deploy OpenStack, what is the Web tools？ | 14:38 |
| cloudnull | Xu_Chao: web tools? | 14:39 |
| mattt | horizon ? | 14:40 |
| mattt | :) | 14:40 |
| Xu_Chao | yes | 14:40 |
| cloudnull | mattt: idk , your openrc role is passing the centos gate https://jenkins06.openstack.org/job/gate-openstack-ansible-openstack_openrc-dsvm-ansible-functional-centos7-nv/1/console | 14:40 |
| Xu_Chao | Implementing web installation | 14:40 |
| cloudnull | maybe something wonky with the ubuntu image ? | 14:40 |
| mattt | cloudnull: the functional test is a noop at the moment :P | 14:41 |
| cloudnull | right . | 14:41 |
| *** jthorne has joined #openstack-ansible | 14:41 | |
| cloudnull | but passed :) | 14:41 |
| *** jthorne has quit IRC | 14:42 | |
| mattt | all three lint jobs fail for some reason, hrrmph. | 14:42 |
| *** jthorne has joined #openstack-ansible | 14:42 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: Implements IRR for the repo_build role https://review.openstack.org/284896 | 14:43 |
| cloudnull | I just updated all of this for my repo_build role, so ill see if i hit the same thing there. | 14:43 |
| Xu_Chao | After considering the use of Ansible UI to deploy OpenStack? | 14:44 |
| *** furlongm_ has quit IRC | 14:44 | |
| cloudnull | Xu_Chao: if your talking about tower we've not done much with it . | 14:45 |
| cloudnull | there are a few folks exploring with tower though | 14:45 |
| *** ric has joined #openstack-ansible | 14:46 | |
| *** neilus has quit IRC | 14:46 | |
| *** neilus has joined #openstack-ansible | 14:47 | |
| *** furlongm_ has joined #openstack-ansible | 14:47 | |
| Xu_Chao | Look forward to | 14:48 |
| *** sdake has quit IRC | 14:49 | |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-openstack_openrc: [WIP] IRR-ify https://review.openstack.org/286623 | 14:50 |
| mattt | cloudnull: ok, i think one issue is i had no bash scripts so that job was failing | 14:50 |
| Xu_Chao | The ansible project to deploy OpenStack HA architecture, what time can be achieved | 14:51 |
| *** neilus has quit IRC | 14:51 | |
| Xu_Chao | thanks | 14:51 |
| cloudnull | Xu_Chao: time all depends on how many nodes you have. | 14:52 |
| *** Mudpuppy has joined #openstack-ansible | 14:52 | |
| *** Mudpuppy has quit IRC | 14:52 | |
| *** Mudpuppy has joined #openstack-ansible | 14:53 | |
| cloudnull | In my experience, an install of around <10 nodes can be done in ~1-2 hours depending on network speed. Adding nodes to a deployment can be done at whenever and will take different time based on whats being added. | 14:54 |
| odyssey4me | o/ back from being afk | 14:55 |
| bsv | wb o7 | 14:55 |
| cloudnull | odyssey4me mattt it would seem that https://review.openstack.org/#/c/284896/ and https://review.openstack.org/#/c/286625/ are the reason the gate is blocked ? | 14:55 |
| odyssey4me | cloudnull the repo_server role seems to not be working in copenstack-ci any more, likely based on infra changes to the image - I'm looking into it | 14:56 |
| *** sdake has joined #openstack-ansible | 14:56 | |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-openstack_openrc: [WIP] IRR-ify https://review.openstack.org/286623 | 14:57 |
| cloudnull | odyssey4me: : seems a package may be missing ? | 14:57 |
| cloudnull | http://logs.openstack.org/10/284410/2/check/gate-openstack-ansible-repo_server-dsvm-ansible-functional-trusty/4b2915f/console.html#_2016-03-01_12_44_10_899 | 14:57 |
| cloudnull | maybe related to the culing of the available repositories apt has access to ? | 14:58 |
| odyssey4me | cloudnull yep, I'll figure it out - not to worry | 14:58 |
| cloudnull | okiedokie | 14:58 |
| odyssey4me | cloudnull if you can pick a role to work on from https://etherpad.openstack.org/p/openstack-ansible-role-breakout then it'll help get things moving again | 14:59 |
| cloudnull | im working on repo_build already | 14:59 |
| odyssey4me | cloudnull ah, I see you're doing the repo_build - thanks | 14:59 |
| cloudnull | os_keystone should be good to go too | 14:59 |
| odyssey4me | I'll likely figure out the fix for repo_server in the next hour | 14:59 |
| *** busterswt has joined #openstack-ansible | 15:02 | |
| mattt | cloudnull: can you update readme.rst in your skeleton ? | 15:02 |
| mattt | (uppercase README) | 15:02 |
| *** ShannonM has joined #openstack-ansible | 15:02 | |
| *** fawadkhaliq has joined #openstack-ansible | 15:02 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server: Implement human readable logging for functional tests https://review.openstack.org/286591 | 15:03 |
| cloudnull | mattt: done | 15:03 |
| *** KLevenstein has joined #openstack-ansible | 15:04 | |
| mattt | cloudnull: thanks! | 15:04 |
| jmccrory | odyssey4me: i think repo_server is breaking because of the ovh image. similar problem here until mhayden updated test with forced apt update https://review.openstack.org/#/c/285483/ | 15:06 |
| mhayden | the ovh environment is a little interesting | 15:07 |
| odyssey4me | jmccrory yeah, it does seem like the ovh image has some brokenness - I'll have to feed that back to infra | 15:07 |
| odyssey4me | and yeah, I think that perhaps the forced apt cache update in the test playbook is a good idea | 15:07 |
| odyssey4me | let me get that done | 15:07 |
| *** cloudtrainme has joined #openstack-ansible | 15:08 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server: Force initial apt-get update as part of test execution https://review.openstack.org/286648 | 15:10 |
| *** admin0 has quit IRC | 15:13 | |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible-os_horizon: [WIP] IRR-ify https://review.openstack.org/286650 | 15:13 |
| *** admin0 has joined #openstack-ansible | 15:16 | |
| odyssey4me | cloudnull feedback in https://review.openstack.org/#/c/286625/ | 15:18 |
| *** sdake_ has joined #openstack-ansible | 15:20 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 15:20 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-repo_build: Updated base repo for IRR https://review.openstack.org/286625 | 15:20 |
| *** sdake has quit IRC | 15:23 | |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_aodh: [WIP] IRR-ify https://review.openstack.org/286655 | 15:23 |
| odyssey4me | cloudnull mattt jmccrory automagically please use the branch name 'bp/independent-role-repositories' for each repo (to ensure that the topic shows this, and the bp is updated)... also I think a good subject line would simply be something like 'Implement base configuration for independent repository' | 15:23 |
| palendae | Yeah, "IRR-ify" will mean nothing in the future | 15:24 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-openstack_openrc: [WIP] IRR-ify https://review.openstack.org/286623 | 15:25 |
| *** raddaoui has joined #openstack-ansible | 15:25 | |
| mattt | palendae: i used that while a WIP, i'll set it once i know everything that is changing | 15:25 |
| mattt | but at this point i don't | 15:25 |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible-os_horizon: [WIP] Implement base configuration for independent repository https://review.openstack.org/286650 | 15:26 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-repo_build: Updated base repo for IRR https://review.openstack.org/286625 | 15:26 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add missing runtime information to upgrade script https://review.openstack.org/286251 | 15:26 |
| cloudnull | odyssey4me: updated the repo_build PRs | 15:27 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add Ansible bootstrapping to Liberty upgrade https://review.openstack.org/286269 | 15:28 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible: Docs: Add missing horizontal rule / navigation links https://review.openstack.org/286657 | 15:30 |
| odyssey4me | jmccrory thanks for the pointer to those inconsistencies! ^ | 15:30 |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible-os_horizon: [WIP] Implement base configuration for independent repository https://review.openstack.org/286650 | 15:32 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add upgrade config change tools https://review.openstack.org/285564 | 15:32 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-repo_build: Updated repo_build for IRR https://review.openstack.org/286625 | 15:34 |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible-os_keystone: [WIP] Implement base configuration for independent repository https://review.openstack.org/286660 | 15:34 |
| admin0 | if anyone knows: i have a instance where cloud-init assigns a secondary IP .. but I do not find that IP in the database under ipallocations or in the dnsmasq file , so where does cloud-init get this ip from ? | 15:36 |
| *** asettle has quit IRC | 15:36 | |
| *** javeriak has joined #openstack-ansible | 15:36 | |
| odyssey4me | admin0 no idea | 15:37 |
| *** spotz_zzz is now known as spotz | 15:37 | |
| cloudnull | is it a floating IP or maybe an address that was assigned with neutron "allowed-address-pairs" | 15:37 |
| admin0 | it is not a floating-ip but a direct -attached | 15:38 |
| busterswt | cloudnull looks like the arp proxy patch got merged | 15:38 |
| cloudnull | WOOT! | 15:38 |
| *** Mudpuppy has quit IRC | 15:39 | |
| cloudnull | https://review.openstack.org/#/c/278597/ <-- for reference | 15:39 |
| cloudnull | busterswt: good looking out. | 15:39 |
| busterswt | cloudnull i was in the middle of building an environment to test :P | 15:39 |
| *** ric has quit IRC | 15:39 | |
| busterswt | time to work on the backport requests! | 15:40 |
| cloudnull | +1 | 15:40 |
| cloudnull | I think its reasonable for that to be backported to liberty | 15:40 |
| *** keedya has joined #openstack-ansible | 15:40 | |
| cloudnull | mattt: in the repo_build role i found that the linters were failing because bashate and flake8 were not installe.d | 15:42 |
| cloudnull | I updated the TEMPLATE repo to have those reqs in the test-reqs file | 15:43 |
| cloudnull | once i added them the tests were passing fine | 15:43 |
| *** Bofu2U has quit IRC | 15:43 | |
| mattt | cloudnull: good stuff | 15:44 |
| rromans | does the run-upgrade.sh script need to be run as root? | 15:45 |
| *** rparulek has quit IRC | 15:46 | |
| odyssey4me | cloudnull feedback in https://review.openstack.org/286625 | 15:48 |
| odyssey4me | rromans everything we implement currently assumes that you're running as root | 15:48 |
| rromans | odyssey4me: thanks | 15:49 |
| cloudnull | odyssey4me: i don’t thing they should be separate (which is why i didnt remove the fixes from the patch) the role cant work without those updates. | 15:49 |
| cloudnull | w/ a 5 hour lead time on the gate more patches is just more waiting | 15:49 |
| odyssey4me | cloudnull the role has no functional tests with the base config - please don't combine patches - as per the guidelines, each patch should be focused on doing one thing | 15:50 |
| odyssey4me | regardsless of the wait time, that's ok | 15:50 |
| cloudnull | however per the spec for IRR combining the patches is ok. | 15:50 |
| odyssey4me | it can be a follow on patch, and it will merge quickly after the base config patch assuming we get the votes on it in the same time frame | 15:50 |
| cloudnull | or it can go all in that one patch . we're just creating more work | 15:51 |
| odyssey4me | please focus the patches on solving one problem at a time | 15:51 |
| cloudnull | and the role cant work without it | 15:51 |
| cloudnull | thats the problem ^ | 15:51 |
| cloudnull | i know it has not functional tests, however the osa main repo is blocked because that patch didnt hit master due to other issues. | 15:52 |
| odyssey4me | the role will work just fine once both patches merge - the interim status of the role not working is not relevant right now | 15:52 |
| cloudnull | it is to me | 15:52 |
