Tuesday, 2018-02-20

ANKITAeven in containers i have to install nfs-commn to mount it manually00:00
*** masber has joined #openstack-ansible00:21
*** weezS has quit IRC00:25
*** weezS_ has quit IRC00:25
*** dave-mccowan has joined #openstack-ansible00:31
openstackgerritMerged openstack/openstack-ansible-os_nova master: tests: test-nova-upgrades.sh: Drop Ansible verbosity  https://review.openstack.org/53737900:32
*** pmannidi has quit IRC00:37
*** pmannidi has joined #openstack-ansible00:40
*** weezS has joined #openstack-ansible00:42
*** lbragstad has quit IRC00:43
*** dave-mccowan has quit IRC00:53
*** weezS has quit IRC00:57
*** armaan has quit IRC01:00
*** acormier_ has quit IRC01:08
*** acormier has joined #openstack-ansible01:09
*** chyka_ has joined #openstack-ansible01:13
*** acormier has quit IRC01:13
*** acormier has joined #openstack-ansible01:14
*** acormier has quit IRC01:14
*** acormier has joined #openstack-ansible01:15
*** chyka has quit IRC01:16
*** dave-mccowan has joined #openstack-ansible01:18
*** ANKITA has quit IRC01:22
*** dave-mccowan has quit IRC01:29
*** woodard_ has joined #openstack-ansible01:39
*** woodard has quit IRC01:39
*** chyka_ has quit IRC01:49
*** chyka has joined #openstack-ansible01:50
*** chyka has quit IRC01:55
*** armaan has joined #openstack-ansible01:55
*** esberglu has quit IRC02:04
*** esberglu has joined #openstack-ansible02:04
*** jafeha has joined #openstack-ansible02:08
*** esberglu has quit IRC02:09
*** jafeha__ has quit IRC02:10
*** weezS has joined #openstack-ansible02:14
*** vishwanathj has joined #openstack-ansible02:20
*** esberglu has joined #openstack-ansible02:31
*** esberglu has quit IRC02:39
*** esberglu has joined #openstack-ansible02:40
*** ANKITA has joined #openstack-ansible02:41
*** weezS has quit IRC02:41
*** esberglu has quit IRC02:44
*** lbragstad has joined #openstack-ansible03:01
*** dave-mccowan has joined #openstack-ansible03:10
*** gkadam has joined #openstack-ansible03:39
*** acormier has quit IRC03:50
*** esberglu has joined #openstack-ansible03:51
*** esberglu has quit IRC03:56
*** jafeha__ has joined #openstack-ansible04:03
*** jafeha has quit IRC04:04
*** chyka has joined #openstack-ansible04:06
*** chyka has quit IRC04:10
*** udesale has joined #openstack-ansible04:13
*** lbragstad has quit IRC04:26
cloudnullevenings04:38
*** esberglu has joined #openstack-ansible04:48
*** esberglu has quit IRC04:52
idlemindo704:54
prometheanfireo7 cmdr04:55
*** dave-mccowan has quit IRC05:02
*** poopcat has quit IRC05:02
ANKITAGE cloudnull thanks for the help; guidance & support odyssey4me; evrardjp and all members frm d community despite of my limited knowledge;  have finally completed the deployment05:07
ANKITAhave a g8 day05:07
ANKITATONS OF GRATITUDE05:09
ANKITAFOUND https://docs.openstack.org/openstack-ansible/pike/contributor/inventory-and-vars.html TO BE HELPFUL05:14
*** dariko has joined #openstack-ansible05:16
*** gkadam has quit IRC05:43
*** gkadam has joined #openstack-ansible05:44
*** gokhan has joined #openstack-ansible06:01
*** dariko has quit IRC06:01
*** hybridpollo has quit IRC06:19
*** sxc731 has joined #openstack-ansible06:40
ANKITApath for /openstack/log/ansible-logging/ansible.log06:42
*** jafeha__ is now known as jafeha06:45
jafehagood morning everybody06:45
*** sar has quit IRC06:51
*** acormier has joined #openstack-ansible06:51
*** chyka has joined #openstack-ansible06:55
*** acormier has quit IRC06:55
*** chyka has quit IRC06:59
*** sxc731 has quit IRC07:00
*** Sha000000 has joined #openstack-ansible07:05
Taseerevrardjp: Good morning ! any reservations => https://review.openstack.org/#/c/503971/07:09
Taseer>07:10
*** threestrands has quit IRC07:10
*** sar has joined #openstack-ansible07:19
openstackgerritMaxime Guyot proposed openstack/openstack-ansible master: Ceph RadosGW integration  https://review.openstack.org/51785607:32
*** gkadam has quit IRC07:34
*** gkadam has joined #openstack-ansible07:34
*** SmearedBeard has joined #openstack-ansible07:35
*** pcaruana has joined #openstack-ansible07:35
openstackgerritGökhan proposed openstack/openstack-ansible-os_ceilometer master: Deprecate os_endpoint_type option Option "os_endpoint_type" from groups "service_credentials" and "keystone_authtoken" is deprecated[1]. Use option "interface" from groups "service_credentials" and "keystone_authtoken".  https://review.openstack.org/52185707:42
*** alex___ has joined #openstack-ansible07:43
*** alex___ is now known as fusmu07:43
*** Sha000000 has quit IRC07:47
*** mbuil has joined #openstack-ansible07:54
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible master: [Docs] Promote scale the environment  https://review.openstack.org/54559207:58
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible master: [Docs] Promote Backup and Restore  https://review.openstack.org/54559307:58
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible master: [Docs] Flatten out maintenance task  https://review.openstack.org/54559707:58
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible master: [Docs] Move upgrade guides into ops  https://review.openstack.org/54566408:02
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible master: [Docs] Move upgrade guides into ops  https://review.openstack.org/54566408:03
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible master: [Docs] Migrate CLI reference to reference guide  https://review.openstack.org/54579008:03
*** epalper has joined #openstack-ansible08:04
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible master: [Docs] Simplify advanced config  https://review.openstack.org/54579508:09
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible master: [Docs] Merge advanced configuration into reference  https://review.openstack.org/54580508:10
openstackgerritMerged openstack/openstack-ansible master: [Docs] Promote scale the environment  https://review.openstack.org/54559208:20
openstackgerritMerged openstack/openstack-ansible master: [Docs] Promote Backup and Restore  https://review.openstack.org/54559308:24
openstackgerritMerged openstack/openstack-ansible master: [Docs] Flatten out maintenance task  https://review.openstack.org/54559708:24
openstackgerritMerged openstack/openstack-ansible master: [Docs] Move upgrade guides into ops  https://review.openstack.org/54566408:24
openstackgerritMerged openstack/openstack-ansible master: [Docs] Migrate CLI reference to reference guide  https://review.openstack.org/54579008:26
*** sxc731 has joined #openstack-ansible08:26
openstackgerritMerged openstack/openstack-ansible master: [Docs] Simplify advanced config  https://review.openstack.org/54579508:32
*** fusmu has quit IRC08:32
*** fusmu_ has joined #openstack-ansible08:32
openstackgerritMerged openstack/openstack-ansible master: [Docs] Merge advanced configuration into reference  https://review.openstack.org/54580508:35
openstackgerritGökhan proposed openstack/openstack-ansible-os_ceilometer master: Deprecate os_endpoint_type option Option "os_endpoint_type" from groups "service_credentials" and "keystone_authtoken" is deprecated[1]. Use option "interface" from groups "service_credentials" and "keystone_authtoken".  https://review.openstack.org/52185708:36
gunixodyssey4me: cloudnull: i left that centos to do osa aio over night and it freezed on isntalling yum packages. i found a quick solution. i deleted the centos7 vm and i am installing in ubuntu.08:36
openstackgerritGökhan proposed openstack/openstack-ansible-os_ceilometer master: Deprecate os_endpoint_type option Option "os_endpoint_type" from groups "service_credentials" and "keystone_authtoken" is deprecated[1]. Use option "interface" from groups "service_credentials" and "keystone_authtoken".  https://review.openstack.org/52185708:38
*** chyka has joined #openstack-ansible08:44
*** gokhan has quit IRC08:45
*** chyka has quit IRC08:49
kklimondahow can I apply a patch for neutron/use a custom neutron package?08:49
Miougekklimonda: you can specify a custom repo & branch with neutron_git_repo and neutron_git_install_branch see https://docs.openstack.org/openstack-ansible-os_neutron/latest/08:51
kklimondathanks08:51
Miougekklimonda: I think that it is used only if “neutron_developer_mode: true”, so probably not suitable for a production setup.08:53
*** armaan has quit IRC09:00
*** armaan has joined #openstack-ansible09:01
*** shardy has joined #openstack-ansible09:09
*** pmannidi has quit IRC09:16
*** Sha000000 has joined #openstack-ansible09:19
kklimondabummer09:30
*** indistylo has joined #openstack-ansible09:32
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_nova master: Remove systemd conditionals  https://review.openstack.org/54498709:39
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_nova master: Stop running non-voting jobs in gate  https://review.openstack.org/54348609:40
*** aruns has joined #openstack-ansible09:40
*** aruns has quit IRC09:40
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_nova master: Change include: to include_tasks:  https://review.openstack.org/54498609:41
*** aruns has joined #openstack-ansible09:41
*** indistylo has quit IRC09:42
*** aruns has quit IRC09:45
*** indistylo has joined #openstack-ansible09:46
*** ankita_ has joined #openstack-ansible09:49
*** ankita_ has quit IRC09:49
openstackgerritTaseer Ahmed proposed openstack/openstack-ansible master: Integrate Congress with OSA.  https://review.openstack.org/50397109:50
*** Sha000000 has quit IRC09:53
*** aruns has joined #openstack-ansible09:58
*** indistylo has quit IRC10:00
odyssey4meANKITA hooray!\10:15
