Thursday, 2017-03-16

*** DimGR2 has quit IRC00:00
*** DimGR has joined #openstack-ansible00:00
*** pmannidi has quit IRC00:11
*** pmannidi has joined #openstack-ansible00:11
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_keystone master: Source template files from git or deploy host  https://review.openstack.org/44623500:13
*** japestinho has joined #openstack-ansible00:22
*** pwnall138 is now known as zz_pwnall13800:33
*** weezS has quit IRC00:33
*** cathrichardson has quit IRC00:34
*** woodard has quit IRC00:35
*** woodard has joined #openstack-ansible00:36
*** woodard has quit IRC00:52
*** deadnull has quit IRC00:52
*** cuongnv has joined #openstack-ansible00:59
*** SerenaFeng has joined #openstack-ansible01:01
*** foutatoro has quit IRC01:04
*** sanfern has quit IRC01:04
*** sanfern has joined #openstack-ansible01:05
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible stable/ocata: Add better support for multi-OS deployments  https://review.openstack.org/44625501:06
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible-repo_build stable/newton: Modified repo build for better multi-OS support  https://review.openstack.org/44625601:17
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible-repo_build stable/ocata: Modified repo build for better multi-OS support  https://review.openstack.org/44625701:18
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible stable/newton: Add better support for multi-OS deployments  https://review.openstack.org/44625801:24
*** weezS has joined #openstack-ansible01:24
cloudnullanyone around that might want to give this a push https://review.openstack.org/#/c/444252/01:24
openstackgerritMerged openstack/openstack-ansible-haproxy_server stable/ocata: Add http-check options to haproxy service template  https://review.openstack.org/44618701:28
openstackgerritMerged openstack/openstack-ansible-haproxy_server stable/newton: Add http-check options to haproxy service template  https://review.openstack.org/44618601:32
*** sanfern has quit IRC01:39
*** gouthamr has quit IRC01:59
*** crushil has joined #openstack-ansible02:02
*** marst has joined #openstack-ansible02:14
*** kjw3 has quit IRC02:18
*** fandi has quit IRC02:26
*** rmelero has quit IRC02:38
*** rmelero has joined #openstack-ansible02:39
*** rmelero has quit IRC02:43
openstackgerritMerged openstack/openstack-ansible-rsyslog_client stable/ocata: Logrotate not restarting rsyslog  https://review.openstack.org/44425202:51
*** Mahe has quit IRC02:51
*** Mahe has joined #openstack-ansible02:53
*** jwitk0 has joined #openstack-ansible02:59
*** jrobinson has quit IRC03:01
*** jasondotstar has quit IRC03:25
*** jasondotstar has joined #openstack-ansible03:26
*** fandi has joined #openstack-ansible03:33
*** SerenaFeng has quit IRC03:47
*** udesale has joined #openstack-ansible03:48
*** mpotdar has quit IRC03:59
*** NikhilS has joined #openstack-ansible04:01
*** sanfern has joined #openstack-ansible04:03
*** jascott1- has joined #openstack-ansible04:05
*** poopcat has quit IRC04:07
*** mpotdar has joined #openstack-ansible04:11
*** SerenaFeng has joined #openstack-ansible04:11
*** cuongnv has quit IRC04:16
*** cuongnv has joined #openstack-ansible04:16
*** shasha_t_ has joined #openstack-ansible04:21
*** hybridpollo has quit IRC04:22
*** acormier has joined #openstack-ansible04:27
*** acormier has quit IRC04:31
*** chhavi has joined #openstack-ansible04:47
*** SerenaFeng has quit IRC04:49
*** haasn has quit IRC05:00
*** neith has quit IRC05:00
*** neith has joined #openstack-ansible05:01
*** haasn has joined #openstack-ansible05:06
*** chhavi has quit IRC05:07
*** chhavi has joined #openstack-ansible05:13
*** neith has quit IRC05:14
*** neith has joined #openstack-ansible05:15
*** Amit82 has joined #openstack-ansible05:28
*** Jeffrey4l__ has joined #openstack-ansible05:30
*** udesale has quit IRC05:32
*** udesale has joined #openstack-ansible05:32
*** Jeffrey4l_ has quit IRC05:34
*** shausy has joined #openstack-ansible05:37
*** SerenaFeng has joined #openstack-ansible05:44
*** Jack_Iv has joined #openstack-ansible05:50
*** Jack_Iv has quit IRC05:54
*** Jack_Iv has joined #openstack-ansible05:54
*** Jack_Iv_ has joined #openstack-ansible05:56
*** Jack_Iv has quit IRC05:59
*** chhavi has quit IRC06:16
Amit82Hi All06:18
*** lwiecek has joined #openstack-ansible06:26
*** david-lyle_ has joined #openstack-ansible06:26
*** david-lyle has quit IRC06:26
*** weezS has quit IRC06:28
*** weezS has joined #openstack-ansible06:29
*** chhavi has joined #openstack-ansible06:32
*** jrobinson has joined #openstack-ansible06:34
*** SerenaFeng has quit IRC06:45
*** SerenaFeng has joined #openstack-ansible06:52
*** chhavi has quit IRC06:54
*** arbrandes1 has joined #openstack-ansible07:01
*** arbrandes has quit IRC07:03
*** fandi has quit IRC07:15
*** muxdaemon has joined #openstack-ansible07:26
*** manheim has joined #openstack-ansible07:27
*** manheim has quit IRC07:28
*** manheim has joined #openstack-ansible07:28
*** pmannidi has quit IRC07:29
*** sprvn has joined #openstack-ansible07:35
*** askb has quit IRC07:38
*** udesale has quit IRC07:40
*** Jeffrey4l__ has quit IRC07:41
*** chhavi has joined #openstack-ansible07:52
*** chhavi has quit IRC07:58
*** shardy has joined #openstack-ansible08:09
*** jrobinson has quit IRC08:22
*** acormier has joined #openstack-ansible08:28
*** pcaruana has joined #openstack-ansible08:30
*** acormier has quit IRC08:33
*** ArchiFleKs has quit IRC08:36
*** chhavi has joined #openstack-ansible08:40
*** shausy has quit IRC08:50
*** shausy has joined #openstack-ansible08:50
*** deadnull has joined #openstack-ansible08:55
*** david-lyle_ has quit IRC08:58
*** DimGR has quit IRC08:59
*** qiliang28 has joined #openstack-ansible08:59
*** DimGR has joined #openstack-ansible09:00
*** qiliang27 has quit IRC09:02
*** qiliang28 is now known as qiliang2709:02
*** chhavi has quit IRC09:03
admin0good morning .. on newton AIO build with telemetry, how do I auth/see gnocci metrics ?09:04
*** Disova_ has joined #openstack-ansible09:05
*** jbadiapa_ has quit IRC09:11
*** Jack_Iv_ has quit IRC09:16
*** Jack_Iv has joined #openstack-ansible09:17
openstackgerritAndy McCrae proposed openstack/openstack-ansible-os_neutron master: Update paste, policy and rootwrap configurations 2017-03-16  https://review.openstack.org/44638109:20
*** Amit82 has quit IRC09:23
*** Amit82 has joined #openstack-ansible09:23
Amit82Hi All09:24
*** karimb has joined #openstack-ansible09:24
Amit82I have deployed Openstack Newton release using Openstack-Ansible09:24
Amit82I have 1 Compute and 1 Controller Node in my setup. Each physical node is having two nics. eth0 of each machine is connected to a switch for communication b/w compute and controller node. eth1 of each machine is connected to my lab n/w (192.168.255.XXX).09:24
Amit82I want to ping/access my external lab n/w through my VMs but not able to do so.09:25