| odyssey4me | yes, and both patches will easily merge within minutes of each other | 15:52 |
| *** izaakk has joined #openstack-ansible | 15:53 | |
| mattt | we're going to have to rescan all these roles afterwards to ensure nothing has been missed, if the patches are separated it will be easier to tell if subsequent changes have gone in | 15:53 |
| cloudnull | are there any other patches for the repo_build role ? | 15:54 |
| * cloudnull knows there are not | 15:54 | |
| *** javeriak has quit IRC | 15:55 | |
| odyssey4me | cloudnull please respect other reviewers by adhering to the guidelines which have been set out by the project | 15:55 |
| *** javeriak has joined #openstack-ansible | 15:55 | |
| cloudnull | ^ WUT | 15:56 |
| cloudnull | we're in a mad dash to patch all of these role because we did all of them against what we were doing in the approved spec | 15:56 |
| *** rsparule has joined #openstack-ansible | 15:56 | |
| cloudnull | the spec that "other" reviewers voted on. | 15:57 |
| cloudnull | so its not about respect. its about getting things done. and separating things for the sake of separating things when there’s no benefit is pointless. | 15:58 |
| *** jiteka has quit IRC | 15:58 | |
| odyssey4me | it's pretty straightforward - I've voted based on the request to split the roles out... other reviewers may add their votes to the same review | 15:58 |
| odyssey4me | it's also about getting things done in a way that we can track and trace why things were done. By combining patches you're obscuring the reasons why the changes are being included and also making it hard for reviewers to review the other patches on the same topic to other repositories. | 15:59 |
| *** javeriak_ has joined #openstack-ansible | 15:59 | |
| *** javeriak has quit IRC | 16:00 | |
| cloudnull | i totally disagree. it wasn’t hard for people to review the infra or hosts IRR work done earlier and those were totally combined. | 16:00 |
| rsparule | Hi, I have a question on the 3 entries that are seen in this link hereby : https://github.com/openstack/openstack-ansible/blob/master/playbooks/defaults/repo_packages/openstack_services.yml#L81 | 16:01 |
| cloudnull | these are simple patches. we're just creating busy work | 16:01 |
| rsparule | what are the 2 entries that follow the git_repo parameter? | 16:01 |
| *** jiteka has joined #openstack-ansible | 16:01 | |
| rsparule | https://github.com/openstack/openstack-ansible/blob/master/playbooks/defaults/repo_packages/openstack_services.yml#L82 | 16:02 |
| rsparule | https://github.com/openstack/openstack-ansible/blob/master/playbooks/defaults/repo_packages/openstack_services.yml#L83 | 16:02 |
| *** Xu_Chao has quit IRC | 16:02 | |
| *** michaelgugino has joined #openstack-ansible | 16:02 | |
| *** bjoern_ has joined #openstack-ansible | 16:02 | |
| michaelgugino | hello all | 16:02 |
| cloudnull | also ive painstakingly made sure all of the history is preserved for these remaining roles so we can follow whats changed and when which was done because of our desire to "track whats been done". | 16:03 |
| cloudnull | o/ michaelgugino | 16:03 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server: Implement human readable logging for functional tests https://review.openstack.org/286591 | 16:04 |
| *** admin0 has quit IRC | 16:04 | |
| odyssey4me | o/ michaelgugino | 16:05 |
| odyssey4me | cloudnull not to track what's been done - but to track why things were done and who did it so that if we don't understand why from the commit then we can ask the author | 16:05 |
| cloudnull | all of that is in the commit message. | 16:06 |
| odyssey4me | rsparule the 'install_branch' parameter is the SHA (it could also be a tag/branch) which should be used as the basis for the service and its requirements | 16:07 |
| *** winggundamth has quit IRC | 16:07 | |
| *** Bjoern__ has joined #openstack-ansible | 16:07 | |
| *** ric has joined #openstack-ansible | 16:07 | |
| odyssey4me | rsparule the dest is used during the repo build as far as I recall - although honestly I'm a little fuzzy on that one withou doing some digging | 16:07 |
| *** bjoern_ has quit IRC | 16:08 | |
| cloudnull | my review has been submitted. and I'll be moving on to the next role. yes im being obstinate. i don't like more work for the sake of more work (especially in a case when the bits cant function without the changes proposed). | 16:08 |
| odyssey4me | rsparule if you wish to build an environment from a different SHA, tag or branch then edit the middle line and leave the rest alone :) if you wish to build from a forked repo then edit the first line (and probably the second too) | 16:08 |
| *** Bjoern__ is now known as bjoernt | 16:09 | |
| rsparule | <odyssey4me>: I want to provide a link to my own repository which holds packages | 16:09 |
| rsparule | <odyssey4me>: for this i assume i will need to provide all 3 lines in there | 16:09 |
| odyssey4me | rsparule what sort of packages? python wheels? | 16:09 |
| rsparule | <odyssey4me>: yes | 16:10 |
| rsparule | <odyssey4me>: also, what purpose does this parameter here serve? is it different from specifying repos? | 16:10 |
| rsparule | https://github.com/openstack/openstack-ansible/blob/master/playbooks/repo-build.yml#L42 | 16:11 |
| odyssey4me | rsparule no, you won't have to - you should just be able to configure extra links in the repo build process - you'll want the repo server to know about the extra source to use (assuming that pypi doesn't hold your packages) - and you'll want to the role (or user config) to specify the packages in question so that the repo build process actually cares to grab the wheels | 16:11 |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_cinder: Implement base configuration for independent repository https://review.openstack.org/286682 | 16:11 |
| rsparule | <odyssey4me>: so you suggest providing my local pypi mirror link here : https://github.com/openstack/openstack-ansible/blob/master/playbooks/repo-build.yml#L42 | 16:12 |
| odyssey4me | rsparule you can override pip_get_pip_options in /etc/openstack_deploy/user_variables.yml to include https://github.com/openstack/openstack-ansible/blob/master/playbooks/inventory/group_vars/hosts.yml#L61-L69 with the additional --find-links for your repo and the additional --trusted-host with your repo host name | 16:12 |
| odyssey4me | rsparule then you need to include an override in user_variables.yml for the role package list to include the defaults and your extras | 16:13 |
| cloudnull | tiagogomes: does it work on retry ? and how many forks are you running? | 16:13 |
| * cloudnull was reading back from a while ago | 16:13 | |
| tiagogomes | cloudnull yes it works on retry. I am using the default (8?) | 16:13 |
| cloudnull | bgmccollum and odyssey4me have some changes to sshd that may help. we used to have a retry function on ssh failure which was upstreamed, sadly we wont see that until we go to ansible 2 | 16:15 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-openstack_openrc: Implement base configuration for independent repository https://review.openstack.org/286623 | 16:16 |
| *** Mudpuppy has joined #openstack-ansible | 16:17 | |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_aodh: Implement base configuration for independent repository https://review.openstack.org/286655 | 16:17 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_glance: Implement base configuration for independent repository https://review.openstack.org/286667 | 16:18 |
| *** phiche has quit IRC | 16:19 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-repo_build: Implement base configuration for independent repository https://review.openstack.org/286625 | 16:19 |
| bjoernt | cloudnull: no bug triaging ? | 16:20 |
| cloudnull | odyssey4me: ^ | 16:20 |
| odyssey4me | heh, time ran away with me | 16:20 |
| odyssey4me | my apologies | 16:20 |
| bjoernt | BTW: https://review.openstack.org/#/c/284987/ when does it merge ? | 16:21 |
| automagically | bjoernt: Good catch, been a busy morning | 16:21 |
| rsparule | <odyssey4me>: just modifying these 2 parameters here would not help? i meant specifying my local pypi mirror link url in the 2 links here: | 16:21 |
| rsparule | https://github.com/openstack/openstack-ansible/blob/master/playbooks/repo-build.yml#L42 | 16:21 |
| rsparule | https://github.com/openstack/openstack-ansible/blob/master/playbooks/roles/repo_build/defaults/main.yml#L30 | 16:21 |
| bjoernt | it's blocks quite some commits which we would have to abandon | 16:21 |
| bjoernt | if it merges soon | 16:21 |
| odyssey4me | rsparule no, that is to build from a git source - not to fetch packages from pypi | 16:21 |
| *** Bofu2U has joined #openstack-ansible | 16:21 | |
| odyssey4me | ok, quick bug triage alll | 16:21 |
| rromans | \o | 16:22 |
| odyssey4me | bug triage cloudnull, mattt, andymccr, d34dh0r53, hughsaunders, b3rnard0, palendae, Sam-I-Am, odyssey4me, serverascode, rromans, erikmwilson, mancdaz, _shaps_, BjoernT, claco, echiu, dstanek, jwagner, ayoung, prometheanfire, evrardjp, arbrandes, mhayden, scarlisle, luckyinva, ntt, javeriak, automagically, spotz, vdo, jmccrory, alextricity25, jasondotstar, KLevenstein, admin0, michaelgugino, ametts, v1k0d3n, sever | 16:22 |
| odyssey4me | ion, bgmccollum | 16:22 |
| prometheanfire | hi | 16:22 |
| odyssey4me | reference: https://bugs.launchpad.net/openstack-ansible/+bugs?search=Search&field.status=New | 16:22 |
| spotz | o/ | 16:22 |
| automagically | o/ | 16:22 |
| michaelgugino | here | 16:22 |
| odyssey4me | first up https://bugs.launchpad.net/openstack-ansible/+bug/1542391 | 16:22 |
| jmccrory | o/ | 16:22 |
| openstack | Launchpad bug 1542391 in openstack-ansible "ceilometer swift pollster returns 403" [Undecided,New] | 16:22 |
| izaakk | o/ | 16:23 |
| odyssey4me | hmmm, I would have asked alextricity25 to chip in, but he's the reporter | 16:23 |
| odyssey4me | is there anyone that's happy to look into this and confirm the issue? | 16:23 |
| odyssey4me | we can leave it as 'new' and revisit it next week as I know we have other work in progress now | 16:24 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_swift: Implement base configuration for independent repository https://review.openstack.org/286692 | 16:24 |
| raddaoui | o/ | 16:24 |
| odyssey4me | next up https://bugs.launchpad.net/openstack-ansible/+bug/1544425 | 16:24 |
| openstack | Launchpad bug 1544425 in openstack-ansible "swift dispersion.conf auth version mismatch" [Undecided,New] - Assigned to Hector Gonzalez (theizaakk) | 16:24 |
| mattt | \o | 16:25 |
| tiagogomes | If I want to install the required packages from upstream, is there an easy way to accomplish that? | 16:25 |
| odyssey4me | hmm, it seems that the dispersion config there is v2 only anyway and needs to be updated to v3 | 16:25 |
| *** daneyon has joined #openstack-ansible | 16:25 | |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_heat: Implement base configuration for independent repository https://review.openstack.org/286693 | 16:25 |
| odyssey4me | izaakk I see that you've self assigned? | 16:25 |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_neutron: Implement base configuration for independent repository https://review.openstack.org/286694 | 16:25 |
| izaakk | odyssey4me: my bad, click by mistake | 16:26 |
| automagically | odyssey4me: I’ll take that one | 16:26 |
| *** daneyon_ has joined #openstack-ansible | 16:26 | |
| odyssey4me | automagically cool - I've added a note | 16:27 |
| odyssey4me | my proposal is that it's of medium importance, considering that the dispersion report is used in monitoring | 16:27 |
| openstackgerrit | Robb Romans proposed openstack/openstack-ansible: Add notes on running the upgrade script https://review.openstack.org/285419 | 16:27 |