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_nova stable/queens: Add SELinux contexts for nova log directory  https://review.openstack.org/54610310:20
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_nova stable/queens: tests: test-nova-upgrades.sh: Drop Ansible verbosity  https://review.openstack.org/54610410:21
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_nova stable/pike: tests: test-nova-upgrades.sh: Drop Ansible verbosity  https://review.openstack.org/54610510:21
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_build master: [WIP] Make the git clone and wheel build optional  https://review.openstack.org/53701510:30
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_build master: Ensure virtualenv is not broken for CentOS/SuSE too  https://review.openstack.org/54599110:34
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_build master: [WIP] Make the git clone and wheel build optional  https://review.openstack.org/53701510:34
ThomasShi10:38
odyssey4meevrardjp hwoarang I'd like to get https://review.openstack.org/545849 merged asap to prevent any confusion. If you could take a peek that'd be great.10:40
odyssey4mealso https://review.openstack.org/545697 needs one more review10:41
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron stable/queens: Stop running non-voting jobs in gate  https://review.openstack.org/54610910:42
*** jamespage has quit IRC10:43
hwoarangok10:43
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron stable/queens: Add SELinux contexts for neutron log directory  https://review.openstack.org/54588410:44
*** jamespage has joined #openstack-ansible10:46
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/pike: Ensure that constraints are used for AIO bootstrap  https://review.openstack.org/54589410:47
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/ocata: Ensure that constraints are used for AIO bootstrap  https://review.openstack.org/54589510:47
odyssey4methanks evrardjp hwoarang - one more bug fix which needs to get back into queens: https://review.openstack.org/54599110:53
odyssey4meI just rebased it a moment ago which is why it's still running tests.10:54
*** fusmu_ has quit IRC10:59
hwoaranghmm i thought i voted on it10:59
*** gkadam has quit IRC11:06
*** gkadam has joined #openstack-ansible11:07
odyssey4mehwoarang this is a follow-on from another patch which you may have voted on11:08
*** pbandark has joined #openstack-ansible11:09
*** gkadam has quit IRC11:12
*** gkadam has joined #openstack-ansible11:12
*** foutatoro has joined #openstack-ansible11:14
*** gkadam has quit IRC11:18
*** aruns__ has joined #openstack-ansible11:19
*** aruns has quit IRC11:22
*** armaan has quit IRC11:23
*** gkadam has joined #openstack-ansible11:30
hwoarangodyssey4me: hmm some of the jobs still fail on the upgrade test11:40
hwoarangbut sometimes they also pass11:40
hwoarangveeeeeery strange11:40
odyssey4mesome depend on others having the right things done - eg: nova need neutron/glance/keystone to be right before it works right11:41
openstackgerritMerged openstack/openstack-ansible-os_octavia stable/queens: Updated from OpenStack Ansible Tests  https://review.openstack.org/54493211:41
odyssey4mebut yeah, there are still some jobs failing, and I unfortunately don't have time right now to look into it - evrardjp knows the reasons for those failures and has patched many of them up, so maybe looking into those patches would give the right clues11:42
*** armaan has joined #openstack-ansible11:42
odyssey4mehwoarang for https://review.openstack.org/545991 I wasn't sure if ansible_plg_mgr was the best way to do the platform differentiation - suggestions are welcome11:43
*** udesale has quit IRC11:43
*** stuartgr has joined #openstack-ansible11:43
hwoarangyeah should be fine11:51
openstackgerritMerged openstack/openstack-ansible-os_nova master: Stop running non-voting jobs in gate  https://review.openstack.org/54348612:04
*** rpittau has quit IRC12:05
*** acormier has joined #openstack-ansible12:06
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_nova stable/queens: Stop running non-voting jobs in gate  https://review.openstack.org/54612812:06
*** gkadam has quit IRC12:06
*** acormier has quit IRC12:06
*** acormier has joined #openstack-ansible12:07
openstackgerritPeriyasamy Palanisamy proposed openstack/openstack-ansible master: Make Opendaylight as the BGP speaker using Quagga  https://review.openstack.org/52390712:16
openstackgerritPeriyasamy Palanisamy proposed openstack/openstack-ansible master: Make Opendaylight as the BGP speaker using Quagga  https://review.openstack.org/52390712:20
*** gkadam has joined #openstack-ansible12:22
*** chyka has joined #openstack-ansible12:25
*** chyka has quit IRC12:30
openstackgerritMerged openstack/openstack-ansible-os_tempest master: Use realpath filter instead of the stat/dir dance  https://review.openstack.org/54582912:31
*** sxc731 has quit IRC12:32
*** foutatoro has quit IRC12:36
openstackgerritMerged openstack/openstack-ansible master: Replace host_need_pip with standard override mechanism  https://review.openstack.org/54569712:37
*** acormier has quit IRC12:37
*** acormier has joined #openstack-ansible12:37
*** acormier has quit IRC12:42
openstackgerritMerged openstack/openstack-ansible-os_magnum stable/pike: Zuul: Remove project name  https://review.openstack.org/53864212:42
*** indistylo has joined #openstack-ansible12:49
*** aruns__ has quit IRC12:51
*** sm806 has joined #openstack-ansible12:54
JohnnyOSAHello all.  Have a question about my OSA deploy configuration (see: http://paste.openstack.org/show/678432/)12:59
JohnnyOSAI have a flat provider network I plan on using for both a floating IP range to for an IP for the external lb vip address using ens3 (untagged).13:00
JohnnyOSA(..."and for" rather than "to for"...)13:01
JohnnyOSAens3 is captured by br-vlan to map the floating ip network range with neutron, and the haproxy_keepalived_external_interface is set to br-vlan13:02
JohnnyOSAThe default gateway in the interfaces file is the management network's gateway, 172.29.236.1, which is routable to the internet.13:04
*** sm806_ has joined #openstack-ansible13:05
JohnnyOSAMost everything seems to work ok, except for Horizon.  Without a few extra "ip route" and "ip rule" commands which are seen in the paste above under the br-vlan device, the external lb VIP is accessible only from the same subnet that the external VIP is on (10.9.198.0/24).13:06
*** sm806 has quit IRC13:06
*** sm806_ has quit IRC13:06
JohnnyOSAThis appears to be due to a source-routing issue where an incoming http/s request to the external LB VIP leaves the haproxy server out the default gateway (172.29.236.1) rather than the same interface that the external VIP is on, and that traffic then appears to get dropped.13:08
JohnnyOSA(before it reaches the requesting computer).13:09
JohnnyOSAIf I add a few lines to the configuration for br-vlan, such as ip route and ip rule see in the paste, to ensure that traffic coming into the br-vlan interface leaves also leaves the br-vlan interface on an appropriate gateway rather than a different interfaces gateway, I can get to Horizon just fine from other external subnets.13:11
*** sm806 has joined #openstack-ansible13:11
JohnnyOSAHowever, the issue now is that Horizon throws 504 page response timeouts on certain Horizon pages, such as the network topology page and the keypair page13:12
*** sm806 has quit IRC13:15
JohnnyOSAIs there a better approach than what I'm doing here?  Has anyone else run into what seems like a source-routing issue like this?13:15
openstackgerritJean-Philippe Evrard proposed openstack/openstack-ansible-os_nova master: Updated from global requirements  https://review.openstack.org/53506713:16
*** sm806 has joined #openstack-ansible13:16
*** chyka has joined #openstack-ansible13:17
JohnnyOSA(will be on road shortly -- be back soon)13:17
mhaydenbuenos dias13:20
*** chyka has quit IRC13:22
*** acormier has joined #openstack-ansible13:22
mhaydenevrardjp: i can go change some of these to "import_tasks" but i had some peculiar results when i did that in ansible-hardening https://review.openstack.org/54498613:26
evrardjpmhayden: could you link to the ansible-hardening attempt? That's the nova change.13:27
*** acormier has quit IRC13:27
mhaydenthe ansible-hardening one was the one we reverted ;)13:27
evrardjplink?13:28
evrardjplet me find that13:28
* mhayden searches13:28
mhaydenhttps://review.openstack.org/#/c/534327/13:28
mhaydenbah, CI jobs are missing13:29
evrardjpmhayden: I am confused13:29
evrardjpI thought the revert was to support more than ansible 2.413:29
mhaydenthe revert removed the patch entirely13:30
mhaydenand restored the warnings13:30
mhaydenand ansible < 2.4 support13:30
evrardjpok so I was right on that part.13:31
mhaydengenerally you want to use import_tasks where you once used "static: yes"13:31
evrardjpyes.13:31
mhaydeninclude_tasks is where you would have used static: no13:31
evrardjpthat's true13:31
evrardjpand?13:31
mhaydenso on the nova patch, you suggested using import_tasks more often13:31
evrardjpanything that should be serialized is import_task13:31
mhaydenwhich is okay, but that might break13:31
evrardjpwell13:31
evrardjpmy question is why13:32
*** sm806 has quit IRC13:32
mhaydeni have the same question :P13:32
mhaydenit seems to break things in odd ways13:32
*** sm806 has joined #openstack-ansible13:32
sm806In my compute node i see this error been printed in /var/log/neutron/neutron-openvswitch-agent.log --> "Bridge br-vlan for physical network vlan does not exist. Agent terminated!"13:32