*** vnogin has joined #openstack-ansible09:25
*** SerenaFeng has quit IRC09:25
*** Jeffrey4l has joined #openstack-ansible09:25
Amit82I had to comment br-vlan "flat" configuration from my openstack_user_config.yml and re-configure neutron09:26
Amit82reason was that eth12 used in flat n/w was non-existent on my compute node09:26
Amit82so I had removed09:26
admin0Amit82: how does your config yml looks like ?09:26
Amit82paste.openstack.org is not working?09:27
admin0gist ?09:29
muxdaemonAmit82 I had the same issue - had to set host_bind_override to the actual NIC on the physical compute host09:31
Amit82my openstack_user_config.yml file is here: https://onetimesecret.com/secret/qqh63iiozf7k6hoapuezk3x3r072vkq09:31
Amit82muxdaemon: I have two NICs on each compute and Controller09:33
Amit82so, should I give container_interface: "eth1" and host_bind_override: "eth1"?09:34
muxdaemontwo NICs for the VM networks?09:34
Amit82eth1 is the NIC connected to external n/w09:34
Amit82I have 1 Compute and 1 Controller Node in my setup. Each physical node is having two nics. eth0 of each machine is connected to a switch for communication b/w compute and controller node. eth1 of each machine is connected to my lab n/w (192.168.255.XXX).09:34
muxdaemoncontainer interface stays as is is, host_bind_overide to whatever ip link shows the nic name to be on the compute node09:35
muxdaemonso from the looks of that, yes, eth109:35
muxdaemonhttps://gist.github.com/anonymous/1d5097c951c2e05a9e9383e47c17498a09:35
muxdaemonis what I have09:35
Amit82so, br-vlan in /etc/network/interfaces/ will have bridge-ports as eth0 or eth1 in my case?09:37
muxdaemonno, just br-vlan I think09:37
Amit82are you able to install and access external machines through VMs?09:37
muxdaemonThat is to be tested in the next 30 minutes :)09:37
Amit82aha...do let me know your results :)09:39
Amit82btw this is my /etc/network/interfaces file on Controller node: https://gist.github.com/anonymous/5c5fa98ebcb20936c51d78f3b1b1484a09:40
Amit82it is having br-vlan with bridge-port as eth009:40
Amit82?09:40
Amit82what does your file have09:40
Amit82do you have two NICs for each Controller and Compute?09:41
muxdaemonI have 3 NICs - one mgmt, one external, one with 2 VLANs for storage and vxlan09:41
muxdaemonlet me try getting a VM up09:41
muxdaemonsee if it works09:42
*** Disova_ has quit IRC09:42
openstackgerritAndy McCrae proposed openstack/openstack-ansible master: Update all SHAs for Pike 2017-03-16  https://review.openstack.org/44639709:45
*** kysse has quit IRC09:52
*** kysse has joined #openstack-ansible09:54
muxdaemonAmit82: I'm seeing similar to what you're getting - I can see ping packets go out the router but no reply from the outside world10:00
muxdaemonlooks like a network config issue outside of OS tho'10:02
*** jbadiapa has joined #openstack-ansible10:03
muxdaemonok, it was.10:03
muxdaemonit's working as expected now and I can ssh to the instance10:04
*** electrofelix has joined #openstack-ansible10:06
*** cuongnv has quit IRC10:06
Amit82muxdaemon: what was it?10:07
Amit82I mean the issue10:07
*** qiliang27 has quit IRC10:08
*** qiliang27 has joined #openstack-ansible10:10
Amit82muxdaemon: could you please provide your network connectivity diagram sort of thing, if you have10:10
muxdaemonissue was that I hadn't set the VLAN to be untagged to the port on br-vlan10:11
muxdaemonI don't have a diagram, but the port, eth1 in your case, is just a port on the network you want to use10:12
Amit82muxdaemon: could you provide you /etc/network/interfaces file from the nodes10:12
muxdaemonyep10:12
muxdaemontwo mins10:12
Amit82and also the openstack_user_config.yml10:12
Amit82thanks, probably it will help me in understanding the difference10:12
muxdaemonhttps://gist.github.com/anonymous/9e5d429b811d59fdda43e3b67961948f10:13
*** deadnull has quit IRC10:13
*** manheim_ has joined #openstack-ansible10:14
*** mpranjic has joined #openstack-ansible10:15
*** ArchiFleKs has joined #openstack-ansible10:15
*** manheim has quit IRC10:16
Amit82"ens3f0" is connected to your external n/w?10:16
muxdaemonyes10:17
Amit82Your IP's 10.42.XXX.XXX are internal to openstack environment, right10:19
Amit82?10:19
Amit82These have nothing to do with outside n/w, is it true?10:20
muxdaemonyes10:21
muxdaemonwell, 10.42.110.0/24 is the external network10:22
*** woodard has joined #openstack-ansible10:22
muxdaemon100-102 are mgmt, storage, vxlan10:22
muxdaemon110 was only set inside neutron when creating the subnet10:22
*** Jeffrey4l has quit IRC10:23
Amit82In my case, I am using the IPs (172.29.xxx.xxxx) provided in the example test environment as you can notice from my /etc/network/interfaces that I had shared10:23
muxdaemonI'm trying to set up glance to use swift. In the docs (for v13) glance_swift_store_auth_address: '{{ auth_identity_uri }}' is set, but auth_identity_uri doesn't exist. What should it be now?10:23
Amit82but my external n/w is on 192.168.xxx.xxx. Should that be creating any issue?10:24
muxdaemonso eth1 should be on br-vlan. eth1 should have no IP config10:25
andymccrmuxdaemon: you're using mitaka (so 13.x.x)?10:25
muxdaemonandymccr, v1510:25
muxdaemonI couldn't find docs for the current version10:26
andymccrmuxdaemon: ahh ok - lemme double check i thought we had set it so it would automatically use swift if you had a swift cluster10:26
*** woodard has quit IRC10:26
Amit82eth1 is having IP assigned from external lab n/w only10:27
*** Andrew_jedi has joined #openstack-ansible10:27
muxdaemonandymccr you're right I think - I see swift config in glance10:27
muxdaemoneth1 doesn't need an IP10:28
muxdaemonneutron will handle the IP allocation to the VMs10:28
Amit82This is required for me to ssh to Controller and Compute Node from lab n/w10:29
muxdaemonI shouldn't break anything, I don't think10:29
muxdaemonjust be sure to set the subnet IP range accordingly in the subnet for the external network10:30
Amit82hmm...it is fine10:30
muxdaemonbut yeah, eth1 should be on br-vlan and that might fix it10:31
Amit82Can you share openstack_user_config.yml file once?10:31
Amit82And I used these commands for creating the router, external net, tenant n/w: https://gist.github.com/anonymous/0e94505f55268a372caf2f3bdf12520010:32
Amit82do you find any issue with them?10:32
muxdaemonhttps://gist.github.com/anonymous/ec63ea80245b87aaf9237ef6b0ddb57a10:33
*** Andrew_jedi has quit IRC10:33
Amit82are you using the similar method for providing outside access to your VMs?10:33
muxdaemonI used the web ui10:33
Amit82but created the same things, I mean router etc.?10:35
Amit82what was your external n/w type? VLAN or FLAT?10:36
muxdaemonflat10:36
muxdaemonit's only me using it :)10:36
muxdaemonandymccr - I can see objects on the disk in swift that look about the right size for the images, so that working. Thanks.10:37
Amit82but do you think the way my configuration is in /etc/network/interfaces, I am not making any basic mistake10:38
Amit82?10:38
Amit82I am going to configure flat n/w in openstack_user_config.yml by only changing "host_bind_override" to "eth1"10:39