| odyssey4me | anyone got an issue with that or have an alternative proposal? | 16:27 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_swift_sync: Implement base configuration for independent repository https://review.openstack.org/286696 | 16:28 |
| tiagogomes | If I want to install the required apt packages from upstream, is there an easy way to accomplish that? | 16:28 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_swift: Implement base configuration for independent repository https://review.openstack.org/286692 | 16:28 |
| tiagogomes | That is, I don't want to build an repo server | 16:28 |
| odyssey4me | tiagogomes what apt packages? | 16:28 |
| stevelle | medium seems fair | 16:28 |
| automagically | odyssey4me: +1 on medium | 16:28 |
| odyssey4me | tiagogomes not building a repo server is currently not a tested code path, so we're not really in a position to advise with any conviction | 16:29 |
| tiagogomes | Everyone used, including rabbit, galera and all openstack | 16:29 |
| tiagogomes | Ok. | 16:29 |
| cloudnull | tiagogomes: you'd need to change https://github.com/openstack/openstack-ansible/blob/master/playbooks/inventory/group_vars/hosts.yml#L38-L39 to the online repo | 16:29 |
| odyssey4me | next up https://bugs.launchpad.net/openstack-ansible/+bug/1544889 | 16:30 |
| openstack | Launchpad bug 1544889 in openstack-ansible "Conditionally set UID/GID for Nova user" [Undecided,New] | 16:30 |
| *** daneyon has quit IRC | 16:30 | |
| prometheanfire | the docs are going to be the harder part there | 16:30 |
| bjoernt | yeah that's a issue when using nfs for nova for example. So far we have not managed userids... | 16:30 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_aodh: Implement base configuration for independent repository https://review.openstack.org/286655 | 16:31 |
| odyssey4me | hmm, this looks to me like a wishlist item | 16:31 |
| bjoernt | it's a bug to me | 16:31 |
| odyssey4me | it seems valid to me as a desired request | 16:31 |
| cloudnull | ++ | 16:31 |
| bjoernt | since we need to make sure that the uids/gids are consistent | 16:31 |
| bjoernt | on all hosts | 16:31 |
| cloudnull | that seems like a good request | 16:31 |
| bjoernt | not so much which id | 16:31 |
| prometheanfire | this type of thing should probably be done with all users | 16:32 |
| odyssey4me | bjoernt there are many methods to make that happen which do not require inclusion in the project code | 16:32 |
| odyssey4me | there are easy workarounds too | 16:32 |
| cloudnull | I'd be awesome to spec out the UID/GIDs we want everything to be assigned and have that implemented | 16:32 |
| bjoernt | osa creates the users so we need to make sure that it is consistent | 16:32 |
| odyssey4me | so I do not see this as a bug in something we're doing, but rather a desire for improvement | 16:32 |
| bjoernt | at least that's my point on it | 16:32 |
| prometheanfire | start at 42424 as the uid and cout up, or something, use ldap, etc | 16:32 |
| odyssey4me | yes, I agree - but that puts it into the wishlist category IMO | 16:33 |
| cloudnull | ++ odyssey4me not a bug -- but would be a good wishlist item to work on | 16:33 |
| prometheanfire | many options | 16:33 |
| michaelgugino | ++not a bug | 16:33 |
| prometheanfire | wishlist++ | 16:33 |
| automagically | +1 on wishlist | 16:33 |
| michaelgugino | I like the idea of documenting the userid map | 16:33 |
| cloudnull | ^ +1 | 16:33 |
| *** fawadkhaliq has quit IRC | 16:33 | |
| prometheanfire | ya, this would be a larger spec/refacter then just nova (to do it right) | 16:34 |
| bjoernt | yes agreed, that should be the same for all users | 16:34 |
| cloudnull | we can codify those changes (should be fairly smallish/easy) too but i think it should be a spec. | 16:34 |
| bjoernt | perhaps a user_config option | 16:34 |
| odyssey4me | confirmed as a wishlist item - I expect that this could be looked at for the next cycle as it's a bit late in this cycle for new features | 16:34 |
| *** phiche1 has joined #openstack-ansible | 16:34 | |
| odyssey4me | and this meeting is not really a place to discuss how to implement new features either :) | 16:35 |
| bjoernt | jup | 16:35 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-openstack_openrc: Implement base configuration for independent repository https://review.openstack.org/286623 | 16:35 |
| odyssey4me | next up: https://bugs.launchpad.net/openstack-ansible/+bug/1544898 | 16:35 |
| openstack | Launchpad bug 1544898 in openstack-ansible "Host layout is missing the Object Storage" [Undecided,New] | 16:35 |
| odyssey4me | ah, this is a docs bug | 16:35 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_aodh: Implement base configuration for independent repository https://review.openstack.org/286655 | 16:35 |
| odyssey4me | it's also a hangover from the original docs structure | 16:36 |
| odyssey4me | we need to update the docs to cover the inclusion of swift better | 16:36 |
| odyssey4me | and perhaps also ceph and other options | 16:36 |
| odyssey4me | we had some discussion earlier today about relating the docs to the openstack guide reference architectures | 16:37 |
| odyssey4me | this would be great to do | 16:37 |
| odyssey4me | any volunteers to tackle this particular bug? | 16:37 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_heat: Implement base configuration for independent repository https://review.openstack.org/286693 | 16:37 |
| michaelgugino | I'll take it | 16:38 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_tempest: Implement base configuration for independent repository https://review.openstack.org/286701 | 16:38 |
| *** ric has quit IRC | 16:38 | |
| *** cloudtrainme has quit IRC | 16:38 | |
| odyssey4me | thanks michaelgugino | 16:38 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_glance: Implement base configuration for independent repository https://review.openstack.org/286667 | 16:38 |
| odyssey4me | next up https://bugs.launchpad.net/openstack-ansible/+bug/1545505 | 16:39 |
| openstack | Launchpad bug 1545505 in openstack-ansible "Repo stored version of wheel does not match version installed on repo containers" [Undecided,New] | 16:39 |
| *** phiche has joined #openstack-ansible | 16:39 | |
| *** phiche1 has quit IRC | 16:39 | |
| odyssey4me | ah, I'll take this one | 16:39 |
| jmccrory | https://review.openstack.org/#/c/284977/ ? | 16:39 |
| odyssey4me | there are some patches in progress around this | 16:39 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add missing runtime information to upgrade script https://review.openstack.org/286251 | 16:40 |
| odyssey4me | next up https://bugs.launchpad.net/openstack-ansible/+bug/1546225 | 16:40 |
| openstack | Launchpad bug 1546225 in openstack-ansible "run-upgrade script assumes rings are locally held for deployment " [Undecided,New] | 16:40 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add Ansible bootstrapping to Liberty upgrade https://review.openstack.org/286269 | 16:40 |
| *** weezS has joined #openstack-ansible | 16:40 | |
| *** mgoddard_ has quit IRC | 16:40 | |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add upgrade config change tools https://review.openstack.org/285564 | 16:40 |
| *** mgoddard has joined #openstack-ansible | 16:40 | |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Upgrades: Cleanup RabbitMQ / vhost https://review.openstack.org/272652 | 16:41 |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible-os_horizon: [WIP] Implement base configuration for independent repository https://review.openstack.org/286650 | 16:41 |
| odyssey4me | I see that bjoernt has already proposed a patch for that, assigning | 16:42 |
| palendae | Is that relevant for liberty, bjoernt? | 16:42 |
| bjoernt | no only kilo | 16:42 |
| odyssey4me | next up https://bugs.launchpad.net/openstack-ansible/+bug/1546796 | 16:42 |
| openstack | Launchpad bug 1546796 in openstack-ansible "sysstat.default.j2 should convert true and false to lower case" [Undecided,New] | 16:42 |
| palendae | I haven't gotten to doing swift stuff, not sure if rings or anything will need to be adjusted post upgrade | 16:42 |
| bjoernt | liberty we start working on the upgrade process so may be relevant but too early to say | 16:42 |
| cloudnull | bjoernt: in juno the rings we're stored on the deployment server as the default. | 16:42 |
| bjoernt | yeah in kilo it seem to have changed | 16:43 |
| palendae | odyssey4me: For that one izaakk did wor | 16:43 |
| palendae | work* | 16:43 |
| odyssey4me | this was covered in https://review.openstack.org/283337 right? | 16:43 |
| odyssey4me | so this is a duplicate | 16:43 |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible-os_horizon: [WIP] Implement base configuration for independent repository https://review.openstack.org/286650 | 16:44 |
| cloudnull | bjoernt: yes now they're on the swift nodes in kilo instead of the deployment node. so if the rings are not found when doing a juno to kilo upgrade you likely have other issues. | 16:44 |
| izaakk | yes, It's a duplicate | 16:44 |
| odyssey4me | next up https://bugs.launchpad.net/openstack-ansible/+bug/1546818 | 16:44 |
| openstack | Launchpad bug 1546818 in openstack-ansible "No ansible checks to ensure loaded ebtables module" [Undecided,New] - Assigned to Ala Raddaoui (raddaoui-ala) | 16:44 |
| palendae | raddaoui: ^ | 16:44 |
| raddaoui | yeah I will be working on it but I was trying to reproduce the error and I couldn't. | 16:44 |
| odyssey4me | andymccr and I chatted about this yesterday I think | 16:44 |
| raddaoui | I will check with the reporter | 16:44 |
| prometheanfire | raddaoui: you might be able to check sysfs to see if a value is reported or dir exists | 16:45 |
| prometheanfire | raddaoui: that's a better way of checking then lsmod or modinfo | 16:45 |
| odyssey4me | this is absolutely right, so I can confirm it | 16:45 |
| cloudnull | odyssey4me: seems like we need to add the ebtables module to https://github.com/openstack/openstack-ansible-openstack_hosts/blob/master/defaults/main.yml#L35 | 16:45 |
| odyssey4me | it only affects master, or liberty if you enable arp spoofing protection | 16:45 |
| raddaoui | ahah I unloaded the ebtables module but I didnt see any errors in neutron agents | 16:46 |
| odyssey4me | cloudnull wouldn't it be better to delegate the apt install of the package to whichever host hosts the neutron agents container? | 16:46 |
| odyssey4me | otherwise we're just installing it everywhere | 16:46 |
| odyssey4me | I suppose that's not necessarily a major issue | 16:46 |
| michaelgugino | I'd like to see something portable for the network host to load additional modules/packages | 16:47 |
| cloudnull | we are for all of those other modules too. | 16:47 |
| cloudnull | it'd be great to break that list down | 16:47 |
| cloudnull | and maybe move the functionality into the play | 16:47 |
| cloudnull | but putting it specifically in the role will cause problems when executed in the container (for some modules) | 16:48 |
| prometheanfire | zgrep /proc/config.gz | 16:48 |
| odyssey4me | I am thinking that we should look at some of this refactoring in the next cycle | 16:48 |
| prometheanfire | raddaoui: BRIDGE_NF_EBTABLES is what you are looking for right? | 16:49 |
| odyssey4me | but we can leave it to raddaoui to come up with a creative solution | 16:49 |
| odyssey4me | :) | 16:49 |
| cloudnull | +1 | 16:49 |
| odyssey4me | I've set it to medium priority - any alternative suggestions? | 16:49 |
| raddaoui | yes prometheanfire | 16:49 |
| prometheanfire | raddaoui: zgrep BRIDGE_NF_EBTABLES /proc/config.gz | 16:50 |
| odyssey4me | ok, next up https://bugs.launchpad.net/openstack-ansible/+bug/1547541 | 16:50 |
| openstack | Launchpad bug 1547541 in openstack-ansible "galera_client fails without apt-transport-https" [Undecided,New] | 16:50 |
| prometheanfire | just make sure that returns 0 | 16:50 |