sm806I have not configured anible to use ovs. So i am hoping my openstack setup would use only linux bridges. I see this issue with vlan network only. vxlan and flat network seems to be working fine.13:33
sm806Logs, "brctl show" and "ovs-vsctl show" --> https://paste.ubuntu.com/p/FqGHf9J3T6/13:33
evrardjpI thought the jinja in the filenames was now allowed with import.13:33
mhaydensm806: do you have a bridge called "br-vlan"?13:33
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_tempest stable/queens: Use realpath filter instead of the stat/dir dance  https://review.openstack.org/54614713:33
evrardjpanyway13:33
mhaydenevrardjp: well, i could flip them all to "import" and then change the ones that need to be changed13:33
evrardjpdo what you want :)13:33
* mhayden has the powerrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrrr13:34
sm806mhayden : yes its is created as linuxbridge. Looks like it is looking for a ovs bridge !!!13:34
mhaydensm806: that is a bit unusual13:34
evrardjpmhayden: you already had it.13:34
evrardjpit's called +w .13:34
mhaydensm806: so you want to use linuxbridge exclusively? no OVS at all?13:34
sm806mhayden : yes i want try our linuxbridge only for now.13:35
mhaydensm806: here's my global_overrides from /etc/openstack_deploy/openstack_user_config.yml https://paste.fedoraproject.org/paste/DUIl-nqUTCnQoyhFUlnzBA/raw13:36
mhaydeni am using linuxbridge only13:36
mhaydensm806: i believe the "group_binds" is what controls that13:36
mhaydenbut it's been a while since i've dug into that13:37
openstackgerritMerged openstack/openstack-ansible stable/pike: Update all SHAs for 16.0.9  https://review.openstack.org/54541013:38
*** woodard_ has quit IRC13:38
*** woodard has joined #openstack-ansible13:39
mhaydenevrardjp: also, on the CentOS + AIO basekit, i figured it might be a nice way to get CentOS into a voting gate13:39
mhaydensince it reduces the services it deploys13:39
sm806mhayden : https://paste.fedoraproject.org/paste/W1r-7KKAYQ9CcvgggvVklA this is my openstack_user_config.yml13:39
* mhayden ganders13:39
mhaydensm806: okay, that looks fine13:40
mhaydenhow about user_variables.yml?13:41
mhaydenfeel free to mask/hide anything sensitive in there ;)13:41
sm806mhayden: this is my user_variahles.yml https://paste.fedoraproject.org/paste/tTzpI6RWf70lT55-eo13Lw13:44
sm806mhayden: most of the options are commented :)13:44
mhaydentaking a look13:45
mhaydenhmm, i don't see any neutron ml2 stuff set there, so you should get linuxbridge by default13:45
mhaydensm806: which branch are you dploying?13:45
sm806i am using pike.13:46
sm806mhayden : i tried to enable ovs in my setup few days back. Ran to some issues and i reverted back to linuxbridge.13:48
*** esberglu has joined #openstack-ansible13:48
sm806mhayden : reverted to my previous openstack_user_config.yml anduser_variables.yml version.13:49
sm806reinstalled ubuntu in my infra and compute nodes and re-executed the anisible playbooks13:49
*** chyka has joined #openstack-ansible13:50
sm806before re-executing the playbooks, i cleaned up following files and dirs -- cd /etc/openstack_deploy; rm -rf ansible_facts; rm backup_openstack_inventory.tar; rm openstack_hostnames_ips.yml; rm openstack_inventory.json13:52
openstackgerritMerged openstack/openstack-ansible-os_tempest master: Updated from OpenStack Ansible Tests  https://review.openstack.org/54466313:55
*** chyka has quit IRC13:55
*** Smeared_Beard has joined #openstack-ansible13:56
mhaydensm806: ah, did you go out and manually stop those ovs services?13:56
mhaydenthe ansible playbooks won't go and shut them off if they were started previously13:56
mhaydenyou'll need to manually stop/disable them13:56
*** electrofelix has quit IRC13:57
*** SmearedBeard has quit IRC13:57
*** Sha000000 has joined #openstack-ansible13:58
*** dave-mccowan has joined #openstack-ansible13:59
sm806mhayden: I reinstalled ubuntu on the infra and compute hosts.So expected ansible script wont install them if ovs is not enabled in config files.14:00
sm806i see ovs is running in the host. I will stop them and try creating a vlan network14:01
sm806mhayden : looks like neutron ovs agent plugin is still trying to contact ovs service.14:04
sm806https://paste.fedoraproject.org/paste/oIM~DT2fMQHOYNnxtGznmQ14:05
openstackgerritMajor Hayden proposed openstack/openstack-ansible-tests master: Set SELinux to permissive mode for tests  https://review.openstack.org/54615314:05
sm806i stoped ovs service using "/etc/init.d/openvswitch-switch stop"14:05
openstackgerritMajor Hayden proposed openstack/openstack-ansible master: Set SELinux to permissive mode for tests  https://review.openstack.org/54615514:08
openstackgerritMajor Hayden proposed openstack/openstack-ansible master: Use aio_basekit scenario for CentOS 7  https://review.openstack.org/54545514:10
openstackgerritMajor Hayden proposed openstack/openstack-ansible master: Use package_state=present for CentOS 7 testing  https://review.openstack.org/54101714:11
openstackgerritMajor Hayden proposed openstack/openstack-ansible master: CentOS 7 integrated gate optimization  https://review.openstack.org/54545514:13
mhaydensm806: i'm not entirely sure what to do there14:14
mhaydenyou may want to look under /etc/neutron to ensure there's not an ovs-related configuration file there14:14
*** aruns__ has joined #openstack-ansible14:15
*** indistylo has quit IRC14:17
openstackgerritMajor Hayden proposed openstack/openstack-ansible master: CentOS 7 integrated gate optimization  https://review.openstack.org/54545514:25
evrardjpdear cloudnull, DimGR, andymccr, d34dh0r53, hughsaunders, b3rnard0, palendae, odyssey4me, serverascode, rromans, erikmwilson, mancdaz, _shaps_, BjoernT, claco, echiu, dstanek, jwagner, ayoung, prometheanfire, evrardjp, arbrandes, mhayden, scarlisle, luckyinva, ntt, javeriak, spotz, vdo, jmccrory, alextricity25, jasondotstar, admin0, michaelgugino, ametts, bgmccollum, darrenc, JRobinson__, asettle,14:25
evrardjpcolinmcnamara, thorst, adreznec, eil397, qwang,nishpatwa_, cathrichardson, drifterza, hwoarang, cshen, ullbeking14:25
evrardjpthe osa bug triage will start in 1h30'. Please have a look at the bug list before starting: https://etherpad.openstack.org/p/osa-bugtriage14:25
*** lbragstad has joined #openstack-ansible14:29
JohnnyOSARe: my query above on flat routing.  I'm making an assumption that the ip route/rule additions to the interfaces file caused the Horizon 504 timeouts on select pages since those errors happened as soon as I introduced rules/route cmds.  I'll verify that by removing and seeing if the 504 Horizon errors disappear.  If not, I'll start looking for another root cause.14:29
*** jwitko_ has joined #openstack-ansible14:29
openstackgerritMerged openstack/openstack-ansible-os_nova master: Remove systemd conditionals  https://review.openstack.org/54498714:30
openstackgerritMerged openstack/openstack-ansible-os_nova stable/pike: tests: test-nova-upgrades.sh: Drop Ansible verbosity  https://review.openstack.org/54610514:30
*** sar has quit IRC14:34
*** gillesMo has joined #openstack-ansible14:36
gillesMoHello, After an upgrade from Ocata to Pike, My Ceph cluster has been upgraded from jewel to luminous. But... There's no ceph-mgr started, the cluster is not healthy, and I cant' see any pool... I can't see if there is something to add in openstack_user_config.yml to add ceph mgr demons14:39
*** Sha000000 has quit IRC14:43
evrardjpgit-harry: could you file a bug in OSA for your galera issue?14:52
*** acormier has joined #openstack-ansible14:52
evrardjpthis way we can track it at both ways14:52
evrardjpalso I'd add it to the triage of today14:52
*** udesale has joined #openstack-ansible14:56
*** indistylo has joined #openstack-ansible14:56
*** aruns__ has quit IRC14:58
evrardjpgit-harry: wrong wording on my side, maybe that's clearer: https://review.openstack.org/#/c/545844/115:01
*** gokhan has joined #openstack-ansible15:02
*** kstev has joined #openstack-ansible15:26
*** armaan has quit IRC15:30
*** indistylo has quit IRC15:46
openstackgerritMerged openstack/openstack-ansible-os_nova master: Updated from global requirements  https://review.openstack.org/53506715:47
*** sxc731 has joined #openstack-ansible15:53
gillesMoAnyone using Ceph Luminous deployed with openstack-ansible (>pike) ?15:54
evrardjpgillesMo: yeah there are ppl like logan- andymccr15:57
shananigansVery happy to see the ssh retries being updated as its a pain on every build.  Upping the hosts MaxStartups and MaxSessions seems to help a lot in the ssh connection issues as well.15:58
gillesMoevrardjp: thanks15:58
andymccri know logan is using it in prod, but if you have mgr services that is not good - the mgr services should run on the mon hosts15:58
evrardjpshananigans: oh so you have the issue too ?15:58
evrardjpshananigans: do you know if https://review.openstack.org/#/c/545844/1 could fix your issue?15:58
shananigansI see ssh issues all the time.  On every new environment I usually up the retries, timeout and set the maxsessions and maxstartups.15:59
evrardjpthe MaxSessions and MaxStartups are your host things, so we shouldn't change that15:59
evrardjpbut the sessions/forks are something we could do15:59
evrardjpshananigans: could you review this, and give us your opinion? https://review.openstack.org/#/c/545844/116:00
gillesMoandymccr: I see that when I lauvnh ceph-install.yml, it tries to configure and launch it in the ceph mon container. But it fails because it tries to copy a non-existant keyring in /var/lib/ceph/mgr/ceph-*/keyring16:00
shananigansIt is the source of the issue being worked around by the retries from my experience.16:00