openstackgerritAndy McCrae proposed openstack/openstack-ansible stable/ocata: Update all SHAs for 15.0.1  https://review.openstack.org/44644510:39
ArchiFleKsHi, do you know if thereis a way to run everything in developer mode ? like openstack_developer_mode ? and bypass repo or is it somehting that has to be done on a per component basis ?10:40
Amit82and I will re-configure neutron and also add bridge-port as eth1 in br-vlan and see if it works10:40
Amit82One more question, tenant n/w being used by your VMs was VLAN based or FLAT?10:41
andymccrArchiFleKs: it looks like we only have it on a per repo basis - if you want, file a bug and it's a pretty easy one to setup - we already do something similar for endpoint protocols10:42
*** Jeffrey4l has joined #openstack-ansible10:46
*** sanfern has quit IRC10:48
muxdaemonAmit82 the tenant network was the default, which is vxlan10:48
muxdaemonthat's all handled by neutron10:48
*** sunjon_ has joined #openstack-ansible10:50
-openstackstatus- NOTICE: paste.openstack.org is down, due to connectivity issues with backend database. support ticket has been created.10:56
*** ChanServ changes topic to "paste.openstack.org is down, due to connectivity issues with backend database. support ticket has been created."10:56
*** vnogin has quit IRC11:01
*** Jack_Iv has quit IRC11:02
*** Jack_Iv has joined #openstack-ansible11:03
openstackgerritAndy McCrae proposed openstack/openstack-ansible stable/newton: Update all SHAs for 14.1.2  https://review.openstack.org/44645411:03
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-galera_server master: Upgrade: Make removal list distribution-specific  https://review.openstack.org/44312311:05
*** Jack_Iv has quit IRC11:08
odyssey4meArchiFleKs what do you hope to achieve by turning everything into developmer mode in an integrated build?11:09
openstackgerritAndy McCrae proposed openstack/openstack-ansible-os_aodh stable/mitaka: Update paste, policy and rootwrap configurations 2017-03-16  https://review.openstack.org/44645911:10
*** stuartgr has joined #openstack-ansible11:18
openstackgerritAndy McCrae proposed openstack/openstack-ansible stable/mitaka: Update all SHAs for 13.3.17  https://review.openstack.org/44646811:20
*** Jack_Iv has joined #openstack-ansible11:21
*** deadnull has joined #openstack-ansible11:24
odyssey4meandymccr if you have a moment, I'd like to discuss https://review.openstack.org/#/q/topic:bug/1641131 and how we approach releasing it into the stable branches11:25
odyssey4meevrardjp I like https://review.openstack.org/446219 but would prefer your thoughts as you know haproxy better11:27
*** sanfern has joined #openstack-ansible11:38
*** Jack_Iv has quit IRC11:39
odyssey4meevrardjp and others, I'd appreciate a review of https://review.openstack.org/443123 asap to resolve the next problem found in the periodic upgrade test11:39
evrardjpsure11:39
*** 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"11:44
-openstackstatus- NOTICE: paste.openstack.org service is back up - turns out it was a networking issue, not a database issue. yay networks!11:44
*** Jack_Iv has joined #openstack-ansible11:44
*** retreved has joined #openstack-ansible11:46
bwwAmit82, I am trying to solve the same use case for external access for AIO. Can you let me know if what you try works? I have tried several things and couldnt get it to work. We are trying to get the same exact thing working11:56
*** SerenaFeng has joined #openstack-ansible11:58
*** retreved has quit IRC12:00
mgariepygood morning everyone12:02
*** retreved has joined #openstack-ansible12:03
openstackgerritMerged openstack/openstack-ansible-openstack_hosts stable/ocata: Fix rdo package installation  https://review.openstack.org/44615712:04
ArchiFleKsodyssey4me: nothing special, just wondering ^^12:11
mgariepyinteresting http://logs.openstack.org/14/446214/1/check/gate-openstack-ansible-os_nova-ansible-func-ubuntu-xenial/b3056ee/logs/openstack/openstack1/neutron/neutron-l3-agent.log.txt.gz?level=ERROR12:12
Amit82bww: I am still stuck12:13
Amit82muxdaemon: when I am adding eth1 on br-vlan, I am not able to access Controller and Compute Node via their lab n/w IPs (192.168.255.xxx range IPs) on eth1.12:14
bwwok, i have tried several things as well and have been unable to get the current AIo to work with external access :(12:14
bwwand I know it wasnt designed for that but would be nice to have12:15
Amit82which release you are using?12:15
mgariepynice. https://github.com/openstack/neutron/commit/3b22541a2aa9a5b06e2bff256701dbe24554c17c12:16
*** schwicht has joined #openstack-ansible12:16
*** gouthamr has joined #openstack-ansible12:16
bwwstable/Mitaka12:17
*** woodard has joined #openstack-ansible12:19
*** woodard has quit IRC12:20
*** woodard has joined #openstack-ansible12:21
openstackgerritMarc Gariépy proposed openstack/openstack-ansible-os_neutron master: Fix neutron issue caused by upstream change  https://review.openstack.org/44648812:22
*** schwicht has quit IRC12:23
mgariepy^^ odyssey4me, evrardjp12:23
odyssey4memgariepy can you rebase that on top of https://review.openstack.org/#/c/446381/1 ?12:24
odyssey4meand approve that patch12:25
mgariepyha12:25
mgariepydamn :(12:25
mgariepyalready included in12:25
*** schwicht has joined #openstack-ansible12:26
mgariepyoops.12:26
mgariepymissing the haproxy stuff12:26
mgariepy** pkg.12:26
*** karimb has quit IRC12:28
openstackgerritMarc Gariépy proposed openstack/openstack-ansible-os_neutron master: Update paste, policy and rootwrap configurations 2017-03-16  https://review.openstack.org/44638112:28
mgariepyodyssey4me, ^^12:28
odyssey4meurgh, that's not what I meant :p12:29
muxdaemonAmit82 - it sounds like putting the br-mgmt on eth1 might be the way to go. So both br-mgmt and br-vlan or on eth1, eth0 is for the internal traffic12:29
*** esberglu has joined #openstack-ansible12:29
odyssey4meI meant that you can approve the patch as-is (it passed the tests), but rebase yours on it12:29
mgariepyhmm sorry,12:30
mgariepyi miss understood what you meant. should I revert then ?12:30
openstackgerritMerged openstack/openstack-ansible-galera_server master: Upgrade: Make removal list distribution-specific  https://review.openstack.org/44312312:31
* mgariepy need more coffee.12:32
Amit82muxdaemon: thanks, I will give it a try12:32
muxdaemonremember, networking is fun :)12:33
bwwAmit82: Let me know if that works for you12:34
Amit82bww: sure12:37
*** udesale has joined #openstack-ansible12:37
Amit82muxdaemon: :) for me it is getting fun out of control12:38
*** Jack_Iv has quit IRC12:38
openstackgerritLogan V proposed openstack/openstack-ansible-os_nova stable/newton: Fix default live migration configuration  https://review.openstack.org/44649812:39