| raddaoui | okay sounds cool | 16:50 |
| odyssey4me | hmm, we should either default to not using https - or we should ensure that we add the required package | 16:50 |
| cloudnull | odyssey4me: i think just about all of our apt repos require the https package | 16:51 |
| prometheanfire | raddaoui: make sure it ends with =y or =m, if =m then check via modinfo | 16:51 |
| automagically | question with that latest bug is do we support Ubuntu with tasksel minimal or do we presume server | 16:51 |
| prometheanfire | I'll add that to the bug | 16:51 |
| michaelgugino | I'm personally a fan of assuming minimal everywhere | 16:51 |
| raddaoui | okay prometheanfire | 16:51 |
| cloudnull | its a hard one to nail down. you have to have the apt-https-trans package to use https repos, but if you have https repos and do not already have that package apt will not work | 16:51 |
| michaelgugino | if the package is already on the host, no big deal. | 16:51 |
| odyssey4me | yeah, I prefer not requiring https if at all possible - for the default | 16:52 |
| michaelgugino | I think going away from https by default is not a great idea. That's the direction of the industry overall | 16:52 |
| bjoernt | afaik as I have tracked it down it is part of our trusty base image | 16:52 |
| automagically | michaelgugino: +1 | 16:52 |
| odyssey4me | that is a fair point | 16:52 |
| odyssey4me | bjoernt your comment is right, but not kinda off topic | 16:53 |
| odyssey4me | this is a comment with regards to the independent use of the repo for galera_client | 16:53 |
| bjoernt | this can be a chicken/egg thing where you need to have it installed to even install base packages | 16:53 |
| michaelgugino | there's always dpkg | 16:53 |
| cloudnull | I dont know that we've ever done a 100% minimal os. | 16:53 |
| michaelgugino | we're freezing packages anyway | 16:54 |
| odyssey4me | a simple option here could be to install that package if the repo includes https at the start? | 16:54 |
| bjoernt | lol base64 the package into the container | 16:54 |
| odyssey4me | the only reason this matters in this case is because we're adding a repo | 16:54 |
| cloudnull | we could download the package and dpkg install it as a separate task | 16:54 |
| odyssey4me | no, that's just crazy talk - we need this because were adding a repo | 16:54 |
| odyssey4me | the host should already have ubuntu repositories | 16:55 |
| cloudnull | but it is a chicken and egg issue | 16:55 |
| odyssey4me | so all we need to do is install the package from the ubuntu repository if the mariadb repo has https at the start of it | 16:55 |
| odyssey4me | how so? | 16:55 |
| cloudnull | if you add an https repo and dont have apt-https-tans installed apt will be broken | 16:55 |
| bjoernt | yes if the host points to SSL | 16:55 |
| odyssey4me | if the host uses https already then it will already have the right transport | 16:56 |
| odyssey4me | the ubuntu repo is outside of the scope of this role | 16:56 |
| stevelle | ^ | 16:56 |
| cloudnull | correct. however ubuntu doesnt install that package by default | 16:56 |
| *** fawadkhaliq has joined #openstack-ansible | 16:56 | |
| cloudnull | it outside the scope of this role | 16:56 |
| *** mikelk has quit IRC | 16:56 | |
| odyssey4me | if the mariadb repo uses the https transport, then install the package before adding the repo - it's pretty straightforward | 16:56 |
| michaelgugino | okay, I will back off the minimal server stance, but if it's in the upstream lxc container, then I say it's no problem | 16:57 |
| cloudnull | odyssey4me: there are other roles / repos that use https. we simply assume that package is there because ubuntu server includes. | 16:57 |
| odyssey4me | cloudnull the ubuntu server, and the base image, do not include this package - we add it | 16:58 |
| bjoernt | yes let's just put a target on the bug and work on it | 16:58 |
| odyssey4me | anyway, yes - any volunteers to work on this? | 16:58 |
| michaelgugino | if people are rolling their own container without packages, that's on them. But I think we should support linuxcontainers.org containers | 16:58 |
| cloudnull | an iso install of ubuntu server includes. | 16:58 |
| *** mgoddard_ has joined #openstack-ansible | 16:58 | |
| odyssey4me | next up https://bugs.launchpad.net/openstack-ansible/+bug/1547598 | 16:58 |
| openstack | Launchpad bug 1547598 in openstack-ansible "install-guide define contents of /etc/network/interfaces for both the controller node and compute node that differ" [Undecided,New] | 16:58 |
| cloudnull | did we finish the last one ? | 16:59 |
| odyssey4me | cloudnull it's set to low priority and can be picked up by anyone inclined | 16:59 |
| cloudnull | ok | 16:59 |
| odyssey4me | we're out of time anyway | 17:00 |
| michaelgugino | we did start 22 minutes late ;) | 17:00 |
| cloudnull | ^ that | 17:00 |
| odyssey4me | lol, ok - is everyone good to continue? | 17:00 |
| bjoernt | +1 | 17:00 |
| cloudnull | +1 | 17:00 |
| michaelgugino | +1 | 17:00 |
| mattt | -1, need to leave momentarily :) | 17:01 |
| odyssey4me | alright, we'll go with whoever can make it | 17:01 |
| odyssey4me | https://bugs.launchpad.net/openstack-ansible/+bug/1547598 | 17:01 |
| openstack | Launchpad bug 1547598 in openstack-ansible "install-guide define contents of /etc/network/interfaces for both the controller node and compute node that differ" [Undecided,New] | 17:01 |
| cloudnull | doc issue, low IMHO . | 17:01 |
| odyssey4me | this is a docs bug, and is good commentary on the consistency of the work | 17:01 |
| *** mgoddard has quit IRC | 17:01 | |
| odyssey4me | I would really like to see us getting rid of two sources of truth (docs and .examples files) and move towards just docs | 17:02 |
| odyssey4me | this is a consistent headache | 17:02 |
| cloudnull | +1 | 17:02 |
| *** jmccrory_ has joined #openstack-ansible | 17:02 | |
| cloudnull | the .example files are very much an openstack-ism but i think clearer docs are better for the project. | 17:03 |
| automagically | +1 | 17:03 |
| odyssey4me | any volunteers to pick this up? | 17:03 |
| izaakk | o/ | 17:03 |
| odyssey4me | great, thanks izaakk | 17:03 |
| odyssey4me | izaakk can you self assign it please? | 17:03 |
| odyssey4me | next up https://bugs.launchpad.net/openstack-ansible/+bug/1547679 | 17:04 |
| openstack | Launchpad bug 1547679 in openstack-ansible "config_template.py dies when using host_vars inside openstack_user_config" [Undecided,New] | 17:04 |
| izaakk | odyssey4me: done | 17:04 |
| odyssey4me | thanks izaakk | 17:04 |
| odyssey4me | this bug looks to be kilo only, due to a backport not being present as jmccrory has pointed out | 17:05 |
| *** ntpttr has joined #openstack-ansible | 17:05 | |
| cloudnull | not a bug | 17:05 |
| cloudnull | fixed in later versions of kilo | 17:05 |
| odyssey4me | ah yes - commented and marked invalid | 17:06 |
| odyssey4me | https://bugs.launchpad.net/openstack-ansible/+bug/1550040 | 17:07 |
| openstack | Launchpad bug 1550040 in openstack-ansible "Capther 4, configure Keystone missing identity_hosts" [Undecided,New] | 17:07 |
| *** pcaruana has quit IRC | 17:07 | |
| ntpttr | Hey everyone - after running haproxy-install.yml should I be able to ping the external_lb_vip? I'm running that playbook and it's completing successfully, but then the installation is failing when galera tries to install pip packages and needs to connect to the load balancer to find them | 17:08 |
| odyssey4me | two docs bugs https://bugs.launchpad.net/openstack-ansible/+bug/1550040 and https://bugs.launchpad.net/openstack-ansible/+bug/1550041 | 17:08 |
| openstack | Launchpad bug 1550040 in openstack-ansible "Chapter 4, configure Keystone missing identity_hosts" [Undecided,New] | 17:08 |
| openstack | Launchpad bug 1550041 in openstack-ansible "Chapter 4, configure Block Storage missing storage-infra_hosts" [Undecided,New] | 17:08 |
| odyssey4me | can anyone validate the statements there? I'm not familiar enough with that stuff without hunting | 17:09 |
| cloudnull | infra_hosts is a catch all | 17:09 |
| cloudnull | identity_hosts is only for keystone | 17:09 |
| *** ametts has joined #openstack-ansible | 17:10 | |
| odyssey4me | it should be either/or rather than both, right? | 17:10 |
| cloudnull | i'd like us to not use infra_hosts. | 17:10 |
| cloudnull | its really only their for legacy compat | 17:10 |
| michaelgugino | I don't see a problem with storage_hosts. But, I'm not versed in that aspect of deploys | 17:11 |
| cloudnull | odyssey4me: +1 not both. | 17:11 |
| michaelgugino | those bug reports are not really well versed | 17:11 |
| cloudnull | I mean it'll result in the same thing, however the later provides more control. | 17:11 |
| odyssey4me | I've added a comment to the keystone one - any volunteers for picking up the documentation update to be clearer? | 17:12 |
| odyssey4me | spotz / mhayden ? :) | 17:12 |
| spotz | odyssey4me I think your note on 1550040 is wrong | 17:12 |
| spotz | I'll take it if you make the note clear:) | 17:13 |
| odyssey4me | spotz it may very well be - whoever picks it up will need to discuss the best information with the channel and come up with the best solution | 17:13 |
| spotz | You said to remove and add the same thing | 17:13 |
| odyssey4me | spotz my advice is to use identity_hosts for keystone in the documentation | 17:14 |
| odyssey4me | but I don't know what I'm talking about! :D | 17:14 |
| spotz | I picked it up, I think it's remove infra and add identity_hosts | 17:14 |
| prometheanfire | odyssey4me: ping for when triage is done | 17:14 |
| odyssey4me | spotz so that's a yes on self assigning the bug then? | 17:14 |
| spotz | yep already took it | 17:14 |
| odyssey4me | aweosme, thanks | 17:15 |
| odyssey4me | next up https://bugs.launchpad.net/openstack-ansible/+bug/1550041 ? | 17:15 |
| openstack | Launchpad bug 1550041 in openstack-ansible "Chapter 4, configure Block Storage missing storage-infra_hosts" [Undecided,New] | 17:15 |
| odyssey4me | another docs bug | 17:15 |
| odyssey4me | I guess that storage-infra_hosts is specifically for hosting the cinder-api service? | 17:16 |
| *** tiagogomes has quit IRC | 17:16 | |
| *** tiagogomes has joined #openstack-ansible | 17:16 | |
| bgmccollum | odyssey4me correct | 17:16 |
| *** kysse_ has quit IRC | 17:17 | |
| odyssey4me | if someone can spend a little time providing some accurate info in the bug, then I'm sure someone will jump in to patch it | 17:17 |
| odyssey4me | but the inventory is a bit of a dark art to most people | 17:17 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_swift: Implement base configuration for independent repository https://review.openstack.org/286692 | 17:18 |
| *** raddaoui has quit IRC | 17:18 | |
| automagically | dark art indeed | 17:18 |
| odyssey4me | are my comments in https://bugs.launchpad.net/openstack-ansible/+bug/1550041 accurate? | 17:19 |
| openstack | Launchpad bug 1550041 in openstack-ansible "Chapter 4, configure Block Storage missing storage-infra_hosts" [Undecided,New] | 17:19 |
| odyssey4me | clear as mud? | 17:19 |
| *** raddaoui has joined #openstack-ansible | 17:19 | |
| michaelgugino | the first bit is clear, then you muddy'd it up | 17:20 |
| odyssey4me | heh, of course | 17:20 |
| spotz | I think I can work with it. I'll grab it and pester on channel | 17:21 |
| odyssey4me | thanks spotz | 17:21 |
| odyssey4me | next up https://bugs.launchpad.net/openstack-ansible/+bug/1550044 | 17:21 |
| openstack | Launchpad bug 1550044 in openstack-ansible "Chapter 2. Deployment host missing network information" [Undecided,New] | 17:21 |
| cloudnull | I think most of the networking section of the documentation needs much love | 17:22 |
| odyssey4me | I think that it's implying that the deployment host needs to be on the same network? | 17:22 |