evrardjpmm interesting. Forks isn't the cause?16:00
gillesMoandymccr: when you say mon hosts, you really mean bare metal ?16:01
evrardjpshananigans: We have a bug triage right now, but your input is extremely valuable.16:01
shananigansThe higher the forks, the more ssh connections to all of the lxc containers on the host.  Back when originally looking into this, it looks like the host was dropping connections once it hit the session/startup limits.16:01
andymccrgillesMo: no more wherever you have specified your mon service to exist, that could be metal, but i think the default is containers for that16:01
evrardjpthat's until I wrote a more efficient connection plugin16:01
shananiganssorry.16:01
evrardjpshananigans: exactly16:01
evrardjpso limiting the amount of forks is generally good16:02
evrardjpno need for retries.16:02
andymccrgillesMo: the group-vars are set to make the mgr_group_name = mon_group_name (i.e. the mgr service will get installed on the same hosts as the mon service)16:02
gillesMoandymccr: yes, that's what I see. But ceph-mgr can't start because of an empty keyring...16:03
andymccrgillesMo: hmm16:03
andymccrgillesMo: so the service is there but it just wont start properly?16:03
gillesMoYes16:03
evrardjpshananigans: https://docs.openstack.org/openstack-ansible/latest/admin/maintenance-tasks.html#ansible-forks16:03
evrardjpBug triage cloudnull, DimGR, andymccr, d34dh0r53, hughsaunders, b3rnard0, palendae, odyssey4me, serverascode, rromans, erikmwilson, mancdaz, _shaps_, BjoernT, claco, echiu, dstanek, jwagner, ayoung, prometheanfire, evrardjp, arbrandes, mhayden, scarlisle, luckyinva, ntt, javeriak, spotz, vdo, jmccrory, alextricity25, jasondotstar, admin0, michaelgugino, ametts, bgmccollum, darrenc, JRobinson__,16:04
evrardjpasettle, colinmcnamara, thorst, adreznec, eil397, qwang,nishpatwa_, cathrichardson, drifterza, hwoarang, cshen, ullbeking16:04
evrardjpHere is our bug list for today https://etherpad.openstack.org/p/osa-bugtriage16:04
evrardjp#startmeeting openstack_ansible_meeting16:04
openstackMeeting started Tue Feb 20 16:04:15 2018 UTC and is due to finish in 60 minutes.  The chair is evrardjp. Information about MeetBot at http://wiki.debian.org/MeetBot.16:04
openstackUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.16:04
*** openstack changes topic to " (Meeting topic: openstack_ansible_meeting)"16:04
d34dh0r53o/16:04
openstackThe meeting name has been set to 'openstack_ansible_meeting'16:04
hwoarangο/16:04
evrardjpgood, we are already 3.16:04
jmccroryo/16:05
evrardjp4!16:05
evrardjpomg!16:05
andymccro/16:05
hwoaranglol16:05
hwoarangquick before we disappear16:05
andymccrhaha16:05
openstackgerritMerged openstack/openstack-ansible-os_nova master: Change include: to include_tasks:  https://review.openstack.org/54498616:06
evrardjpok let's move on to the agenda then!16:06
evrardjp#topic focus of the week16:07
*** openstack changes topic to "focus of the week (Meeting topic: openstack_ansible_meeting)"16:07
evrardjpthis week is!16:07
evrardjpdrumroll....16:07
evrardjpWrapping up Newton, stabilization of queens branch by fixing bugs16:07
evrardjpso basically newton is close of EOL16:07
evrardjpI will send a message to ML soon to do a last warning sign :)16:07
*** indistylo has joined #openstack-ansible16:08
evrardjpfor the rest, I think it would be nice to fix the upgrades to queens and improve stability of queens in general.16:08
evrardjpok let's move on to bug triage16:08
evrardjp#topic bugtriage16:08
*** openstack changes topic to "bugtriage (Meeting topic: openstack_ansible_meeting)"16:08
evrardjp#link https://bugs.launchpad.net/openstack-ansible/+bug/175024116:08
openstackLaunchpad bug 1750241 in openstack-ansible "creating OS::Neutron::FloatingIP with OS::Neutron::LBaaS::LoadBalancer" [Undecided,New]16:08
evrardjpI don't understand this. Is that our thing?16:09
* hwoarang has no clue16:09
evrardjpIt looks like it's usage16:10
evrardjpany idea?16:11
evrardjpshould I say invalid?16:11
evrardjpincomplete?16:11
andymccrhmm16:11
andymccrthats a heat template16:11
evrardjpyeah16:11
evrardjpit looks like a depends on something16:11
evrardjpwhich we don't do, and probably shouldn't16:11
andymccrim not sure what the issue is at all16:11
evrardjpahah welcome to the club16:12
andymccrincomplete?16:12
evrardjpyeah incomplete16:12
evrardjpI asked a question there16:12
evrardjp#link https://bugs.launchpad.net/openstack-ansible/+bug/175023616:13
openstackLaunchpad bug 1750236 in openstack-ansible "os::aodh::alarm via heat stable/pike ubuntu & centos http 503" [Undecided,New]16:13
andymccrok so since its the same reporter16:13
evrardjpomg.16:13
andymccrim guessing the issue is that using heat with OSA right now is causing issues?16:13
andymccrbut tl;dr aodh is broken there id guess :D16:13
evrardjpwith gnocchi and autoscaling.16:13
evrardjpso I guess unless someone has the time to confirm, we should move to another bug16:14
evrardjpok for everyone?16:14
evrardjpI assume yes.16:15
evrardjpnext16:15
evrardjp#link https://bugs.launchpad.net/openstack-ansible/+bug/175023316:15
openstackLaunchpad bug 1750233 in openstack-ansible "corrupted dynamic_inventory.py backup file" [Undecided,New]16:15
evrardjpok this one is VERY nice.16:15
odyssey4meI suspect those heat issues relate to bad config - endpoints for example.16:16
evrardjpI want to work on it, but I am still lacking cycles, so I'd like to classify this, so we know what we can do.16:16
odyssey4meI don't know heat very well, but it does require access to valid endpoints, and theire showing gateway errors and internal server errors.16:16
odyssey4meheh, yeah - our inventory is designed for single-use only16:17
evrardjpodyssey4me: can you comment on the bug, to ask for more detailed information maybe? Or should I do it?16:17
odyssey4memultiple accessors at the same time *will* break it16:17
evrardjpodyssey4me: yeah, that inventory thing show how many issues we will hit if we do things in parallel.16:17
spotzheat does require valid endpoints16:17
odyssey4mewe should probably implement some sort of lock file or whatever16:17
odyssey4meor not allow it to change itself ;)16:18
evrardjpor moved to static inventory, or safe system16:18
evrardjpyeah16:18
evrardjpso16:18
evrardjpI think this is a feature addition, which is kinda whishlist16:18
evrardjpalthough, this is a SEVERE bug in usability16:18
andymccrhmm16:19
andymccrcan i ask what the usecase is there?16:21
andymccri guess its still broken but hmm16:21
evrardjpyou can ask :D16:21
matttthe usecase of multiple simultaneous ansible runs?16:21
*** pcaruana has quit IRC16:21
andymccrmattt: ok so the aim is to run separate things at the same time? e.g. deploy nova and something else in the inventory at the same time?16:21
andymccrthat kinda makes sense16:22
andymccr(this is why we need a distributed inventory!) :D16:22
matttandymccr: :)  yeah not entirely sure, maybe shananigans can chime in if he's free16:23
evrardjpnah we don't need distributed inventory, we need it to be multithread safe.16:23
evrardjpdon't go for too complex :)16:23
odyssey4meandymccr it could also be multiple people doing multiple things at the same time, but with the same inventory16:23
andymccrwe could avoid doing the tar in the script and instead create a backups dir that gets tarred up once per day as part of a cron?16:23
odyssey4mefor example - jenkins is executing the upgrade, meanwhile I'm doing some maintenance16:23
openstackgerritGerman Eichberger proposed openstack/openstack-ansible-os_octavia master: Fixes Lint errors and improve tests  https://review.openstack.org/54411716:24
andymccrthat way we only tar once and each run will create its own backup inventory16:24
andymccrso that should avoid corruption16:24
odyssey4meif we're worried just about the tarball, we could just timestamp it and maintain a history of x versions16:24
andymccri think we do - but the tarball is a tar of those to save space16:24
evrardjpyeah that's what we do indeed.16:24
andymccrso we create timestamped backup confs that we then tar up, and we untar add the new one and tar up again on each run16:24
andymccrat least thats my idea16:25
odyssey4mebut the issue is actually that the inventory itself is a static file, and that file can't really be safely modified at the same time by multiple executions of the script... but in this case that's happening16:25
evrardjpthat's correct16:25
*** acormier has quit IRC16:25
andymccrso if we just dont tar as part of the inventory script16:25
evrardjpwell16:25
evrardjpwe just append I think16:25
andymccrhmm16:25
*** acormier has joined #openstack-ansible16:25
andymccrok so the issue is the inventory script itself16:25
andymccrthats a problem then16:25
evrardjpI think this should be moved out of the inventory16:25
evrardjpyeah that's definitely inventory script failure there16:25
odyssey4mewow, simpler would just be to have the 'openstac-ansible' wrapper do the tarballing and take it out of the inventory script16:25
andymccryeah ^ that'd work16:26
andymccrbut i think it sounds like its still broken16:26
andymccrhmm16:26
evrardjpwell let's just triage this first16:26
evrardjpinstead of thinking of solutions16:27
evrardjpis that a new feature we want to provide, or do we think it's a bug?16:27
andymccryou could classify as either16:27
andymccrwe didnt design it to be run multiple times as jesse said16:27
evrardjpIMO it's a new feature we provide, but it proves a big bug we have pending on our noses.16:27
andymccryeh16:27
evrardjpso I'd like to classify this as something medium or high.16:28
odyssey4meI'd say this is high.16:28
odyssey4meConfirmed.16:29