*** NikhilS has quit IRC12:40
*** Jack_Iv has joined #openstack-ansible12:40
*** Jack_Iv has quit IRC12:44
*** SerenaFeng has quit IRC12:45
*** manheim has joined #openstack-ansible12:48
*** shardy is now known as shardy_lunch12:50
*** schwicht has quit IRC12:50
*** manheim_ has quit IRC12:51
openstackgerritAndy McCrae proposed openstack/openstack-ansible-os_cinder master: Add support for cinder v3 api  https://review.openstack.org/44650312:53
*** klamath has joined #openstack-ansible12:53
*** klamath has quit IRC12:54
*** sprvn has quit IRC12:54
*** klamath has joined #openstack-ansible12:54
*** sprvn has joined #openstack-ansible12:54
*** LinStatSDR has joined #openstack-ansible12:56
*** kylek3h has joined #openstack-ansible12:56
*** LinStatSDR has quit IRC13:05
*** LinStatSDR has joined #openstack-ansible13:05
*** schwicht has joined #openstack-ansible13:05
*** Amit82 has quit IRC13:06
openstackgerritAndy McCrae proposed openstack/openstack-ansible-os_nova master: Use the cinderv3 endpoint by default  https://review.openstack.org/44650813:06
*** LinStatSDR has quit IRC13:10
andymccrodyssey4me: will take a look when i get a chance - thanks for pointing it out13:10
*** shausy has quit IRC13:11
*** jbadiapa has quit IRC13:12
*** shausy has joined #openstack-ansible13:12
*** udesale has quit IRC13:14
*** foutatoro has joined #openstack-ansible13:15
ArchiFleKsxgerman: I managed to deployed Octavia :) fromwhat I understand if you also have the HAproxy driver installed in Neutron you can have both right ? in neutron.conf service_provider I have both, is there a way to choose which one you want to use when you create an LB ?13:20
foutatorohi all, I'm trying to create trusty image "openstack image create --file /home/alioune/custome-ubuntu-trusty.img   trusty "on my OSA platform13:20
foutatorobut I get this error : Error finding address for http://172.29.236.100:9292/v2/images/b1617ef7-a09a-441b-ae8e-4324bd866095/file: [Errno 32] Broken pipe13:21
xgermanyes, there is a —provider flag on some commands13:21
xgermanthere is also a way to list providers13:21
*** shausy has quit IRC13:21
xgermanoptionally you can define flavors with the flavor framework and pit to specific providers13:22
xgermanthen it would be —flavor13:22
foutatorosomeone knows how to solve that ,13:23
ArchiFleKsok thanks, do you have more info about the flavor stuff13:23
*** crushil has quit IRC13:24
xgermanthere shoudl be sufficient help with neutron help13:24
xgermanI remember there being a flavor-create command13:25
xgermanand then you cna use —flavor upon lbaas-loadbalancer-create13:25
*** cjloader has joined #openstack-ansible13:26
xgermannow, flavors are probably a bit unstable there is a proposal to enhance that drastically13:27
xgermanhttps://review.openstack.org/#/c/392485/13:28
openstackgerritMarkos Chandras proposed openstack/openstack-ansible-galera_server master: Add support for SUSE Leap distributions  https://review.openstack.org/44524813:29
*** karimb has joined #openstack-ansible13:30
*** shardy_lunch is now known as shardy13:31
*** NikhilS has joined #openstack-ansible13:31
*** Jeffrey4l has quit IRC13:34
*** jbadiapa has joined #openstack-ansible13:37
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-galera_server stable/ocata: Upgrade: Make removal list distribution-specific  https://review.openstack.org/44652013:41
*** LinStatSDR has joined #openstack-ansible13:41
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-galera_server stable/newton: Upgrade: Make removal list distribution-specific  https://review.openstack.org/44652213:45
odyssey4meandymccr I'm thinking that we should do a minor bump when https://review.openstack.org/#/q/topic:bug/1641131 has merged, but also include an upgrade release note to warn of the repo structure change.13:47
odyssey4methose patches need to go in to facilitate a multi-OS migration, for example being able to deploy to both Ubuntu Trusty and Ubuntu Xenial at the same time13:47
*** Jeffrey4l has joined #openstack-ansible13:47
andymccryeah i understood that one - i think its super important to get it in.13:47
andymccrodyssey4me: does it require a repo rebuild though?13:48
andymccre.g. what is the upgrade impact?13:48
*** crushil has joined #openstack-ansible13:48
odyssey4meandymccr all upgrades require a repo rebuild, so this isn't really any different13:51
odyssey4methe old data stays, it just adds a new structure13:51
andymccrodyssey4me: i meant more in branch13:51
andymccror true13:51
andymccryeah13:51
andymccrhmm13:51
odyssey4methe awareness needed is mainly just in case there is any downstream tooling using that structure, for example we're using the MANIFEST file and have tooling to rebuild that structure in the RPC staging process13:52
odyssey4methere may be others doing much the same13:52
odyssey4meI'd like to know if it impacts logan- in any way too13:53
andymccrahh ok that makes sense13:53
andymccrthat sounds like a sensible approach then more just a warning that hey if youre doing something this'll change13:53
odyssey4meyep13:54
odyssey4meI'll adjust cloudnull's patches to include the upgrade warning.13:54
odyssey4meI figure that since we've just done a release, we can now merge those and have two weeks to test before releasing this.13:54
*** FrankZhang has joined #openstack-ansible13:56
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron master: Fix neutron issue caused by upstream change  https://review.openstack.org/44648813:58
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron master: Update paste, policy and rootwrap configurations 2017-03-16  https://review.openstack.org/44638114:01
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron master: Fix neutron issue caused by upstream change  https://review.openstack.org/44648814:02
odyssey4meevrardjp ^ neutron metadata proxy is now using haproxy14:03
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-os_neutron master: Add haproxy to neutron distro packages  https://review.openstack.org/44648814:05
*** FrankZhang_ has joined #openstack-ansible14:05
odyssey4memgariepy ^ I rebased your patch on andymccr's patch and tweaked the commit message to be a bit more explicit.14:06
mgariepyodyssey4me, perfect thanks14:06
*** LinStatSDR has left #openstack-ansible14:07
foutatoroany suggestion about the glance image error ?14:08
admin0hi all .. . on newton AIO build with telemetry, how do I auth/see gnocci metrics ?14:09
*** FrankZhang has quit IRC14:09
admin0how/where14:10
*** FrankZhang_ has quit IRC14:11
*** FrankZhang has joined #openstack-ansible14:11
*** jmckind has joined #openstack-ansible14:15
cloudnullmornings14:17
mgariepygood morning cloudnull14:18
*** kstev has joined #openstack-ansible14:19
*** manheim has quit IRC14:19
cloudnullhows it mgariepy ?14:20
*** acormier has joined #openstack-ansible14:20
mgariepyfighting with selinux again :)14:20
andymccrodyssey4me: is there an easy way to delete a branch in an upstream repo without asking infra?14:21
andymccri know i can create a branch14:21
mgariepyand verifying that my python skills are very low.14:21