| odyssey4me | if so, that would definitely be a recommendation to limit failures. | 17:22 |
| git-harry | https://review.openstack.org/#/q/topic:liberty_upgrade is a list of patches for liberty upgrades. If any cores get time please review. | 17:22 |
| automagically | same network or network acls permit traffic to transit between deployment host network and the container mgmt vlan | 17:23 |
| odyssey4me | please add additional comments to the bug - I've added some | 17:24 |
| *** ric has joined #openstack-ansible | 17:24 | |
| *** admin0 has joined #openstack-ansible | 17:24 | |
| odyssey4me | automagically you are correct :) | 17:24 |
| * automagically commented | 17:24 | |
| odyssey4me | spotz keen for another one? unless izaakk wants to pick that one up? | 17:24 |
| spotz | let me read the notes:) | 17:25 |
| *** raddaoui has quit IRC | 17:25 | |
| odyssey4me | fair point automagically - it could just be noted as a prerequisite | 17:25 |
| michaelgugino | I've never cared for calling it the container management vlan | 17:25 |
| spotz | Yeah I'll pick that fix up, just not re-writing networking:) | 17:26 |
| automagically | michaelgugino: Comment on the bug | 17:26 |
| odyssey4me | it should just be the management network I would guess - but that's a topic of much bike shedding which I thought we had worked out | 17:26 |
| odyssey4me | last one up https://bugs.launchpad.net/openstack-ansible/+bug/1550442 | 17:26 |
| openstack | Launchpad bug 1550442 in openstack-ansible "V-38496 check includes non-system accounts" [Undecided,New] - Assigned to Major Hayden (rackerhacker) | 17:26 |
| michaelgugino | k | 17:27 |
| odyssey4me | I see that mhayden has picked this up. Any comments would be useful to him I'm sure. | 17:27 |
| odyssey4me | thank you all for taking the time, and extra time! | 17:27 |
| cloudnull | cheers | 17:28 |
| admin0 | back .. did i miss a big fight :D ? | 17:29 |
| *** kstepniewski has joined #openstack-ansible | 17:31 | |
| *** cloudtrainme has joined #openstack-ansible | 17:32 | |
| *** cloudtrainme has quit IRC | 17:33 | |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible-os_horizon: [WIP] Implement base configuration for independent repository https://review.openstack.org/286650 | 17:36 |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_cinder: Implement base configuration for independent repository https://review.openstack.org/286682 | 17:38 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_nova: Implement base configuration for independent repository https://review.openstack.org/286736 | 17:40 |
| *** jwagner is now known as jwagner_lunch | 17:41 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_ceilometer: Implement base configuration for independent repository https://review.openstack.org/286674 | 17:44 |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_neutron: Implement base configuration for independent repository https://review.openstack.org/286694 | 17:46 |
| *** ntpttr has left #openstack-ansible | 17:47 | |
| *** Bjoern_ has joined #openstack-ansible | 17:48 | |
| *** michaelgugino has quit IRC | 17:49 | |
| cloudnull | git-harry palendae what would you like to see in the commit message for https://review.openstack.org/#/c/271591 that's otherwise unclear? | 17:49 |
| cloudnull | in regards to the changes in defaults ? | 17:50 |
| *** bjoernt has quit IRC | 17:50 | |
| palendae | cloudnull: https://review.openstack.org/#/c/271591/4/playbooks/roles/galera_server/defaults/main.yml anything for the wsrep addresses. The galera_upgrade I get | 17:50 |
| cloudnull | i mean whats unclear about their purpose ? | 17:51 |
| palendae | Mostly not sure why they're necessary for the major version upgrade to work | 17:51 |
| *** sdake_ has quit IRC | 17:51 | |
| cloudnull | if the upgrade fails, or the cluster for that matter, youre going to want the ability to set the cluster node name, address and the like to recover it | 17:51 |
| palendae | And that's not doable with the previos values? | 17:52 |
| cloudnull | i can pull those changes out, but they seem like things you'd want. | 17:52 |
| palendae | Sure, what I think we're looking for is "why would we want them" | 17:53 |
| palendae | Looking at the review, I do not know the why | 17:53 |
| palendae | I'm not asking for it to be pulled out necessarily | 17:53 |
| *** Bjoern_ has quit IRC | 17:53 | |
| cloudnull | ill yank those updates in the next review | 17:57 |
| odyssey4me | cloudnull the ask has always been to explain the 'why' - not necessarily to yank them | 17:57 |
| odyssey4me | if they are not appropriate to the focal point of the patch, then yank them - if they are, then explain why they are | 17:58 |
| odyssey4me | the issue at hand is that the patch does not match the master patch, so therefore the reviewer has to re-ask the question 'why' | 17:58 |
| *** ametts has quit IRC | 17:59 | |
| cloudnull | i think the "why" is self evident. you now know that you can set the wsrep nodename, address, and cluster addresses. which are useful in recovery situations. | 18:01 |
| cloudnull | if you think that blub makes the commit message clearer ill add it. | 18:01 |
| *** weezS has quit IRC | 18:02 | |
| palendae | cloudnull: 2 people don't think it's self evident though | 18:02 |
| cloudnull | this is true | 18:02 |
| palendae | What you typed right there is enough | 18:02 |
| palendae | For me | 18:02 |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-rabbitmq_server: [WIP] Multi-distro support for rabbitmq-server role https://review.openstack.org/286282 | 18:03 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server: Force initial apt-get update as part of test execution https://review.openstack.org/286648 | 18:06 |
| *** busterswt has quit IRC | 18:07 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_server: Force initial apt-get update as part of test execution https://review.openstack.org/286648 | 18:07 |
| *** busterswt has joined #openstack-ansible | 18:07 | |
| odyssey4me | automagically ^ you're right - I missed that bit - hopefully this resolves it | 18:07 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: Added major version upgrade support (galera) https://review.openstack.org/271591 | 18:09 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add configuration migration for Liberty https://review.openstack.org/286753 | 18:10 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add aodh, ceilometer, and neturon config upgrades https://review.openstack.org/286753 | 18:11 |
| *** openstackgerrit has quit IRC | 18:18 | |
| *** openstackgerrit has joined #openstack-ansible | 18:18 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_swift: Implement base configuration for independent repository https://review.openstack.org/286692 | 18:19 |
| *** thorst_ has joined #openstack-ansible | 18:19 | |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add aodh, ceilometer, and neturon config upgrades https://review.openstack.org/286753 | 18:19 |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible-lxc_hosts: Removing unneeded with_items usage for clarity https://review.openstack.org/286758 | 18:21 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_swift_sync: Implement base configuration for independent repository https://review.openstack.org/286696 | 18:21 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add aodh, ceilometer, and neturon config upgrades https://review.openstack.org/286753 | 18:21 |
| prometheanfire | odyssey4me: what do you think of this way to check for modules/kernel support of things, useful for multi-os. https://bugs.launchpad.net/openstack-ansible/+bug/1546818 | 18:22 |
| openstack | Launchpad bug 1546818 in openstack-ansible "No ansible checks to ensure loaded ebtables module" [Medium,Confirmed] - Assigned to Ala Raddaoui (raddaoui-ala) | 18:22 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add aodh, haproxy, and neturon config upgrades https://review.openstack.org/286753 | 18:22 |
| *** raddaoui has joined #openstack-ansible | 18:22 | |
| *** weezS has joined #openstack-ansible | 18:23 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-repo_build: Implement base configuration for independent repository https://review.openstack.org/286625 | 18:24 |
| openstackgerrit | Merged openstack/openstack-ansible: Automate setting of glance_api_version https://review.openstack.org/284182 | 18:25 |
| admin0 | is today a special day or a very busy day for openstackgerrit | 18:26 |
| admin0 | looks like a floodgate of new changes | 18:26 |
| *** serverascode has quit IRC | 18:26 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_nova: Implement base configuration for independent repository https://review.openstack.org/286736 | 18:26 |
| *** phiche has quit IRC | 18:27 | |
| automagically | admin0: There is a bug push today to get testing/gate squared away for all of the individual role repos | 18:27 |
| *** weshay has quit IRC | 18:27 | |
| admin0 | ha .. that went above my head :D | 18:27 |
| automagically | admin0: https://github.com/openstack/openstack-ansible-specs/blob/master/specs/mitaka/independent-role-repositories.rst | 18:28 |
| automagically | s/bug/big | 18:28 |
| admin0 | oh | 18:28 |
| admin0 | this is a good thing | 18:28 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_tempest: Implement base configuration for independent repository https://review.openstack.org/286701 | 18:29 |
| *** permalac has quit IRC | 18:31 | |
| *** serverascode has joined #openstack-ansible | 18:32 | |
| *** permalac has joined #openstack-ansible | 18:34 | |
| *** permalac has quit IRC | 18:34 | |
| *** phiche has joined #openstack-ansible | 18:41 | |
| *** jwitko has joined #openstack-ansible | 18:44 | |
| *** phiche1 has joined #openstack-ansible | 18:44 | |
| cloudnull | mattt: you around? looks like the additional role requirements file and entries for the os_ role needs to be craeted/updated | 18:45 |
| cloudnull | odyssey4me: also on the IRR roles it seems that we're going to have to add the apt-cache update change | 18:45 |
| cloudnull | http://logs.openstack.org/60/286660/1/check/gate-openstack-ansible-os_keystone-dsvm-ansible-functional-trusty/7932786/console.html#_2016-03-01_16_23_32_671 | 18:45 |
| cloudnull | it might be wise to do an "apt-get clean && apt-get update" instead of the cache update task | 18:45 |
| cloudnull | the apt module has a few issues in it that could cause the cache to not be updated and wont be fixed until we go to Ansible 2 | 18:46 |
| *** phiche has quit IRC | 18:48 | |
| *** Mudpuppy has quit IRC | 18:48 | |
| *** itsuugo has quit IRC | 18:50 | |
| *** Bofu2U has quit IRC | 18:52 | |
| *** furlongm_ has quit IRC | 18:52 | |
| *** finchd has quit IRC | 18:52 | |
| *** furlongm_ has joined #openstack-ansible | 18:54 | |
| *** openstack has joined #openstack-ansible | 19:17 | |
| cloudnull | for sure. | 19:18 |
| *** openstackgerrit has joined #openstack-ansible | 19:19 | |
| cloudnull | so this merged https://review.openstack.org/#/c/286250/ (thanks for the review folks) now we need to make a big push to watch for var changes and ensure we're using the new deprecated filter. | 19:22 |
| *** maximov_ has quit IRC | 19:25 | |
| *** bryan_att has quit IRC | 19:25 | |
| *** errr has quit IRC | 19:25 | |
| *** alextricity25 has quit IRC | 19:25 | |
| *** odyssey4me has quit IRC | 19:25 | |
| *** lane_kong has quit IRC | 19:25 | |
| *** mhayden has quit IRC | 19:25 | |
| *** git-harry has quit IRC | 19:25 | |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible-os_horizon: [WIP] Implement base configuration for independent repository https://review.openstack.org/286650 | 19:26 |
| -cloudnull- please watch the following reviews and provide feedback on them so that we can get the latest batch of IRR moving forward https://review.openstack.org/#/q/status:open+topic:bp/independent-role-repositories | 19:26 | |