odyssey4meThe bug problem itself can be worked around with a band-aid, but the feature request for multi-user inventory should be registered separately.16:29
evrardjpok I validate this then.16:29
evrardjpI agree.16:29
odyssey4meThat feature request is not actually asked for here though, so don't comment that.16:29
odyssey4meWe're deriving a feature request based on the bug.16:30
shananigansmatt, andymcrr: Sorry, was in a meeting.  We were seeing bug 1750233 in a larger environment where multiple admins are running various plays to get information from the servers.16:30
openstackbug 1750233 in openstack-ansible "corrupted dynamic_inventory.py backup file" [High,Confirmed] https://launchpad.net/bugs/175023316:30
evrardjpwell it's the first line of the bug.16:30
evrardjpok let's move on then, now that we have triaged the bug16:30
matttshananigans: yeah, in my mind i could imagine a deployment running ansible routinely to obtain information about the deployment, and hitting this race conditino16:30
evrardjpnext16:31
evrardjp#link https://bugs.launchpad.net/openstack-ansible/+bug/174999016:31
openstackLaunchpad bug 1749990 in openstack-ansible "Fails to update the a-r-r file properly when stable branch is used" [Undecided,New]16:31
evrardjpI think this can be marked as confirmed and low.16:32
evrardjpok for everyone ?16:32
evrardjplet's move on, we have many bugs open today.16:32
evrardjp#link https://bugs.launchpad.net/openstack-ansible/+bug/174968016:32
openstackLaunchpad bug 1749680 in openstack-ansible "Ensure apt operations have retries" [Undecided,New]16:32
hwoarangok16:32
hwoarangdont know what to say about that. by the same argument, every network operation should have a retry option16:33
evrardjpprobably yes.16:33
evrardjpansible is notoriously unreliable.16:33
*** sxc731 has quit IRC16:33
evrardjpwow.16:34
evrardjpthat's not what I meant without proper context.16:34
evrardjpI think that it's possible that you have a process issue if a network disruption happens.16:34
hwoarangand this bug shouldn cover just apt16:35
hwoarangevery network op is subject to this problem16:35
hwoarangoverall i am not sure i agree this is a real bug16:35
evrardjpRetrying is fair on modules that don't implement retries, or for modules whose operations are dependant on somethink that's prone to connectivity issues.16:35
andymccryeah16:35
evrardjphwoarang: that's fair too16:36
andymccryou could argue that if apt module needs retries all the time it should be hardcoded in ansible.16:36
*** udesale has quit IRC16:36
hwoarangevery module is subject to connectivity issues since it talks to remote hosts16:36
andymccrbut if we are doing it everywhere but in a few places it seems sensible to have consistency as well16:36
evrardjpyup16:36
evrardjphwoarang: I mean without counting the connection plugin in itself16:36
odyssey4meit's not a bug, its a wishlist item16:37
hwoarangyeah16:37
evrardjpI think the linting bring consistency16:37
odyssey4meit happens to cause failures, so it matters to us16:37
evrardjpit's a wishlist indeed.16:37
odyssey4mea lint test would help us enforce such a thing, yes16:37
evrardjpwell, I understand both positions -- invalid for the fact it's not really a bug because we shouldn't do it -- and whishlist for consistency16:38
odyssey4mebasically, if we did this, deployments would be more reliable in their results16:38
evrardjpI'd prefer to classify this as whishlist and we can think about improving ansible in the future.16:38
evrardjpjust so you know -- ansible-devel also has lots of flaky tests when installing packages.16:39
andymccryeah im all for consistency tbh so i wouldnt be against adding it in as a feature16:39
evrardjpso there might be improvements coming upstream in the future too, not that it's currently planned.16:39
evrardjplet's mark it as wishlist16:39
evrardjpnext16:40
evrardjp#link https://bugs.launchpad.net/openstack-ansible/+bug/174925516:40
openstackLaunchpad bug 1749255 in openstack-ansible "haproxy healthcheck for API placement is broken" [Undecided,New]16:40
andymccri swear this thing breaks once a cycle16:40
evrardjphaha yeah it broke recently in queens.16:40
andymccrsigh16:41
odyssey4meyeah, every cycle and sometimes in the cycle they change an interface16:41
evrardjpI fixed it I think, so this should be fixed released.16:41
andymccrok cool16:41
andymccrnext!16:41
odyssey4methe return code, or the path16:41
evrardjpI fixed it before the bug was posted I think :p16:41
evrardjprc16:41
evrardjpnext16:41
evrardjp#link https://bugs.launchpad.net/openstack-ansible/+bug/174908316:41
openstackLaunchpad bug 1749083 in openstack-ansible "Nova, Glance, Cinder, ... downtime during O to P upgrade" [Undecided,New]16:41
evrardjpoh yeah. I am on it.16:42
evrardjpwhen not busy on other things.16:42
evrardjpConfirmed and high, ok for everyone?16:42
evrardjpit's not breaking gates, it's counting that's wrong16:42
odyssey4meyeah16:43
evrardjpok moving on16:43
evrardjp#link https://bugs.launchpad.net/openstack-ansible/+bug/174895116:43
openstackLaunchpad bug 1748951 in openstack-ansible "Use default sysctl_file in openstack_hosts" [Undecided,New]16:43
evrardjpwishlist?16:43
matttdefinitely wishlist16:43
evrardjphe is overriding the file, so overriding our defaults. I think he should be using the module :)16:43
evrardjpthat would make things work in an idempotent way16:44
evrardjpok commented.16:45
evrardjpnext16:45
evrardjp#link https://bugs.launchpad.net/openstack-ansible/+bug/174768416:45
openstackLaunchpad bug 1747684 in openstack-ansible "Default Values Do Not Allow Image Uploads to Glance from Horizon" [Undecided,New]16:45
evrardjpmaybe we should ship with different defaults?16:46
evrardjpthe changes in horizon in Pike state the default is to use external endpoints16:46
evrardjpso it would make sense to work on this16:47
mattti'll need 30 mins to process that bug, so no comment16:47
evrardjpand provide better defaults.16:47
mattt:)16:47
odyssey4meevrardjp if we use the public endpoint by default, due to the self-signed cert it will be broken out the box16:47
odyssey4meso to change that we have to ensure that we also configure the cert CA16:48
odyssey4meultimately, yes, we should use the public endpoint by default16:48
jwitko_ Hey All, is it possible to specify the index of the product of an intersected group of hosts? For example,  if i do  "hosts: group1:&group2[0]",  this intersects group1 with the index0 host from group216:48
evrardjpunless the default should be to use HTTP.16:48
odyssey4mebecause the endpoints shown to the user should be public, not internal16:48
logan-i think the self signed cert might be breaking it anyway16:48
odyssey4mewe could do that, but then we'll not catch bugs where the wrong endpoint is being used16:48
logan-because iirc horizon fixed the "bug" for direct uploads and it now directs the client to use public glance endpoints regardless of what endpoints horizon uses16:49
odyssey4meso yes - the intent is good, but some plumbing needs to be done16:49
evrardjpis there no insecure flags that can be set?16:49
odyssey4mehaha, well that's always fun16:49
odyssey4methere probably is - not sure, but that's not doing anyone any favors either16:50
evrardjpI think it's the same conversation over and over again16:50
evrardjpdoes someone deserve to run openstack if he hasn't valid certs?16:50
evrardjphahaha16:50
evrardjpwhat do we do?16:51
evrardjpand who could work on this?16:51
logan-in the case where self signed certs are used, legacy mode will work fine. maybe we should make sure our developer and AIO example configs reflect that16:51
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_nova stable/queens: Change include: to include_tasks:  https://review.openstack.org/54623116:51
evrardjpFYI: https://review.openstack.org/#/c/525491/16:51
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_nova stable/queens: Remove systemd conditionals  https://review.openstack.org/54623216:51
openstackgerritMajor Hayden proposed openstack/openstack-ansible master: Enable profile_tasks callback for AIO bootstrap  https://review.openstack.org/54623316:53
evrardjpso what's the situation ,  what do we decide?16:53
evrardjpThat looks like confirmed -- not so sure how we're gonna classify this16:53
logan-mediumish? the defaults need improvement but with configuration (easiest way is to set upload_mode legacy) this can be easily worked around16:54
evrardjpok16:55
*** chyka has joined #openstack-ansible16:55
evrardjpnext16:56
evrardjp#link https://bugs.launchpad.net/openstack-ansible/+bug/174762916:56
openstackLaunchpad bug 1747629 in openstack-ansible "A worker was found in dead state" [Undecided,New]16:56
evrardjpI think it's valid to cleanup the translations build, in order to make it pass.16:57
evrardjpConfirmed and high.16:57
evrardjpok for everyone?16:57
odyssey4meyeah, perhaps better to move to using the non-container build for it to save up some resource usage16:58
odyssey4meand look through all the services to see if we've ensured that they're all properly constrained16:58
matttwhat exactly does the translations job do?16:58
odyssey4meie there should not be a default number of workers/threads/processes for each service - but instead a smaller set as per the AIO config16:58
odyssey4memattt it installs every service that has a horizon plugin, and is used by the translations team to validate whether the language conversations are working right16:59
evrardjpok let's do a last one for today17:00
evrardjp#link https://bugs.launchpad.net/openstack-ansible/+bug/174528117:00