odyssey4meandymccr I honestly have no idea - you can always try using the git CLI and if it fails due to a permission issue, then ask infra14:21
andymccrhaha yeh fair14:21
odyssey4meandymccr https://gerrit-review.googlesource.com/Documentation/project-configuration.html#branch-deletion14:22
*** tonytan4ever has joined #openstack-ansible14:22
*** lucasxu has joined #openstack-ansible14:23
cloudnullmgariepy: that sounds like a blast !14:24
mgariepyhaha14:24
*** pcaruana has quit IRC14:24
andymccrbenefits of being at the ops midcycle is that you can ask people in person to delete random branches :P14:27
*** lwiecek has quit IRC14:27
*** udesale has joined #openstack-ansible14:29
cloudnullodyssey4me: are you creating the reno for the two multi-os patches or should i?14:29
odyssey4mecloudnull I'm happy to do a quick update now.14:29
odyssey4mecloudnull doing it live :)14:30
cloudnullhttps://www.youtube.com/watch?v=O_HyZ5aW76c14:30
*** udesale has quit IRC14:31
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_build stable/ocata: Modified repo build for better multi-OS support  https://review.openstack.org/44625714:31
*** udesale has joined #openstack-ansible14:31
openstackgerritJesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-repo_build stable/newton: Modified repo build for better multi-OS support  https://review.openstack.org/44625614:31
*** messy has quit IRC14:31
odyssey4meI think that's good enough.14:31
*** Guest69399 has joined #openstack-ansible14:32
odyssey4mecloudnull I think we should get https://review.openstack.org/#/q/topic:sha-update+status:open in before we merge those though14:32
*** messy has joined #openstack-ansible14:32
odyssey4mejust in case downstream consumers get caught in the middle14:32
cloudnullmakes sense14:33
odyssey4meI was also thinking that we discuss it at the community meeting this afternoon before we get it merged14:33
cloudnullalso makes sense14:35
odyssey4meandymccr FYI I also put this together last night to set a pattern for the policy/api-paste files - I'll do one for nova too which will also include the rootwrap files14:36
odyssey4mehttps://review.openstack.org/44623514:36
odyssey4meI just want to do the integrated build patch too and validate that it all works right before releasing it.14:37
odyssey4meI suspect the gnocchi one is quite broken because the way of fetching the file for the integrated repo is broken.14:37
odyssey4mecloudnull another one: https://review.openstack.org/44644514:39
*** charz_ has quit IRC14:43
openstackgerritMarc Gariépy proposed openstack/openstack-ansible-openstack_hosts master: set yum priority  https://review.openstack.org/44549314:44
*** cathrichardson has joined #openstack-ansible14:44
*** Jack_Iv has joined #openstack-ansible14:44
openstackgerritMarc Gariépy proposed openstack/openstack-ansible-pip_install master: set yum priority  https://review.openstack.org/44550214:45
*** tonytan4ever has quit IRC14:45
openstackgerritMerged openstack/openstack-ansible-os_aodh stable/mitaka: Update paste, policy and rootwrap configurations 2017-03-16  https://review.openstack.org/44645914:45
*** marst has quit IRC14:45
*** tonytan4ever has joined #openstack-ansible14:46
*** Disova_ has joined #openstack-ansible14:46
*** charz has joined #openstack-ansible14:46
*** logan- has quit IRC14:46
*** logan- has joined #openstack-ansible14:46
*** cathrich_ has joined #openstack-ansible14:47
openstackgerritMarc Gariépy proposed openstack/openstack-ansible-lxc_hosts master: set yum priority  https://review.openstack.org/44553514:47
*** manheim has joined #openstack-ansible14:47
*** kstev has left #openstack-ansible14:48
*** weezS has quit IRC14:49
*** cathrichardson has quit IRC14:50
*** manheim has quit IRC14:52
*** marst has joined #openstack-ansible14:53
cloudnullandymccr: we might want to also backport this to ocata, maybe newton https://review.openstack.org/#/c/446503 ?14:58
cloudnullwe might also want to do a follow up patch to remove v1. thoughts?14:58
andymccrcloudnull: yeah i agree on the backport.14:58
andymccrthe removal thing im good with too - we can just do that in master though. i really dont think there is a reason for v1 although it is apparently still there (but beyond deprecated)14:59
odyssey4meperhaps it'd be good if we ensure that it can be turned off by a var if the deployer wants to, but we leave the current state14:59
odyssey4meI'm talking in ocata/newton... in master we just rip it out.15:00
odyssey4mewe could perhaps deprecate it in ocata seeing as almost no-one is using it15:00
* odyssey4me looks at mhayden ;)15:00
andymccrodyssey4me: yeah agreed - you can already turn it off if you want there is an "enabled' var for each api version15:00
*** Disova_ has quit IRC15:01
odyssey4meah awesome, then maybe we turn it off by default in ocata and deprecate it15:01
odyssey4methen master we rip it out15:01
odyssey4meit'll help with our memory usage :)15:01
andymccri mean its not even listed on cinder docs anymore haha15:01
andymccrbut i asked and it's still "there"15:02
*** karimb has quit IRC15:02
andymccrbut yeah it makes sense we can just deprecate it and leave it as is for now but rip it out for pike15:02
odyssey4mewell, if we are going to do a minor rev for the repo change, perhaps we just add the removal in ocata to it too?15:02
andymccrsounds good to me15:02
odyssey4meI'm just concerned about knock on effects15:02
palendaeRemoving without deprecation is a bad idea IMO15:02
odyssey4mewe need a little time to figure those out and get them all squared away15:02
palendaeIt's not at all respectful of users15:03
*** cathrich_ is now known as cathrichardson15:04
*** karimb has joined #openstack-ansible15:04
spotzpalendae: So you're saying deprecate pike and remove queens?15:05
palendaespotz: I'm fine with Ocata15:05
spotzahh ok:)15:05
palendaeJust don't yank it with 0 warning15:05
openstackgerritMerged openstack/openstack-ansible-os_neutron master: Update paste, policy and rootwrap configurations 2017-03-16  https://review.openstack.org/44638115:06
*** phalmos has joined #openstack-ansible15:06
odyssey4mepalendae although, upstream has delivered the deprecation for years now15:08
palendaeI guess...but we've kept it the default despite that15:08
stevelle+1 for deprecate ocata, remove pike15:08
stevelletoo late to purge in pike15:08
stevelleerr too late to purge in ocata15:09
odyssey4meyeah, I agree15:09
odyssey4meI'm just playing devil's advocate15:09
stevelleappreciate having someone in the room check15:10
palendaeWe're admittedly in a crap scenario15:10
*** udesale has quit IRC15:10
palendaeBecause we're unable to keep in lock step with upstream deprecations15:10
stevelletempest tests for v1 are going away in pike, it looks like15:10
stevelleso it will bit rot, if cinder doesn't purge v1 in this cycle15:11
odyssey4mewell, when I last checked nova was still incompatible with glance v2 and cinder v215:11
odyssey4meI think, however, we switched to using cinder v2 at some point for nova?15:11
stevellenova switched to glance v215:11