| automagically | cloudnull: Love that filter! | 19:26 |
| cloudnull | sweet! i think itll be real useful | 19:27 |
| spotz | pretty | 19:28 |
| *** jmccrory_ has quit IRC | 19:28 | |
| *** git-harry has joined #openstack-ansible | 19:31 | |
| *** maximov_ has joined #openstack-ansible | 19:31 | |
| *** bryan_att has joined #openstack-ansible | 19:31 | |
| *** errr has joined #openstack-ansible | 19:31 | |
| *** lane_kong has joined #openstack-ansible | 19:31 | |
| *** alextricity25 has joined #openstack-ansible | 19:31 | |
| *** odyssey4me has joined #openstack-ansible | 19:31 | |
| *** mhayden has joined #openstack-ansible | 19:31 | |
| *** mhayden has quit IRC | 19:31 | |
| *** mhayden has joined #openstack-ansible | 19:32 | |
| *** Mudpuppy has joined #openstack-ansible | 19:32 | |
| cloudnull | just checked https://github.com/ansible/ansible-modules-core/pull/1517 -- seems that its still not merged so apt cache update issue are still present in ansible at this point. | 19:34 |
| cloudnull | waiting on "CORE" review | 19:34 |
| *** evrardjp has joined #openstack-ansible | 19:35 | |
| *** evrardjp has left #openstack-ansible | 19:35 | |
| cloudnull | hopefully we can get https://github.com/ansible/ansible/pull/12555 in for 2.1 | 19:37 |
| *** sigmavirus24 is now known as sigmavirus24_awa | 19:37 | |
| *** phiche1 has joined #openstack-ansible | 19:43 | |
| *** phiche has quit IRC | 19:46 | |
| cloudnull | automagically: seems like https://review.openstack.org/#/c/286648/ worked | 19:49 |
| automagically | Awesome, let the mass updating of patchs begin ;) | 19:50 |
| cloudnull | yup :) | 19:50 |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible: [WIP] Enable SSL termination for all services https://review.openstack.org/277199 | 19:51 |
| *** sigmavirus24_awa is now known as sigmavirus24 | 19:52 | |
| cloudnull | if we can get https://review.openstack.org/#/c/286648 in we can unblock the rest of the gate | 19:52 |
| *** weezS has quit IRC | 19:52 | |
| stevelle | looking already | 19:52 |
| cloudnull | scratch that... we need https://review.openstack.org/#/c/284896/ and https://review.openstack.org/#/c/286625/ too | 19:53 |
| *** rromans has quit IRC | 19:54 | |
| admin0 | glad i did not start developing earlier :D | 19:54 |
| admin0 | will do after the gate is fixed:D | 19:54 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible: Implements IRR for the repo_build role https://review.openstack.org/284896 | 19:54 |
| *** raddaoui has joined #openstack-ansible | 19:55 | |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible-os_keystone: [WIP] Implement base configuration for independent repository https://review.openstack.org/286660 | 19:55 |
| admin0 | what does a gate mean anyway? is it a CI/CD term ? | 19:55 |
| *** rromans has joined #openstack-ansible | 19:55 | |
| automagically | admin0: http://docs.openstack.org/infra/zuul/gating.html | 19:56 |
| stevelle | cloudnull: need to add any dependencies or is it just that the critical queue is the sum of the 3? | 19:57 |
| admin0 | thanks automagically | 19:58 |
| admin0 | http://status.openstack.org/zuul/ \o/ | 19:58 |
| *** raddaoui has quit IRC | 19:59 | |
| cloudnull | stevelle: its just those three. | 19:59 |
| cloudnull | the repo-server and build patches go hand in hand. | 19:59 |
| stevelle | automagically: just hit the button on the first one before me | 20:00 |
| automagically | D’oh! | 20:00 |
| stevelle | s'ok. more cores, more speed | 20:00 |
| automagically | We’re going to need hyperthreading here soon | 20:00 |
| *** weezS has joined #openstack-ansible | 20:02 | |
| palendae | Debugging straight openstack-ansible gate failures is gonna be a nightmare from here on out | 20:02 |
| jwitko | hey cloudnull, I am having a lot of trouble running the first playbook for the openstack install. I'm running the kilo branch 11.2.9 release tag. It seems to only be running against two out of my 5 hosts for the initial cluster. and one of those hosts keeps getting "connection timed out", but I can SSH to the host manually without any issues. | 20:02 |
| jwitko | the playbook looks to only be running against my logging host and my single compute host | 20:02 |
| cloudnull | can you limit it or use less forks ? and does it die randomly ? | 20:03 |
| cloudnull | IE on different hots | 20:03 |
| jwitko | sorry not sure I understand... ? | 20:03 |
| cloudnull | is it failing on the same hosts all of the time ? | 20:04 |
| jwitko | yes | 20:04 |
| cloudnull | hum. | 20:04 |
| jwitko | but the host can be accessed by SSH without issue from the command line | 20:04 |
| jwitko | the host is also identical in every way to the host in which it succeeds | 20:04 |
| jwitko | (fresh ubuntu 14.04 release, same ansible run on the hosts to prep them for openstack) | 20:04 |
| *** sdake has quit IRC | 20:05 | |
| *** javeriak has joined #openstack-ansible | 20:05 | |
| jwitko | but more concerning is why is the playbook not running against any of the controller hosts ? | 20:05 |
| cloudnull | are not all of the hosts in the openstack_user_config file ? | 20:06 |
| openstackgerrit | Merged openstack/openstack-ansible-repo_server: Force initial apt-get update as part of test execution https://review.openstack.org/286648 | 20:06 |
| jwitko | cloudnull, they are all in there. I can paste the config file one second | 20:06 |
| cloudnull | and which play are you running ? | 20:06 |
| jwitko | the initial setup-hosts | 20:06 |
| cloudnull | and can you test with --list-hosts to see if the inventory is showing all of the hosts ? | 20:07 |
| *** ric has quit IRC | 20:08 | |
| jwitko | good call I'll try that now | 20:08 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_aodh: Implement base configuration for independent repository https://review.openstack.org/286655 | 20:08 |
| *** javeriak_ has quit IRC | 20:08 | |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_glance: Implement base configuration for independent repository https://review.openstack.org/286667 | 20:10 |
| jwitko | ugh, also every time I run it I have to delete the generated files in /etc/openstack_deploy otherwise it errors | 20:10 |
| jwitko | http://paste.openstack.org/show/488835/ | 20:10 |
| *** phiche1 has quit IRC | 20:10 | |
| *** phiche has joined #openstack-ansible | 20:11 | |
| palendae | That sounds wrong, almost like a wrong python version | 20:12 |
| *** javeriak has quit IRC | 20:12 | |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_glance: Implement base configuration for independent repository https://review.openstack.org/286667 | 20:14 |
| jwitko | cloudnull, indeed the host list is not generating properly | 20:14 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_aodh: Implement base configuration for independent repository https://review.openstack.org/286655 | 20:14 |
| jwitko | its only generating for the compute node and the log host | 20:14 |
| *** javeriak has joined #openstack-ansible | 20:15 | |
| *** sdake has joined #openstack-ansible | 20:16 | |
| automagically | Can I get some core eyes on https://review.openstack.org/#/c/284931 please? | 20:17 |
| *** javeriak has quit IRC | 20:18 | |
| *** raddaoui has joined #openstack-ansible | 20:18 | |
| jwitko | syntax check comes back OK... | 20:20 |
| stevelle | automagically: mind adding a release note into the main repo with a Depends-On for that? | 20:21 |
| cloudnull | jwitko: maybe an indentation issue in the openstack_user_config file ? | 20:21 |
| palendae | jwitko: What version of python are you using? I've never seen it complain about the tarfile object not having an __exit__ method | 20:22 |
| automagically | stevelle: Will do | 20:22 |
| jwitko | palendae, 2.7.9 | 20:24 |
| palendae | Huh, ok | 20:24 |
| jwitko | cloudnull, I just triple checked the user config file and all identation looks good and matching | 20:24 |
| jwitko | redacting and pasting config | 20:26 |
| jwitko | one moment | 20:26 |
| jwitko | cloudnull, so here is my conifg. | 20:30 |
| jwitko | http://paste.openstack.org/raw/488838/ | 20:30 |
| *** sdake has quit IRC | 20:31 | |
| * cloudnull looking | 20:32 | |
| *** phalmos has quit IRC | 20:34 | |
| *** flwang has quit IRC | 20:37 | |
| izaakk | Hello guys, quick question on https://bugs.launchpad.net/openstack-ansible/+bug/1547598 | 20:37 |
| openstack | Launchpad bug 1547598 in openstack-ansible "install-guide define contents of /etc/network/interfaces for both the controller node and compute node that differ" [Low,Confirmed] - Assigned to Hector Gonzalez (theizaakk) | 20:37 |
| izaakk | should we remove the example file? or just Remove any consistency issue and leave it the same as Docs | 20:37 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Update ceilometer environment configuration https://review.openstack.org/286827 | 20:38 |
| *** alejandrito has joined #openstack-ansible | 20:38 | |
| cloudnull | jwitko: is the play hitting oss-ctrl0{1,3} and oss-log01 ? | 20:39 |
| jwitko | cloudnull, i'm thinking it might be something with the IPs? I noticed when I was redacting that file for you that I had a couple IPs incorrect | 20:39 |
| jwitko | it was actually ONLY hitting the incorrect IPs | 20:39 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add aodh, haproxy, and neturon config upgrades https://review.openstack.org/286753 | 20:40 |
| cloudnull | ah . | 20:40 |
| cloudnull | thatll do it | 20:40 |
| jwitko | cloudnull, since I fixed the IPs it is actually hitting NO hosts anymore | 20:40 |
| jwitko | but I don't see myself blocking the IPs in the used_ips block | 20:40 |
| jwitko | so where am I doing this wrong ? | 20:40 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Update ceilometer environment configuration https://review.openstack.org/286827 | 20:40 |
| cloudnull | nuke the openstack_inventory.json file and rerun | 20:41 |
| cloudnull | maybe it was busted from before and having now fixed it up everything will generate correctly ? | 20:41 |
| jwitko | no hosts matched | 20:41 |
| openstackgerrit | Merged openstack/openstack-ansible-lxc_hosts: Allow cert validation to be toggled by deployers https://review.openstack.org/284931 | 20:42 |
| jwitko | cloudnull, after blowing them away and rerunning setup-hosts it does generate a file with the correct IPs in openstack_hostnames_ips.yml | 20:42 |
| jwitko | but it still says "no hosts matched" | 20:42 |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible: Release note for LXC download cert validation https://review.openstack.org/286830 | 20:43 |
| *** sdake has joined #openstack-ansible | 20:43 | |
| palendae | izaakk: My vote would be to make sure they match. | 20:44 |
| *** pcaruana has joined #openstack-ansible | 20:44 | |
| palendae | But it's not clear if Jesse's asking for the example file to be deleted there, or as something to be done later. | 20:44 |
| cloudnull | jwitko: so setup-hosts call openstack-host-setup | 20:45 |
| jwitko | aye | 20:45 |
| jwitko | and that runs against "hosts" | 20:45 |
| cloudnull | can you call that play directly | 20:45 |
| *** jamielennox|away is now known as jamielennox | 20:45 | |
| cloudnull | thats where i was going. seems like youve been down this road :) | 20:45 |
| automagically | palendae: and izaakk - During bug triage there seemed to be consensus that maintaining the example files in favor of docs was a losing proposition | 20:45 |
| jwitko | cloudnull, still same result, no hosts matched | 20:45 |
| jwitko | so for one reason or another my inventory json is not being properly generated? | 20:46 |
| palendae | automagically: I figured. I think that's silly given some of our upgrade mechanisms rely on them | 20:46 |
| palendae | Also it's been demonstrated that the docs in isolation cause misunderstandings | 20:46 |