openstackLaunchpad bug 1745281 in openstack-ansible "galera_server : Create galera users fails on CentOS7" [Undecided,New] - Assigned to Jean-Philippe Evrard (jean-philippe-evrard)17:00
evrardjpwhat17:00
*** weezS has joined #openstack-ansible17:00
evrardjpwhy is this assigned to me?17:00
matttyeah i've seen that bug17:00
matttthat thing is valid17:00
evrardjpyeah17:01
evrardjpmhayden: told me the case was important17:01
odyssey4meevrardjp heh, it shows you assigned yourself :p17:01
evrardjpodyssey4me: yeah during the galera issue period17:01
evrardjpbut that's something different17:01
evrardjpI'll assign that to mhayden17:01
evrardjpanyone against?17:01
evrardjp:D17:01
odyssey4meyup, something is hinky there - I remember mnaser also wondering what the heck was going on17:02
evrardjpyeah17:02
evrardjpit was in galera_client at that time17:02
odyssey4meif mhayden or mgagne can pick that up it'd be nice17:02
matttevrardjp: no objections, i already asked him to look at it :P17:02
evrardjpbut now the link is for galera_server17:02
evrardjpit looks like the package name case doesn't matter after all17:02
evrardjpmattt: hahah great.17:02
evrardjplet's wrap up for today.17:03
evrardjpthanks everyone!17:03
matttthanks!17:03
evrardjp#endmeeting17:03
weezSmorning17:03
*** openstack 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"17:03
openstackMeeting ended Tue Feb 20 17:03:38 2018 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)17:03
openstackMinutes:        http://eavesdrop.openstack.org/meetings/openstack_ansible_meeting/2018/openstack_ansible_meeting.2018-02-20-16.04.html17:03
openstackMinutes (text): http://eavesdrop.openstack.org/meetings/openstack_ansible_meeting/2018/openstack_ansible_meeting.2018-02-20-16.04.txt17:03
openstackLog:            http://eavesdrop.openstack.org/meetings/openstack_ansible_meeting/2018/openstack_ansible_meeting.2018-02-20-16.04.log.html17:03
odyssey4melooks like zuul may have just croaked17:04
openstackgerritMajor Hayden proposed openstack/openstack-ansible-lxc_hosts master: Install common packages into container cache  https://review.openstack.org/54623917:04
evrardjpodyssey4me: zuul was kinda super flaky since saturday17:05
odyssey4meapparently they changed zuul web yesterday, so clear browser cache if it's not working right for you17:06
odyssey4meany takers to review https://review.openstack.org/545991 ? it fixes centos/suse17:08
cloudnullodyssey4me: done17:11
odyssey4methx17:12
spotzodyssey4me: done17:14
*** dariko has joined #openstack-ansible17:14
*** indistylo has quit IRC17:15
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_build master: [WIP] Make the git clone and wheel build optional  https://review.openstack.org/53701517:16
openstackgerritweezer su proposed openstack/openstack-ansible-ops master: Add force: true to avoid dirty git repo  https://review.openstack.org/54545717:19
*** sxc731 has joined #openstack-ansible17:23
*** epalper has quit IRC17:24
*** sm806 has quit IRC17:24
*** weezS has quit IRC17:25
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_build master: [WIP] Make the git clone and wheel build optional  https://review.openstack.org/53701517:27
openstackgerritMerged openstack/openstack-ansible-openstack_hosts master: Remove host_need_pip  https://review.openstack.org/54569817:28
*** pcaruana has joined #openstack-ansible17:36
hwoarangodyssey4me: https://review.openstack.org/#/c/545886/ only queens?17:37
hwoarangnot a backport?17:37
evrardjpfair point17:38
odyssey4mehwoarang it is: https://review.openstack.org/#/q/I9daa4352aa818db03f682eb0d1a65eefff9bb6f617:38
odyssey4mebut it was adjusted, so not a clean pick17:38
evrardjpit's missing the indication17:38
evrardjpok17:38
hwoarangah ok17:38
odyssey4meit wouldn't clean pick as the branch names are different17:38
hwoarangi should have clicked the id instead17:38
evrardjpwell17:39
evrardjpwe generally write manual cherry-pick17:39
evrardjpat least I do it :D17:39
odyssey4methe change id is enough of a reference17:39
hwoarangyeah cherry-pick just makes it obvious it's a backport :D17:40
hwoarangfor lazy people like me :)17:40
evrardjpI remembered the first patch, so it's different :D17:40
odyssey4mewell, fair enough, meh - it's not released, so it's hardly worth a discussion17:40
evrardjpI think it's good17:41
evrardjpbut for once, someone found you doing a mistake (which isn't really one), it's hard to not remember that to you odyssey4me  :p17:41
* odyssey4me makes mistakes all the time :p17:41
evrardjpelse you'd be closer to robot than human17:41
* hwoarang opened a can of worms17:41
evrardjphahaha17:42
evrardjpyou both make me laugh.17:42
odyssey4mein other news, if you haven't discovered https://howchoo.com/g/ytkwotvkztq/using-the-iterm-2-and-tmux-integration yet, then you haven't lived17:43
evrardjpI think I will stop working for today.17:43
evrardjpthat's not new17:43
* odyssey4me is working on his latest megapatch: https://review.openstack.org/#/c/537015/17:43
evrardjpodyssey4me: we discussed this together on top of it!17:44
evrardjpwoot mega patches17:44
odyssey4meyes, I know it's not new - it's my regular love affair with it :)17:44
evrardjpI am working on a series of patches myself.17:44
evrardjpit's smaller but if you merge them together, what a bunch!17:44
openstackgerritJimmy McCrory proposed openstack/openstack-ansible-plugins master: Make connection plugin compatible with Ansible 2.5  https://review.openstack.org/54357617:44
evrardjpand jmccrory is in da place! woot :)17:44
odyssey4methe funny thing about that patch is - once I put one more patch into the integrated build, I'll be deleting somewhere around 800 lines from repo_build ;)17:45
openstackgerritJimmy McCrory proposed openstack/openstack-ansible master: [TEST] Update Ansible to 2.5.0b2  https://review.openstack.org/50181417:45
odyssey4meso I add some, so that we can lose some of what was added, but most of the rest of it17:45
evrardjpodyssey4me: that will be nice!17:46
odyssey4mestep 1 - get repo-build to be more understandable, step 2 - reduce dependencies on it, step 3 - make the builds happen in the roles instead, step 4 - switch models so that repo_build no longer needs to exist at all17:50
odyssey4meI might just managed to get step 1 done by the PTG at this rate. :)17:50
evrardjpthat's good, because we can speak about making roles more generic at the ptg, so linked to step 3... Knowing is half the battle, right?17:54
evrardjp:D17:54
jmccroryodyssey4me wow had no idea iterm2 could do that, awesome17:56
evrardjpjmccrory: it's flaky though :p17:59
openstackgerritMerged openstack/openstack-ansible-repo_build master: Ensure virtualenv is not broken for CentOS/SuSE too  https://review.openstack.org/54599117:59
evrardjpI had so many issues with the -CC in the past I stoped trying.17:59
evrardjpnew sessions are fine, it's when restoring from (deep) sleep18:00
evrardjpbut I might be doing it wrong18:00
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_build stable/queens: Ensure virtualenv is not broken for CentOS/SuSE too  https://review.openstack.org/54625918:02
*** openstackgerrit has quit IRC18:03
jmccroryevrardjp hmm i'll have to see how my sessions look tomorrow then, seems to be working alright for now at least18:04
*** dariko has quit IRC18:06
*** gillesMo has quit IRC18:07
*** openstackgerrit has joined #openstack-ansible18:11
openstackgerritMerged openstack/openstack-ansible-os_tacker stable/queens: Update installation mechanism  https://review.openstack.org/54588618:11
*** pcaruana has quit IRC18:16
*** dariko has joined #openstack-ansible18:18
odyssey4meevrardjp it's not been flaky at all for me18:21
odyssey4mebut the nice thing is that you can tmux the normal way on a fkaly session, then switch to using -CC when you're on a good one18:21
odyssey4meyour window layout stays18:21
*** sxc731 has quit IRC18:22
odyssey4mejmccrory :) ^18:22
odyssey4mehas anyone else seen this dbu error popping up in tests lately? http://logs.openstack.org/15/537015/15/check/openstack-ansible-functional_selective-ubuntu-xenial/09b4064/logs/ara/result/3f774412-746a-4f27-8547-1b3612264bc6/18:23
evrardjpmaybe since the changes that involved dbus?18:23
evrardjpI am off for today18:23
evrardjpI can check tomorrow if you file a bug18:24
odyssey4meevrardjp I suspect that it might just be the role not running something the way it's expected, but it's not a consistent failure... was wondering if it's been seen elsewhere.18:25
*** sxc731 has joined #openstack-ansible18:29
*** weezS has joined #openstack-ansible18:30
odyssey4mejmccrory cloudnull I wonder if this makes any sense to you... I don't know the depths of systemd/lxc/veths/etc well enough...18:34
odyssey4meI've seen this pop up in some tests: http://logs.openstack.org/15/537015/15/check/openstack-ansible-functional_selective-ubuntu-xenial/09b4064/logs/ara/result/3f774412-746a-4f27-8547-1b3612264bc6/18:35
odyssey4meit's a module failure, which makes it look like the connection between the host and container got lost.18:35
odyssey4methat happens to be happening at around this time: http://logs.openstack.org/15/537015/15/check/openstack-ansible-functional_selective-ubuntu-xenial/09b4064/logs/host/syslog.txt.gz#_Feb_20_17_52_0018:36
cloudnullodyssey4me: yup thats thing18:36
odyssey4meand just before it in http://logs.openstack.org/15/537015/15/check/openstack-ansible-functional_selective-ubuntu-xenial/09b4064/logs/host/syslog.txt.gz#_Feb_20_17_51_57 it looks like there's some sort of mac error showing up18:36
openstackgerritPeriyasamy Palanisamy proposed openstack/openstack-ansible master: Make Opendaylight as the BGP speaker using Quagga  https://review.openstack.org/52390718:36