stevellethat was newton iirc15:11
odyssey4meyeah, I remember doing that and then seeing on the ML that there were bugs and it wasn't ready... although we never saw them15:12
odyssey4meand I guess they got solved15:12
stevelleit had to, because they gutted their v1 consuming code15:12
odyssey4meI think I checked cinder at the time and it was problematic, so I didn't change it... but this is part of memory haze here which might be liberty timeframe.15:12
palendaeI do recall the scenario you're thinking of15:13
palendaeThere was talking of deprecating v1 APIs when consumers weren't yet on v215:13
odyssey4meyep15:13
palendaeBut yeah that was L-ish15:13
odyssey4meah ok :)15:13
odyssey4methanks!15:13
*** tonytan4ever has quit IRC15:14
*** tonytan4ever has joined #openstack-ansible15:14
*** acormier has quit IRC15:17
*** Jack_Iv has quit IRC15:18
*** weezS has joined #openstack-ansible15:19
*** acormier has joined #openstack-ansible15:19
*** chris_hultin|AWA is now known as chris_hultin15:19
*** Jack_Iv has joined #openstack-ansible15:19
*** sprvn has quit IRC15:19
*** lucasxu has quit IRC15:22
palendaestevelle, sigmavirus - If either of you has a moment today, I'd appreciate feedback on making https://review.openstack.org/#/c/446131/ a somewhat realistic test. Steve mentioned downloading an image as perhaps a better indicator15:22
*** Jack_Iv has quit IRC15:23
stevellewill look quickly15:24
*** Jack_Iv has joined #openstack-ansible15:24
odyssey4mepalendae image list tests the api/registry, image create/upload tests the api/registry/store write capability, image download tests the api/store read15:25
odyssey4meso it all depends on what you're wanting to see15:25
odyssey4meit may be worth doing them all15:25
palendaeodyssey4me: Looking for API. The thing that had me unsettled was that the test as written never failed as glance nodes restarted15:25
palendaeYeah, I intend to15:25
palendaeBut API's easiest for now15:26
sigmaviruspalendae: keep in mind that glance doesn't do checking of images15:26
palendaeRestarting keystone did, however, lead to failures, same as restarting just keystone15:27
sigmavirusso if you want, you can use an old glance trick of uploading a small plain text file and then downloading it15:27
palendaesigmavirus: ?15:27
palendaeOh, ok15:27
palendaeNice15:27
sigmavirusyeah15:27
sigmavirusit's a simple check of upload/download + integrity of both15:27
palendaeSweet, thanks15:27
palendaeI may refactor this so I only get the keystone information once15:27
*** acormier has quit IRC15:27
palendaeAnd re-use it in the loop15:27
odyssey4mepalendae have you seen https://github.com/Mirantis/kostyor-openstack-ansible ?15:28
palendaeNope15:28
odyssey4meas I understand it, kostyor is a chaos monkey framework, and that's the OSA driver for it15:28
odyssey4meI know about it thanks to vnogin15:29
odyssey4meI haven't looked too deeply though.15:29
palendaeLooks like it only runs upgrades, doesn't take stuff down arbitrarily15:30
sigmaviruspalendae: is bowlingball meant to work on AIOs etc?15:30
palendaesigmavirus: I'm using it on a multinode AIO right now. Presumably it could work on a normal one, though you'd need affinity to get the rolling effect vs just being down15:31
sigmaviruspalendae: fair15:31
palendaeMostly it's so I don't have to manually do the rolling restart and container name look up manually, though :)15:31
palendaeAnd then the test scripts are so I can somewhat quantify the effect15:31
sigmavirusnaturally15:32
odyssey4mepalendae I'm curious why you don't just use ansible for that?15:32
palendaeI'm not looking for high accuracy, more of a "does this thing blow up when we roll through"15:32
palendaeodyssey4me: I do15:32
palendaehttps://github.com/openstack/openstack-ansible-ops/blob/master/bowling_ball/rolling_restart.py#L30-L3315:33
odyssey4meansible -m lxc_container -a "state=restarted" --serial 1 keystone_all15:33
*** LinStatSDR has joined #openstack-ansible15:33
palendaeAh15:33
palendaeWell, this gives me a unified log format to collate15:33
odyssey4mesure, I guess you wanted somethings that won't give you15:34
odyssey4memakes sense then15:34
palendaeI didn't think about that, but also state=restarted restarts it right away, no?15:34
openstackgerritTom Jose Kalapura proposed openstack/openstack-ansible-rsyslog_client stable/ocata: Fix remote logging template  https://review.openstack.org/44660315:38
palendaeKostyor does look interesting though15:39
palendaeLooks like it's mostly for upgrades, not arbitrarily wrecking the environment15:40
stevelletl;dr what is it doing with upgrades palendae?15:42
openstackgerritMerged openstack/openstack-ansible-os_neutron master: Add haproxy to neutron distro packages  https://review.openstack.org/44648815:43
palendaestevelle: It's an orchestration framework that'll take a deployment method as a driver, then upgrade services in the recommended order15:43
palendaeUses celery workers15:43
stevelleany fanciness to how it upgrades a single service?15:44
*** galstrom_zzz is now known as galstrom15:45
palendaeI didn't dig that deeply15:45
palendaeHah, they have a craton inventory plugin15:46
palendaeOr at least proposed one15:47
*** agrebennikov has joined #openstack-ansible15:48
odyssey4meorchestrating the orchestrator, odd15:49
odyssey4meI certainly would prefer it if we bake everything into our plays15:49
odyssey4meideally a new deployment and an upgrade should have the same experience15:49
odyssey4meand a major or minor upgrade should be the same process15:49
palendaeI'm guessing it's a tool to present a common upgrade interface for Mirantis where they have Fuel and OSA installs15:50
*** FrankZhang has quit IRC15:50
odyssey4meyeah, fair enough15:50
palendaeBased on what I'm reading in their docs15:50
stevelleodyssey4me: I am going to disagree on the "install is special case of upgrade" bit still15:51
*** FrankZhang has joined #openstack-ansible15:51
*** furlongm has quit IRC15:51
mgariepyodyssey4me, you are too fast!15:52
stevelleibut it's a nit pick15:52
*** david-lyle has joined #openstack-ansible15:52
odyssey4mestevelle that statement counts in terms of process for our consumers, rather than the details in between15:52
*** lucasxu has joined #openstack-ansible15:53
odyssey4mebut yeah, when you have a large environment it's not really that easy to paint it all with the same brush as a small one15:53
odyssey4meI think it's a good aspiration though15:53
*** asettle has left #openstack-ansible15:55
*** asettle has joined #openstack-ansible15:55
*** david-lyle_ has joined #openstack-ansible15:55
asettlecloudnull, 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, admin0, michaelgugino, ametts, v1k0d3n, severion, bgmccollum, da15:55
asettlerrenc, JRobinson__, asettle, colinmcnamara, thorst, adreznec, eil397, qwang,nishpatwa_, cathrichardson, drifterza, sc68cal15:55
evrardjpda15:55
asettleMeeting in openstack-meeting-4 in 5 minutes!15:55
evrardjprrenc15:55
asettleSorry humans15:55