| automagically | I think the concern for the two falling out of sync or contradicting each other is very valid | 20:46 |
| cloudnull | jwitko: thats what it sounds like | 20:46 |
| palendae | I don't disagree | 20:46 |
| automagically | Not sure that removing the example files is the necessary or advisable solution | 20:47 |
| jwitko | cloudnull, but looking at my config I have no idea what it could be ? | 20:47 |
| jwitko | I don't see anything wrong with my config | 20:47 |
| cloudnull | im going to kick somethings with your config and see | 20:47 |
| jwitko | thank you! | 20:47 |
| palendae | automagically: Agreed there | 20:47 |
| cloudnull | do you have the files in the env.d directory by chance ? | 20:47 |
| jwitko | cloudnull, yes 3 files | 20:47 |
| cloudnull | these files https://github.com/openstack/openstack-ansible/tree/kilo/etc/openstack_deploy/env.d ? | 20:48 |
| jwitko | cloudnull, no. in my previous deploys on 11.2.4 version I did not need those, only the extra containers I wanted | 20:48 |
| izaakk | palendae, automagically : Agree, so will just make them match. I just got confused with the comments on the bug | 20:48 |
| jwitko | is that my problem? those are now required? | 20:48 |
| palendae | izaakk: If that's not what he wants, I'm sure he'll say on reviews | 20:48 |
| palendae | Not sure about that specific example file, but I know a lot of our variables examples files are used by the upgrade scripts | 20:49 |
| palendae | How Mitaka will reconcile that with the IRRs, I have no clue | 20:49 |
| cloudnull | jwitko: in 11.2.4 we have these files there https://github.com/openstack/openstack-ansible/tree/11.2.4/etc/openstack_deploy/env.d | 20:49 |
| cloudnull | did you condense the files by chanc e? | 20:50 |
| *** flwang has joined #openstack-ansible | 20:50 | |
| jwitko | omg im such an idiot | 20:50 |
| jwitko | sorry | 20:50 |
| cloudnull | essentially those are what make the inventory skeleton | 20:50 |
| cloudnull | no need to be sorry . | 20:50 |
| cloudnull | its confusing as shit | 20:50 |
| cloudnull | im sorry | 20:50 |
| cloudnull | :) | 20:50 |
| jwitko | lol | 20:51 |
| palendae | http://docs.openstack.org/developer/openstack-ansible/developer-docs/inventory.html talks a bit about that | 20:51 |
| cloudnull | admin0: was just playing with all of this just the other day :) | 20:51 |
| palendae | env.d + conf.d + dynamic_inventory.py = openstack_inventory.json | 20:51 |
| jwitko | cloudnull, list-hosts shows all of the proper hosts now | 20:51 |
| admin0 | :D | 20:51 |
| cloudnull | w00t! | 20:52 |
| jwitko | I don't even have a conf.d dir | 20:52 |
| jwitko | I need that as well I'm guessing | 20:52 |
| automagically | “confusing as shit” - cloudnull and “clear as mud” - odyssey4me to describe the current state of affairs of inventory. Sounds like we need a blueprint/spec to fix it | 20:52 |
| palendae | automagically: We have one :) | 20:52 |
| palendae | automagically: It's been a long, slow, ongoing process | 20:52 |
| palendae | jwitko: Might not have it; conf.d was a split out version of openstack_user_config.yml | 20:53 |
| cloudnull | automagically: it was the first blueprint https://blueprints.launchpad.net/openstack-ansible/+spec/dynamic-inventory-lib | 20:53 |
| automagically | ORLY…I’ll have to dig it up and see if I can contribute. My special hatred for it currently is not being able to use full/long hostnames | 20:53 |
| automagically | Since clearly I should tailor my hostnames to the eventual result of a dynamically generated LXC container name ;) | 20:53 |
| jwitko | ty palendae | 20:54 |
| cloudnull | yea thats an ssh problem so long as we use generated container names :( | 20:54 |
| cloudnull | however if we could configure that it'd be awesome | 20:54 |
| izaakk | me and raddaoui are trying to get involved in that inventory lib effort! | 20:54 |
| automagically | I’ll see what I can do, once I get out of IRR land | 20:54 |
| cloudnull | ++ | 20:54 |
| izaakk | discussion about it would be helpful | 20:55 |
| palendae | Yeah, izaakk, raddaoui, and neillc have all taken stabs at improving the inventory | 20:55 |
| palendae | There's thoughts to make it a completely indepdent thing, consumeable by anyonme | 20:55 |
| admin0 | i am an ops guy .. i should make blueprints :D | 20:55 |
| palendae | But IMO that's not as urgent as cleaning up the existing code and getting it tested | 20:55 |
| cloudnull | automagically odyssey4me and i have discussed using other inventory storage providers (like a KVS) and trying to change the interaction to that of a client call instead of a lesson in dictionary recursion. | 20:56 |
| admin0 | well, if i created mapping it will be this simple: service_keystone - runs on host_keytone on_metal : false on_containers: true :D | 20:56 |
| automagically | palendae: Refactoring is hard without tests as a safety net | 20:56 |
| admin0 | see , covers all :D | 20:56 |
| *** sdake has quit IRC | 20:56 | |
| palendae | automagically: https://review.openstack.org/#/c/242225/ | 20:56 |
| palendae | We're starting there | 20:56 |
| cloudnull | we've only talked about it, no code as of yet. however palendae and neillc have started that push | 20:56 |
| palendae | I want to add some more, but my priority right now is liberty upgrades | 20:57 |
| cloudnull | that said, +9000 on making inventory better. | 20:57 |
| palendae | That's a start but doesn't provide a tox env yet, and we need a gate job | 20:57 |
| automagically | Thanks folks, glad to know we all feel the pain and want to make it better | 20:57 |
| palendae | Which are fairly easy | 20:58 |
| palendae | automagically: Yeah, like cloudnull said, it was the first BP when we went to launchpad, and it was a scary, dark place even before then | 20:58 |
| admin0 | i lost 2 days and bugged palendae for 2 days before noticing a missed space :D | 20:58 |
| admin0 | in my inventory copy/paste :D | 20:58 |
| palendae | I think it's come down to other things being more pressing, and no one really knowing what the alternate providers would look like | 20:58 |
| *** sigmavirus24 is now known as sigmavirus24_awa | 20:58 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 20:58 | |
| palendae | Cause with config_template, you can get super deep nesting | 20:59 |
| palendae | And per host | 20:59 |
| palendae | So it gets nuts quickly | 20:59 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_nova: Implement base configuration for independent repository https://review.openstack.org/286736 | 20:59 |
| *** sdake has joined #openstack-ansible | 20:59 | |
| *** raddaoui_ has joined #openstack-ansible | 21:00 | |
| *** raddaoui has quit IRC | 21:00 | |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-repo_build: Implement base configuration for independent repository https://review.openstack.org/286625 | 21:02 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_swift: Implement base configuration for independent repository https://review.openstack.org/286692 | 21:03 |
| *** weshay has joined #openstack-ansible | 21:04 | |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible-os_horizon: [WIP] Implement base configuration for independent repository https://review.openstack.org/286650 | 21:10 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Update new user secrets in Liberty https://review.openstack.org/286844 | 21:19 |
| openstackgerrit | Major Hayden proposed openstack/openstack-ansible: Replacing LBaaSv1 with v2 configuration https://review.openstack.org/275440 | 21:24 |
| *** phiche1 has joined #openstack-ansible | 21:25 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_neutron: Implement base configuration for independent repository https://review.openstack.org/286694 | 21:26 |
| openstackgerrit | Matt Thompson proposed openstack/openstack-ansible-os_heat: Implement base configuration for independent repository https://review.openstack.org/286693 | 21:28 |
| *** phalmos has joined #openstack-ansible | 21:28 | |
| *** phiche has quit IRC | 21:28 | |
| *** johnmilton has quit IRC | 21:28 | |
| *** raddaoui_ has quit IRC | 21:29 | |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_openrc: Implement base configuration for independent repository https://review.openstack.org/286623 | 21:34 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add aodh, haproxy, and neturon config upgrades https://review.openstack.org/286753 | 21:35 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Update ceilometer environment configuration https://review.openstack.org/286827 | 21:35 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Update new user secrets in Liberty https://review.openstack.org/286844 | 21:36 |
| *** scarlisle has joined #openstack-ansible | 21:47 | |
| *** raddaoui_ has joined #openstack-ansible | 21:49 | |
| jwitko | hey cloudnull, sorry to bother again but can you tell me how long the step "Prepare cached image" takes where it attempts to speed up the cache image for container creation>? | 21:50 |
| jwitko | its been running for like 30m... I think there are some proxy issues going on where its not recognizing my environment variables | 21:51 |
| palendae | Hey rromans, your changes to https://review.openstack.org/#/c/285419/ resulted in the bash commands being output as comments (see http://docs-draft.openstack.org/19/285419/4/check/gate-openstack-ansible-docs/7545fa2//doc/build/html/upgrade-guide/overview.html) was that your intent? | 21:53 |
| *** fawadkhaliq has quit IRC | 21:54 | |
| cloudnull | jwitko: that should be a couple of min tops | 21:55 |
| cloudnull | apt can hang though (forever) and cause these types of problems | 21:56 |
| cloudnull | likely a proxy/resolver problem | 21:56 |
| cloudnull | to test you can chroot into the container image cache dir | 21:56 |
| cloudnull | and run apt-get update | 21:56 |
| cloudnull | and install some package to see if its hanging | 21:57 |
| *** Bjoern_ has joined #openstack-ansible | 21:58 | |
| spotz | hey guys do you know if there's a qa, non-live version of gerrit? | 22:00 |
| rromans | palendae: no, that wasn't my intent. The `#` should be the prompt | 22:00 |
| palendae | rromans: Ok, cool. I'm looking at pygments (which sphinx uses for highlighting) now, it has a 'bash session' code block | 22:00 |
| palendae | But I can't find the name to use | 22:00 |
| *** pcaruana has quit IRC | 22:01 | |
| palendae | rromans: http://pygments.org/demo/3830263/ looks more correct | 22:01 |
| *** phalmos has quit IRC | 22:02 | |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible: [DOC] Describe core reviewer team responsibilities https://review.openstack.org/286858 | 22:02 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add notes on running the upgrade script https://review.openstack.org/285419 | 22:03 |
| *** fawadkhaliq has joined #openstack-ansible | 22:04 | |
| palendae | rromans: Fixed it. Looks like we want the `console` "language" | 22:04 |
| palendae | `bash` is for scripts, `console` is for interactive input | 22:04 |
| rromans | palendae: fantastic, thanks much | 22:05 |
| *** neilus has joined #openstack-ansible | 22:05 | |
| spotz | automagically Are you sure those lines will wrap? | 22:05 |
| palendae | Will it wrap? | 22:05 |
| palendae | Whooooa | 22:06 |
| jwitko | cloudnull, thanks. it was the chroot, for some reason my proxy environment files don't get processed in a chroot | 22:06 |
| jwitko | so I had to add export commands for proxy variables into the lxc_cache_commands list | 22:06 |
| spotz | long sentence lines in an rst. | 22:06 |
| automagically | spotz: and palendae - I ran `tox docs` and viewed the resulting HTML before pushing for review | 22:06 |
| *** phalmos has joined #openstack-ansible | 22:06 | |
| automagically | I didn’t see any issues with the HTML output | 22:06 |