cloudnullits trying to read from /var/lib/dbus18:36
odyssey4mewhich led me to https://github.com/systemd/systemd/issues/337418:36
cloudnullI've been trying to alleviate some of that within the containers with https://review.openstack.org/#/c/523218/18:37
cloudnullbut have not had a lot of time to circle back on it18:37
odyssey4meah yes, it's been a while since I've seen that coming up... maybe time to get to it?18:37
cloudnullalso we should have a less of those iussues with the recent machine-id changes18:38
odyssey4meit's a bit unfortunate that it's not made queens though :(18:38
cloudnullmaybe this weekend while flying :)18:38
cloudnullodyssey4me: once we can stabalize that set of prs we can easily cherry-pick them back18:40
cloudnulli just need to fix up suse a18:40
cloudnull*and it should be happy18:41
odyssey4mesure, but the longer it takes to stabilize, the harder it is to remember the right bits to pick back18:41
odyssey4mein other news - thank you dmsimard for ARA... finding the failed task using the web interface is so simple... find->'failed'-> profit!18:42
openstackgerritMerged openstack/openstack-ansible-os_octavia stable/queens: Stop running non-voting jobs in gate  https://review.openstack.org/54533418:42
*** shardy has quit IRC18:43
*** dariko has quit IRC18:44
cloudnull++18:44
*** hamzy has quit IRC18:44
*** hamzy has joined #openstack-ansible18:48
* dmsimard ara sense tingling18:50
openstackgerritMajor Hayden proposed openstack/openstack-ansible-tests master: Set SELinux to permissive mode for tests  https://review.openstack.org/54615318:53
*** weezS has quit IRC18:54
dmsimardmhayden: how dare you18:55
mhayden:|18:55
openstackgerritMajor Hayden proposed openstack/openstack-ansible-tests master: Use ARA instead of profile_tasks callback  https://review.openstack.org/54627018:56
mhaydendmsimard: does that please you? ^^18:56
dmsimardthat's pretty cool yeah18:56
dmsimardyou're forgiven18:57
*** sxc731 has quit IRC18:57
openstackgerritMajor Hayden proposed openstack/openstack-ansible master: Use ARA instead of profile_tasks callback  https://review.openstack.org/54627118:57
mhaydendmsimard: twice as nice? ^^18:57
*** hamza21 has joined #openstack-ansible18:57
mhaydenodyssey4me: ^^18:57
dmsimardmhayden: profile_tasks is kinda cool anyway though18:57
mhaydentrue18:58
mhaydenbut odyssey4me told me i am a bad person for expanding its usage ;)18:58
dmsimardI get asked a lot about how to enable ARA in OSA, someone is allegedly writing a blog post about it18:58
*** sar has joined #openstack-ansible18:59
mhaydenyay18:59
dmsimardmhayden: btw are you using the RDO openstack-selinux package ?19:00
odyssey4methere's a cheat sheet in our gating script ;)19:00
dmsimardodyssey4me: yeah but apparently it's a bit awkward to do outside the context of CI, I dunno19:00
mhaydendmsimard: i actually just removed openstack-selinux/container-selinux since their policies didn't match with our deployment19:00
dmsimardmhayden: probably because they're meant to work with the RPM packaging19:00
mhaydenand they took forever to install due to the old way CentOS 7 ocmpiles selinux policies19:00
mhaydenFedoras is much faster19:01
dmsimardmhayden: if you want to look at the source, it's here: https://github.com/redhat-openstack/openstack-selinux19:01
odyssey4memhayden you may wish to add the explanation from https://review.openstack.org/#/c/546270/1 into https://review.openstack.org/#/c/546271/ so that it's clearer why19:01
mhaydenain't nobody got time for that19:01
* mhayden cackles19:02
odyssey4medmsimard hmm, perhaps a nice docs entry would be useful then :)19:02
odyssey4meif the blog post is not yet up by the PTG, then perhaps we'll take the time to document it :)19:02
openstackgerritMajor Hayden proposed openstack/openstack-ansible master: Use ARA instead of profile_tasks callback  https://review.openstack.org/54627119:02
mhaydenfixed, odyssey4me19:02
*** dariko has joined #openstack-ansible19:03
odyssey4meI see what you did there19:03
dmsimardodyssey4me: I'll ping you when it's out :)19:03
*** poopcat has joined #openstack-ansible19:07
mhaydenodyssey4me: do we have a docs page i could link to?19:07
mhaydenor are you insinuating i should cobble one together19:07
* mhayden squints19:07
odyssey4memhayden once dmsimard has a blog post, there will be one :)19:07
odyssey4meor we (you) could do one first ;)19:08
* mhayden adds "Depends-On: dmsimard-blog-post"19:08
odyssey4melol19:08
* odyssey4me likes19:08
odyssey4memhayden as it's not urgent, we can let that lie and see if someone volunteers... or maybe we'll do it at the PTG19:08
odyssey4meor maybe I'll do it this w/end :p19:09
dmsimardwell, someone did tell me they were going to write something soon so I'll wait a bit and let you know if I don't hear back19:10
*** armaan has joined #openstack-ansible19:11
*** weezS has joined #openstack-ansible19:16
*** weezS has quit IRC19:19
*** hamzy has quit IRC19:19
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_build master: [WIP] Make the git clone and wheel build optional  https://review.openstack.org/53701519:20
*** stuartgr has quit IRC19:23
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_build master: [WIP] Make the git clone and wheel build optional  https://review.openstack.org/53701519:24
mhaydenholy moley, i got a CentOS 7 deployment done in exactly 120 minutes with the aio_basekit from andymccr19:24
andymccrYES!19:24
mhaydenwith success even19:24
mhayden- [ Run Time = 7245 seconds || 120 minutes ] -------------------------19:24
mhayden- [ Status: Success ] ------------------------------------------------19:24
andymccr\o/19:24
mhaydenAND THERE WAS MUCH REJOICING19:24
mhaydendistro package installs are still painful :|19:25
mnaserany cores around to +W this patch which already has +W but zuul doing zuul things - https://review.openstack.org/#/c/543467/ (to avoid rechecking)19:25
mnaserif we get another +W it'll get into gate19:25
mhaydenmnaser: looks like it's workflowed19:25
*** hamzy has joined #openstack-ansible19:25
mhaydenmnaser: i put extra workflow on it to show zuul who's the boss19:26
mnasermhayden: yeah but it's stuck and not in gate (probably lost with one of the zuul restarts)19:26
mnaserand it's in gate now, thank you mhayden :)19:26
mhaydenhooray!19:26
mhaydengalera package installs are painful19:26
mnaserwe should just use mongodb tbh its web scale19:26
mhaydentrolololol19:27
mhaydenwe could shave 5 minutes if i could avoid pulling in python-monascaclient19:27
mnaseractually though i always wondered it would be nice to experiment with that database that was presented in boston keynote19:27
mhaydencockroachdb?19:27
mnaserit talks mysql but uses some distributed storage19:27
mnaseryes19:27
mhaydenif heat would switch to install_requires, that'd be nice19:28
* mhayden likes how gnocchi does that19:29
odyssey4memhayden  a report from a timeout: http://logs.openstack.org/49/545849/1/check/openstack-ansible-deploy-ceph-centos-7/d2c484e/logs/ara/ :)19:30
mhaydenah okay19:30
odyssey4megiven that the ARA report is done in the post section of the zuul execution, it now collects all the things19:31
odyssey4meevrardjp methinks we should add mnaser as core for the telemetry roles...19:32
mnaseri don't mind keeping an eye on telemetry stack reviews19:33
odyssey4methat would be marvellous19:33
odyssey4mewith your work, perhaps we could add some telemetry tests into the periodics so that they're actively tested again19:33
mnaserodyssey4me: i think the one step before that is os_panko but i can't remember who mentioned that they already started it19:33
odyssey4meah yes, gokhan perhaps?19:34
odyssey4mecloudnull did you push up a patch to import those extra roles? perhaps we should do a batch19:34
mnaserdon't remember who exactly :( but yeah, that would be nice to add to the stack19:34
cloudnulli have not.19:35
cloudnulli suck19:35
odyssey4mecloudnull I can put something together now.19:36
odyssey4mecloudnull doing it live!19:38
evrardjp[m]me also thinks odyssey4me19:38
odyssey4meevrardjp want me to do it now?19:39
evrardjp[m]if he is okay yeah. i can do it too, i have to add mbuil to neutron too.19:41
odyssey4meI can do that for you now if you'd like. Seems sensible to me.19:41
odyssey4memnaser also, confirmed that gokhan has the panko role: http://eavesdrop.openstack.org/irclogs/%23openstack-ansible/%23openstack-ansible.2018-02-06.log.html#t2018-02-06T14:06:2319:41
evrardjp[m]i was busy with all governance stuff today19:42
odyssey4megokhan we need to revisit that conversation - happy to pull the role in if you'r happy to go ahead? we might want a few tweaks (like flattening the git history)19:42
odyssey4meit doesn't matter if it's broken - it's easy enough to get fixed once we have people working on it as a group.19:43
evrardjp[m]i still have a etherpad ;)19:44
mnaseryeah once it's in, we can fix the issues with proper gate checks19:45
mnaserand i'm okay with it19:45
odyssey4meawesome, thanks mnaser - I'll verify that the gerrit ACL's are properly set as I'm doing a housekeeping patch anyway to pull cloudnull's nspawn roles in19:45
mnaserodyssey4me: ok cool, i'm going to setup a gerrit dashboard for this then, aodh+ceilometer+gnocchi+panko ?19:47
odyssey4memnaser yep, and maybe horizon for the related plugins there?19:47
* odyssey4me has a long gerrit query: project:"^(openstack|openstack-infra)/(monitorstack|openstack-ansible|ansible-|devstack-gate).*"19:48
mnaserodyssey4me: sure, i can do horizon reviews if you want (i have the pleasure of working with it often, ha) or leave them19:48
mnaserodyssey4me: i use multiple dashboards for different projects/components19:48