asettleGot distracted15:55
asettleIt'll be hosted by me, so naturally, will be all about me15:55
asettle;)15:55
*** david-lyle has quit IRC15:55
-openstackstatus- NOTICE: The Gerrit service on review.openstack.org is being restarted to address hung remote replication tasks, and should return to an operable state momentarily15:56
evrardjpwhen isn't it that?15:56
*** david-lyle__ has joined #openstack-ansible15:57
cloudnullcron job for restarts and be done with it :)15:57
evrardjphahaha cloudnull15:57
cloudnulljust do it 3x a day15:57
evrardjpI meant for asettle 'though15:57
*** acormier has joined #openstack-ansible15:57
evrardjpbut it works too.15:57
asettleevrardjp: RIGHT15:57
evrardjpYEAH15:58
*** marst_ has joined #openstack-ansible15:58
*** marst has quit IRC15:59
*** vnogin has joined #openstack-ansible15:59
*** david-lyle has joined #openstack-ansible15:59
*** david-lyle__ has quit IRC15:59
*** karimb has quit IRC16:00
asettleSeriously, meeting, evrardjp and odyssey4me - not time for the two of you to walk off16:00
* palendae leaves16:00
*** david-lyle_ has quit IRC16:00
asettlecloudnull, 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, admin0, michaelgugino, ametts, v1k0d3n, severion, bgmccollum, da16:01
asettlerrenc, JRobinson__, asettle, colinmcnamara, thorst, adreznec, eil397, qwang,nishpatwa_, cathrichardson, drifterza, sc68cal16:01
asettleMeeting now homies16:01
asettleMove yo selves16:01
asettleopenstack-meeting-416:01
*** chris_hultin is now known as chris_hultin|AWA16:01
*** vnogin has quit IRC16:02
*** karimb has joined #openstack-ansible16:02
*** vnogin has joined #openstack-ansible16:02
*** david-lyle has quit IRC16:05
*** kstev has joined #openstack-ansible16:05
*** japestinho has quit IRC16:10
*** karimb has quit IRC16:12
*** NikhilS has quit IRC16:14
*** karimb has joined #openstack-ansible16:17
*** marst_ has quit IRC16:20
*** marst has joined #openstack-ansible16:20
*** lucasxu has quit IRC16:30
agrebennikovevrardjp morning16:37
agrebennikovregarding your comment in https://review.openstack.org/#/c/446219/116:37
agrebennikovit is really an objection that all tunables will show up when you activate one?16:37
*** karimb has quit IRC16:38
evrardjpmorning! agrebennikov16:39
evrardjpwell the only objection is for backport16:39
evrardjpfor master only it's fine16:39
agrebennikov?16:39
agrebennikovhm16:39
agrebennikovthe whole point was to bring it to newton telling the truth16:39
evrardjpif you plan to backport this, you'll change the integrity of the file, which would trigger a restart of haproxy which would mean downtime for no value16:40
agrebennikovand re-use on the prod customer's cloud16:40
agrebennikovahh....16:40
agrebennikovyeah16:40
agrebennikovbut then16:40
evrardjpso if you don't modify the content of the file that would be great16:41
agrebennikovmaybe I can just customize ocata and newton backports a little bit and make just bufsize to be unique?16:41
agrebennikovlike maxconn16:41
agrebennikovwill that wirk?16:41
agrebennikov*work16:41
evrardjpyeah16:41
evrardjpworks for me16:41
agrebennikovis it allowed?16:42
agrebennikovI mean16:42
evrardjpwhy not?16:42
agrebennikovmake backports different what much16:42
agrebennikovdunno :)16:42
evrardjpwell we can change the original patchset too16:42
*** mpotdar_ has joined #openstack-ansible16:42
agrebennikovyou guys are driving the project16:42
evrardjpahah no we aren't16:42
evrardjpthe community is :p16:42
agrebennikovso you tell me what;s allowed and what's not16:42
*** mpotdar_ has quit IRC16:42
evrardjpmy opinion is if we backport I'd rather stick closer to what we have16:43
agrebennikovmakes perfect sense to me16:43
evrardjpbut yes, you can either adapt your patchset or adapt the backport16:43
agrebennikovwill do asasp16:43
evrardjpI don't mind, just tell me and I'll vote accordingly :p16:43
agrebennikovthat was my original plan actually - have them all activated together and avoid multiple if's16:43
agrebennikovsince if you know what you are doing and want to tune - everything shows up16:44
*** foutatoro has quit IRC16:45
agrebennikovor I can just make a loop through the params specified, but I'm confused there is a dot in the params names16:45
agrebennikovso I don't know how ansible will wotk if the key will have dot in it16:45
agrebennikovsince it maybe something like http://paste.openstack.org/show/602990/16:47
agrebennikovis it batter?16:47
agrebennikov*better?16:47
evrardjp" "16:48
evrardjpbut have a try16:48
evrardjpthe easiest is you keep bufsize out16:48
evrardjpand we don't speak about it again :p16:49
agrebennikovI mean in general - do you think this is what I have to do instead?16:49
evrardjp(for the backport at least)16:49
evrardjpit's your choice16:49
agrebennikovbufsize will be isolated in the backport16:49
evrardjpok16:49
agrebennikovgreat16:49
agrebennikovthanks!16:49
evrardjpI don't think we should care that much about the current status of this16:49
evrardjpwe should basically improve the jinja for the generation16:49
agrebennikovthat was the original concern that there is no way to ivverride haproxy params16:50
evrardjpand have a look at a dict of allowed directives and if it's user defined, then set it in the config16:50
evrardjpI'm all in for new features on haproxy. I just think the whole thing could deserves a rethink.16:50
evrardjpI had an etherpad on it16:50
agrebennikovthe whole think meaning what?16:51
agrebennikovhow the config is generated in general?16:51
agrebennikovglobal and services16:51
evrardjpyeah. I think haproxy could deserve a lot more modularity16:51
evrardjpI can't find the etherpad.16:52
agrebennikovwelll.... I'd say it doesn't worth it that much.... imho 95% of the cases don't need that16:52
evrardjpBut we can discuss that together if you're interested by stepping up16:52
*** vnogin has quit IRC16:52
agrebennikovplease, just show me what you have and maybe I could be useful for that16:52
*** Jack_Iv has quit IRC16:56
*** Jack_Iv has joined #openstack-ansible16:56
openstackgerritMerged openstack/openstack-ansible-haproxy_server master: Added Haproxy global tunables  https://review.openstack.org/44621916:57
*** muxdaemon has quit IRC16:59
*** Jack_Iv has quit IRC17:01
*** Jack_Iv has joined #openstack-ansible17:02
*** tonytan4ever has quit IRC17:04
*** tonytan4ever has joined #openstack-ansible17:04
*** marst_ has joined #openstack-ansible17:07
*** LinStatSDR has quit IRC17:07
*** muxdaemon has joined #openstack-ansible17:07
*** marst has quit IRC17:09
*** muxdaemon has quit IRC17:13
*** hw_wutianwei has quit IRC17:14
*** jmckind has quit IRC17:32
*** sanfern has quit IRC17:45
*** vnogin has joined #openstack-ansible17:45
*** shardy has quit IRC17:46
*** vnogin has quit IRC17:49
strattaoI’ve got a working OSA deployment, but I’m not entirely clear on the development workflow. Let’s say I need to make a change to the glance code. Is there a document that outlines the developer workflow that i’m not finding in the documentation?17:53