| spotz | Ok just checking, content itself is good:) | 22:07 |
| palendae | Hah, this patch is breaking my browser window | 22:07 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_swift_sync: Implement base configuration for independent repository https://review.openstack.org/286696 | 22:08 |
| *** thorst_ has quit IRC | 22:09 | |
| cloudnull | jwitko: is that something that we might be able to incorporate into the normal container cache create/build process | 22:09 |
| cloudnull | if so mind throwing up a PR for it ? | 22:09 |
| palendae | automagically: fwiw - wrapping at 80 characters or less will make gerrit happier, and still render the HTML cleanly. See https://review.openstack.org/#/c/285419/ and it's rendered output at http://docs-draft.openstack.org/19/285419/4/check/gate-openstack-ansible-docs/7545fa2//doc/build/html/upgrade-guide/overview.html | 22:10 |
| cloudnull | also NOTICE to everyone working on IRR things. Its likely that the linter checks are going to fail | 22:10 |
| cloudnull | https://review.openstack.org/#/c/286696/ | 22:10 |
| automagically | palendae: I’ll submit an update with wrapped lines | 22:10 |
| palendae | automagically: Thanks. I know it's a pain in the ass when writing them, but it is much easier to review when they're side by side | 22:10 |
| gus | Dear openstack-ansible devs, master is broken because 'repo' containers don't actually install virtualenv pip package, so repo_venv_build starts with 'which virtualenv' which promptly fails. | 22:10 |
| gus | kthxbye. | 22:10 |
| automagically | Nah, I get it. And I don’t mind optimizing for review on this one. I expect plenty of feedback | 22:11 |
| palendae | I need to update my vimrc to highlight the 80 character limit in rst files, just been lazy | 22:11 |
| automagically | gus: Just noticed that on a gate | 22:11 |
| palendae | gus: Been seeing that too, thanks | 22:11 |
| palendae | Odd that that's just now breaking | 22:11 |
| openstackgerrit | Kevin Carter proposed openstack/openstack-ansible-os_swift_sync: Implement base configuration for independent repository https://review.openstack.org/286696 | 22:12 |
| palendae | Unless it's the IRRs doing it? | 22:12 |
| cloudnull | gus: yes. this is true | 22:12 |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible: [DOC] Describe core reviewer team responsibilities https://review.openstack.org/286858 | 22:13 |
| cloudnull | we're waiting on a few PRs to get through the gate to fix it | 22:13 |
| *** mariognarly has quit IRC | 22:13 | |
| cloudnull | gate is painfully slow right now :'( | 22:13 |
| palendae | Ahh | 22:13 |
| palendae | Then I'll stop rechecking my master patches | 22:13 |
| spotz | hehehe | 22:13 |
| spotz | Ok and if anyone is interested bgmccollum *hint hint* https://review-dev.openstack.org | 22:14 |
| palendae | My browser's showing me a scary message | 22:14 |
| palendae | I don't trust openstack.org | 22:14 |
| *** neilus has quit IRC | 22:14 | |
| cloudnull | gus once https://review.openstack.org/#/c/284896/ and https://review.openstack.org/#/c/286625/ go in master will be unbroken | 22:15 |
| automagically | palendae: Ha ha | 22:15 |
| spotz | hehe | 22:15 |
| cloudnull | gus: if you pull those patches in you will be good to go | 22:15 |
| spotz | Don't want my summit trainees closing real bugs by accident | 22:15 |
| palendae | spotz: Oh cool. A test instance for doing gerrit stuff? | 22:16 |
| gus | I'm curious (in the the "I'm not actually going to do anything about it" way) what changed to allow the current version to have made it passed the gate checks in the first place. | 22:16 |
| *** mariognarly has joined #openstack-ansible | 22:16 | |
| palendae | gus: Today a lot of the roles in the repository got moved into their own repositories | 22:16 |
| spotz | yep Palendae. So we've got qa launchpad, a git sandbox and now a dev gerrit:) | 22:16 |
| gus | aha, so we just lost some defaults? | 22:17 |
| palendae | gus: https://review.openstack.org/#/q/status:open+topic:bp/independent-role-repositories is a list of new ones, some of the repos are empty til the first commits land | 22:17 |
| palendae | Well, all of them are empty til the first commit lands, but some haven't landed at all | 22:17 |
| gus | so long as people smarter than me are already involved in the post-mortem and ensuring we won't have issues like this again, then I'm happy the process is working. | 22:17 |
| *** alejandrito has quit IRC | 22:18 | |
| cloudnull | gus: no its the process of migrating more roles to independent once out of the main repo. | 22:20 |
| cloudnull | the process causes pain | 22:21 |
| *** Mudpuppy has quit IRC | 22:21 | |
| automagically | cloudnull: __understatement__ of the year | 22:21 |
| *** keedya has quit IRC | 22:21 | |
| cloudnull | it was decided to do all of these in one go. | 22:21 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add aodh, haproxy, and neturon config upgrades https://review.openstack.org/286753 | 22:21 |
| cloudnull | im not sure why | 22:21 |
| *** raddaoui_ has quit IRC | 22:21 | |
| cloudnull | all of the rest have been done in stages | 22:22 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Update ceilometer environment configuration https://review.openstack.org/286827 | 22:22 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Update new user secrets in Liberty https://review.openstack.org/286844 | 22:22 |
| *** ric has joined #openstack-ansible | 22:22 | |
| stevelle | https://review.openstack.org/#/c/286625 came back from the gate, needs more eyeballs | 22:23 |
| cloudnull | i mean i did most of the work however i disagreed with revised process. i would liked to do the os_* roles one at a time. | 22:23 |
| palendae | When you *just* updated a commit and its 2 dependents and see a typo :| | 22:23 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add aodh, haproxy, and neutron config upgrades https://review.openstack.org/286753 | 22:24 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Update ceilometer environment configuration https://review.openstack.org/286827 | 22:24 |
| stevelle | palendae: revision numbers are cheap :P | 22:24 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Update new user secrets in Liberty https://review.openstack.org/286844 | 22:24 |
| palendae | stevelle: Sure, but manual rebasing is a pita | 22:24 |
| stevelle | clouds are a pita | 22:25 |
| palendae | I don't really care about the revision numbers, but the process of looping back and rebasing everything again | 22:25 |
| stevelle | ftfy | 22:25 |
| palendae | Well yes, computers are | 22:25 |
| palendae | Current gerrit interface doesn't have a great way of visualizing dependent patches anymore. Unless I'm missing it | 22:25 |
| openstackgerrit | Nolan Brubaker proposed openstack/openstack-ansible: Add missing runtime information to upgrade script https://review.openstack.org/286251 | 22:26 |
| stevelle | sadly. Would be really nice to have that back | 22:26 |
| palendae | Yeah | 22:26 |
| palendae | That would make this process a lot easier | 22:26 |
| *** phalmos has quit IRC | 22:27 | |
| *** jguy_ has left #openstack-ansible | 22:27 | |
| palendae | I *did* just notice that there's a little green circle next to the parent revision SHA if a rebase is possible/needed | 22:27 |
| *** phalmos has joined #openstack-ansible | 22:27 | |
| palendae | That's handy | 22:27 |
| openstackgerrit | Merged openstack/openstack-ansible-repo_build: Implement base configuration for independent repository https://review.openstack.org/286625 | 22:28 |
| bgmccollum | spotz what am i looking at...?!? | 22:28 |
| spotz | bgmccollum: A test/play gerrit | 22:28 |
| cloudnull | now we need https://review.openstack.org/#/c/284896/ and the gate should be happy once again . | 22:30 |
| *** KLevenstein has joined #openstack-ansible | 22:30 | |
| *** retreved has quit IRC | 22:33 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_ceilometer: Implement base configuration for independent repository https://review.openstack.org/286674 | 22:34 |
| spotz | I'm the last +1 cloudnull | 22:34 |
| spotz | People wake up and review!:) | 22:35 |
| cloudnull | looks like the gate has about a 4 hour lead time. so ill be back in a while . | 22:36 |
| spotz | seeya tomorrow cloudnull | 22:36 |
| *** raddaoui_ has joined #openstack-ansible | 22:36 | |
| cloudnull | for anyone interested these are important for several of our stable maintainers https://review.openstack.org/#/q/status:open+branch:liberty+topic:liberty_upgrade -- if you have review cycles it'd be great to get some feedback on these items | 22:37 |
| cloudnull | and the ongoing initial work for the next batch of IRR roles https://review.openstack.org/#/q/status:open+branch:master+topic:bp/independent-role-repositories | 22:38 |
| * automagically looking cloudnull | 22:38 | |
| -cloudnull- if you have review cycles the following reviews could use your left over cycles https://review.openstack.org/#/q/status:open+branch:master+topic:bp/independent-role-repositories and https://review.openstack.org/#/q/status:open+branch:master+topic:bp/independent-role-repositories -- TYVM :) | 22:39 | |
| *** scarlisle has quit IRC | 22:39 | |
| cloudnull | tyvm automagically .. ill be back in a bit . | 22:39 |
| cloudnull | going to get dog food | 22:39 |
| openstackgerrit | Travis Truman (automagically) proposed openstack/openstack-ansible-os_horizon: [WIP] Implement base configuration for independent repository https://review.openstack.org/286650 | 22:48 |
| *** scarlisle has joined #openstack-ansible | 22:52 | |
| *** raddaoui_ has quit IRC | 22:55 | |
| openstackgerrit | Merged openstack/openstack-ansible: Docs: Add missing horizontal rule / navigation links https://review.openstack.org/286657 | 23:02 |
| *** busterswt has quit IRC | 23:08 | |
| *** busterswt has joined #openstack-ansible | 23:08 | |
| *** busterswt has quit IRC | 23:08 | |
| *** phiche1 has quit IRC | 23:10 | |
| stevelle | palendae: any chance you will push another change set for https://review.openstack.org/#/c/286251 for that extra . | 23:13 |
| *** scarlisle has quit IRC | 23:13 | |
| stevelle | I'm fine with letting it go, it could get cleaned up later | 23:14 |
| spotz | anything you guys need me to look at before I head out in a few? | 23:15 |
| *** jthorne has quit IRC | 23:20 | |
| *** Bjoern_ has quit IRC | 23:24 | |
| *** KLevenstein has quit IRC | 23:26 | |
| openstackgerrit | Merged openstack/openstack-ansible: Add notes on running the upgrade script https://review.openstack.org/285419 | 23:31 |
| *** darrenc is now known as darrenc_afk | 23:33 | |
| *** ric has quit IRC | 23:35 | |
| rromans | stevelle: palendae: I can push the typo fix up now if you want? also fine with leaving it. | 23:38 |
| stevelle | rromans: I vote for fix it later | 23:39 |
| rromans | kk | 23:39 |
| stevelle | was just going to see if he was here still and working on it | 23:39 |
| rromans | it's super minor | 23:39 |
| stevelle | yeah, and there will be lots of changes to those files. It can be reasonably slipped into one of the following changes. | 23:39 |
| *** admin0 has quit IRC | 23:47 | |
| *** sdake has quit IRC | 23:47 | |
| *** spotz is now known as spotz_zzz | 23:47 | |
| palendae | stevelle, rromans Not tonight | 23:48 |
| stevelle | palendae: endorsed. | 23:48 |
| rromans | cheers | 23:48 |
| palendae | Someone wants to fix it, go for it :) | 23:48 |
| palendae | stevelle, Yeah, and more commits to come, though I think I'm approaching the end of my backlog of program changes. Will need to get into the manual process after | 23:49 |
| palendae | Which is more or less writing a doc that shows running each thing in turn | 23:49 |
| *** darrenc_afk is now known as darrenc | 23:50 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!