mnaserex: puppet/project config/etc19:49
odyssey4memakes sense19:52
odyssey4meI basically have a very large 'to review' queue, then a focused one which removes -1 reviews and zuul failures, then another one which shows those owned by me, then another which has my starred reviews, then the merged ones (for context and proactive backporting).19:53
odyssey4meeach is a tab in the browser19:53
odyssey4memaybe we should do a review practise sharing session at the PTG :)19:53
openstackgerritMohammed Naser proposed openstack/openstack-ansible-os_gnocchi stable/pike: Zuul: Remove project name  https://review.openstack.org/54210319:56
*** dave-mccowan has quit IRC19:57
openstackgerritMohammed Naser proposed openstack/openstack-ansible-os_gnocchi stable/queens: Zuul: Remove project name  https://review.openstack.org/54532119:58
openstackgerritMohammed Naser proposed openstack/openstack-ansible-os_gnocchi master: Change default gnocchi ceph pool name to metrics  https://review.openstack.org/51231319:58
odyssey4memhayden you've had some contributors help out with some things - would you want a core group outside of the standard OSA core group?19:59
odyssey4mefor ansible-hardening I mean19:59
mhaydenat the moment, i dunno19:59
mhaydenit could be helpful in the long term19:59
odyssey4meok, I'll leave it then19:59
odyssey4methe option is always there, it's just a project-config patch away20:00
mnaserandymccr: do you have sometime to do the fixups needed here or would you like me to pick it up? https://review.openstack.org/#/c/492580/20:01
odyssey4memnaser I'm quite certain that andymccr would be happy for you to take over that patch.20:03
openstackgerritMohammed Naser proposed openstack/openstack-ansible-os_aodh master: Standardize services list and package installation  https://review.openstack.org/49258020:04
odyssey4mehe was doing a sweep at one point to try and normalise some of the less well looked after roles, and that was part of it20:04
mnasercool :)20:05
*** sxc731 has joined #openstack-ansible20:06
openstackgerritMohammed Naser proposed openstack/openstack-ansible-os_ceilometer stable/pike: [pike only] Stop using 'latest' for pymongo installation  https://review.openstack.org/54629320:07
openstackgerritMohammed Naser proposed openstack/openstack-ansible-os_ceilometer stable/pike: Zuul: Remove project name  https://review.openstack.org/54260120:08
mnaserwould anyone historically know the reason behind this? https://review.openstack.org/#/q/topic:move-out-telem -- is this not necessary anymore because we're maintaining them now?20:10
odyssey4memnaser yep - basically there was one guy trying to maintain them, and he's gone now20:15
odyssey4mebut he was trying to move them out from the integrated build to reduce the effects on our ongoing gating20:15
odyssey4mebut if we now have people picking those up, we can abandon those20:15
mnaseryeah i think we can abandon them because it looks like the author hasn't pushed up anything for the last 7 months20:17
*** mbuil has quit IRC20:28
odyssey4memnaser You're now a core reviewer for os_aodh, os_ceilometer... os_gnocchi will be on the way once we have a dedicated core group for the repo done20:28
odyssey4meSad to see stevelle and alextricity25 go though.20:28
* odyssey4me pours one out for them.20:28
mnaserodyssey4me: thanks and :(20:30
odyssey4mewelcome, and thank you20:30
odyssey4meyou are free to add more cores to the repo core list, just please add the openstack-ansible-core group there as a back-stop to help out with any issues20:31
odyssey4mewe added some core reviewer guidelines stuff to the docs here, for bedtime reading ;) https://docs.openstack.org/openstack-ansible/latest/contributor/core-reviewers.html#general-responsibilities20:31
odyssey4meplease do feel very free to propose changes to those20:32
*** poopcat has quit IRC20:32
*** poopcat has joined #openstack-ansible20:33
*** dave-mccowan has joined #openstack-ansible20:37
*** pcaruana has joined #openstack-ansible20:40
*** oneswig has joined #openstack-ansible20:55
*** dave-mccowan has quit IRC20:56
*** hamza21 has quit IRC21:02
*** flaviosr_ has quit IRC21:12
openstackgerritMerged openstack/openstack-ansible stable/pike: Fix cinder_service_region undefined  https://review.openstack.org/54346721:19
*** armaan has quit IRC21:21
openstackgerritMajor Hayden proposed openstack/openstack-ansible-galera_client master: Allow Galera package downloads over HTTP  https://review.openstack.org/54630821:21
*** armaan has joined #openstack-ansible21:21
openstackgerritMajor Hayden proposed openstack/openstack-ansible-galera_server master: Allow Galera package downloads over HTTP  https://review.openstack.org/54630921:22
mhaydenthat was a fun find ^^21:22
mhaydeni knew yum was slow, but it wasn't THAT slow21:22
openstackgerritMajor Hayden proposed openstack/openstack-ansible master: CentOS 7 integrated gate optimization  https://review.openstack.org/54545521:23
openstackgerritMajor Hayden proposed openstack/openstack-ansible master: Add base kit scenario  https://review.openstack.org/54353421:23
openstackgerritMajor Hayden proposed openstack/openstack-ansible master: CentOS 7 integrated gate optimization  https://review.openstack.org/54545521:24
*** flaviosr_ has joined #openstack-ansible21:25
openstackgerritgit-harry proposed openstack/openstack-ansible-galera_server master: Fix cache update after initial apt_repository fail  https://review.openstack.org/54631221:28
mhaydencloudnull: i'm looking into trimming heat's requirements.txt on the repo server...21:29
mhaydenwe do a selective wheel build but then heat drags in clients that we don't need21:29
mhaydenit would be heat-specific code, which feels dirty :/21:29
*** tux_ has joined #openstack-ansible21:30
*** mbuil has joined #openstack-ansible21:31
tux_Hello Folks!21:33
mhaydenhowdy tux_21:33
tux_Hey Major :) let me tell you i love your blog21:34
mhaydenhah thanks!21:34
* mhayden needs to update it more often ;)21:34
tux_Anyway! i am planning to build openstack in production using OSA and collecting some info,21:34
tux_I have 3 controller node but what about HA Proxy node? do i need separate node?21:35
tux_How do you guys deal with HA proxy node in production?21:35
*** pcaruana has quit IRC21:37
andymccrmajors blog is the best! :)21:38
tux_silence is painful ;)21:39
tux_Can i use HA proxy in all 3 controller node or anything else i need to do?21:42
shananigansYou can set up all three infra nodes with an external and internal floating ip using keepalived in the OSA configs.  https://docs.openstack.org/openstack-ansible-haproxy_server/latest/configure-haproxy.html21:47
shananigansIf you check out the link the 'haproxy_hosts' will determine where haproxy gets installed and the keepalived settings is were you configure your floating ips tied to specific interfaces.21:48
mhaydentux_: sorry, got dragged away21:50
mhaydeni have four nodes running OSA -- each one is a hypervisor, but they run openstack services, too21:50
mhaydenhyperconverged!21:50
mhaydeni picked two of my four to run haproxy21:50
*** mbuil has quit IRC21:51
tux_so that is do-able? i can pick 01 and 02 and setup HA proxy on them.21:51
tux_Does it required hand edit config or OSA does itself?21:51
tux_Does it required hand edit config or OSA does itself?  ( I meant can i pick 2 controller node)21:55
openstackgerritMerged openstack/openstack-ansible stable/queens: Ensure that constraints are used for AIO bootstrap  https://review.openstack.org/54589121:57
*** armaan has quit IRC22:00
shananigansIt is doable and I think the haproxy_server ansible role handles all of the changes based on the OSA configs.22:00
*** sar has quit IRC22:02
*** oneswig has quit IRC22:05
*** sxc731 has quit IRC22:05
openstackgerritMajor Hayden proposed openstack/openstack-ansible-repo_build master: [WIP] Remove unneeded clients from heat  https://review.openstack.org/54631922:08
*** dariko has quit IRC22:09
*** dariko has joined #openstack-ansible22:10
tux_@shananigans - Thanks!22:10
openstackgerritMajor Hayden proposed openstack/openstack-ansible master: [WIP] Test repo build w/trimmed heat requirements  https://review.openstack.org/54633122:10
mhaydencloudnull: i'm a bad person https://review.openstack.org/#/c/546319/22:11
openstackgerritMajor Hayden proposed openstack/openstack-ansible-repo_build master: [WIP] Remove unneeded clients from heat  https://review.openstack.org/54631922:11
*** pbandark has quit IRC22:12
mnaserany cores who would like to push https://review.openstack.org/#/c/546293/ ?22:19
*** threestrands has joined #openstack-ansible22:23
*** dariko has quit IRC22:23
*** kstev has quit IRC22:24
openstackgerritGerman Eichberger proposed openstack/openstack-ansible-os_octavia master: Fixes Lint errors and improve tests  https://review.openstack.org/54411722:26
*** jwitko__ has joined #openstack-ansible22:29
*** jwitko_ has quit IRC22:33
*** tux_ has quit IRC22:33
*** jwitko__ has quit IRC22:36
*** jwitko_ has joined #openstack-ansible23:00
*** Smeared_Beard has quit IRC23:02
*** pmannidi has joined #openstack-ansible23:03
*** openstackgerrit has quit IRC23:04
*** NostawRm has quit IRC23:15
*** NostawRm has joined #openstack-ansible23:19
*** acormier has quit IRC23:35
*** taseer1 has joined #openstack-ansible23:38
*** Taseer has quit IRC23:40
*** Tahvok has quit IRC23:46
*** Tahvok has joined #openstack-ansible23:47
*** tux_ has joined #openstack-ansible23:49
*** taseer2 has joined #openstack-ansible23:51
*** tux_ has quit IRC23:54
*** taseer1 has quit IRC23:54

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!