*** jwitko_ has joined #openstack-ansible17:56
logan-strattao: assuming you have published your glance fork to a git repo somewhere, you can override the glance git source in user_variables like  this http://cdn.pasteraw.com/a8wtkaoczd75k468pj3aqkgghm3e8lv and then run repo-build and the os-glance-install playbooks again, and it should build and pull the updated venv with your forked code17:58
*** sc68cal has quit IRC17:59
logan-the defaults for all of the git sources are here if you want to use them as examples: https://github.com/openstack/openstack-ansible/blob/master/playbooks/defaults/repo_packages/openstack_services.yml#L79-L8217:59
logan-you can confirm it is using     the correct  git source for glance by checking /var/www/repo/os-releases/<osa tag>/requirements* files  in  your repo container18:00
strattaothanks logan- how about debugging? what is your process for that? can you debug right on the box?18:02
strattao(I’m just poking around in the containers and don’t see any python code)18:02
logan-the python is in /openstack/venvs/venv_name/lib/python2.7/site-packages/18:04
*** david-lyle has joined #openstack-ansible18:05
logan-you can enable debugging for services in the user_variables too.. just set debug: True and run the service playbook if you want debug output to your logs18:05
*** sc68cal has joined #openstack-ansible18:06
*** chris_hultin|AWA is now known as chris_hultin18:06
strattaologan- thank you18:08
logan-np18:10
*** sc68cal_ has joined #openstack-ansible18:17
*** sc68cal has quit IRC18:17
*** sc68cal_ has quit IRC18:22
mgariepyany core available: https://review.openstack.org/#/c/445493/ https://review.openstack.org/#/c/445502/ https://review.openstack.org/#/c/445535/18:23
*** sc68cal has joined #openstack-ansible18:24
*** esberglu_ has joined #openstack-ansible18:27
*** esberglu_ has quit IRC18:27
*** tonytan4ever has quit IRC18:28
*** esberglu_ has joined #openstack-ansible18:28
*** MasterOfBugs has joined #openstack-ansible18:28
*** tonytan4ever has joined #openstack-ansible18:28
*** esberglu has quit IRC18:30
openstackgerritMerged openstack/openstack-ansible stable/newton: Update all SHAs for 14.1.2  https://review.openstack.org/44645418:33
evrardjpk18:36
evrardjpmgariepy: topic name: I see what you did there ;)18:38
logan-haha18:38
*** vnogin has joined #openstack-ansible18:39
openstackgerritMerged openstack/openstack-ansible-pip_install master: set yum priority  https://review.openstack.org/44550218:41
openstackgerritMerged openstack/openstack-ansible-os_nova master: Move away from "tenant_id" endpoints  https://review.openstack.org/44372018:42
openstackgerritMerged openstack/openstack-ansible-lxc_hosts master: set yum priority  https://review.openstack.org/44553518:43
openstackgerritMerged openstack/openstack-ansible-openstack_hosts master: set yum priority  https://review.openstack.org/44549318:43
agrebennikovhey folks, may you please take a look at this bug? https://bugs.launchpad.net/openstack-ansible/+bug/167357018:44
openstackLaunchpad bug 1673570 in openstack-ansible "With more than one ceph cluster as the backend nova has secret_uuid hardcoded" [Undecided,New]18:44
agrebennikovI need an opinion18:44
agrebennikovbefore I submit the fix please18:44
openstackgerritMarc Gariépy proposed openstack/openstack-ansible-openstack_hosts stable/ocata: set yum priority  https://review.openstack.org/44669018:45
agrebennikov(and Michael Gugino is not here)18:45
openstackgerritMarc Gariépy proposed openstack/openstack-ansible-lxc_hosts stable/ocata: set yum priority  https://review.openstack.org/44669318:45
openstackgerritMarc Gariépy proposed openstack/openstack-ansible-pip_install stable/ocata: set yum priority  https://review.openstack.org/44669418:45
mgariepyevrardjp, hehe18:59
evrardjpmgariepy: I don't like the alignment of the newlines there https://review.openstack.org/#/c/446693/1 could you start from scratch?19:00
evrardjp:p19:00
openstackgerritMerged openstack/openstack-ansible-pip_install stable/ocata: set yum priority  https://review.openstack.org/44669419:02
mgariepyevrardjp, lol yeah right aligned giant line of code.19:03
openstackgerritMerged openstack/openstack-ansible-openstack_hosts stable/ocata: set yum priority  https://review.openstack.org/44669019:03
evrardjpmgariepy: haha perfect then! :p19:04
evrardjpI'll stop joking for the day and go back home. Thanks for the patches mgariepy :)19:04
*** sc68cal has quit IRC19:05
mgariepyevrardjp, i want the damn thing to work !19:07
mgariepyhaha19:07
openstackgerritMerged openstack/openstack-ansible-lxc_hosts stable/ocata: set yum priority  https://review.openstack.org/44669319:07
mgariepy90% done, only 90% left to be done!19:08
*** electrofelix has quit IRC19:08
*** sc68cal has joined #openstack-ansible19:11
*** smatzek has joined #openstack-ansible19:16
*** Jack_Iv has quit IRC19:18
openstackgerritJimmy McCrory proposed openstack/openstack-ansible stable/newton: [WIP] Test Ansible 2.1.5.0-0.2.rc2  https://review.openstack.org/43700319:21
*** tonytan4ever has quit IRC19:22
*** zz_pwnall138 is now known as pwnall133719:23
*** vnogin has quit IRC19:23
*** tonytan4ever has joined #openstack-ansible19:23
palendaeThat's quite a version number19:27
mgariepycloudnull, andymccr http://paste.openstack.org/show/603018/ ?19:39
mgariepyany idea ?19:40
cloudnullmgariepy: looing19:44
cloudnull**looking19:44
* cloudnull can't type19:44
cloudnullmgariepy: is that from master?19:45
mgariepyyep.19:45
mgariepymaster from this morning-ish.19:45
cloudnulldo you have this commit https://github.com/openstack/openstack-ansible/commit/cacf41028b1fe398175d34cc298ce39042edf09e ?19:45
mgariepyyep.19:47
cloudnullmgariepy: on your repo-server do you have this path venvs/master/centos-7.3-x86_64 ?19:49
mgariepynova containers don't have the cinder* roles.19:49
cloudnullunder /var/www/repo/venvs/master/19:49
mgariepyyes19:49
mgariepythe cinder download for nova works on aio.19:50
cloudnullodd that the checksum url is "http://127.0.0.1/venvs/untagged/ubuntu/cinder.checksum"19:51
cloudnullon a cent host19:51
mgariepyyeah but the url is defined in inventory/group_vars/cinder_all.yml19:51
mgariepythe nova containers are not part of this group19:52
cloudnullwhen the actual download is showing "http://172.29.236.100:8181/venvs/master/centos-7.3-x86_64/cinder-master-x86_64.tgz"19:52
cloudnulloh...19:52
cloudnulli get it ...19:52
cloudnullthis is due to group_var nonsense.19:52
* cloudnull hates group_vars19:52
*** stuartgr has left #openstack-ansible19:53
*** smatzek has quit IRC19:53
cloudnullseems like https://github.com/openstack/openstack-ansible/blob/master/playbooks/inventory/group_vars/cinder_all.yml#L38 and all vars like it should be moved back to all19:53
cloudnull**all.yml19:54
cloudnullbeing that the var could be shared across deployments.19:54
cloudnullmgariepy: is that something specific to your deployment?19:55
cloudnullor are we running cinder venv things on nova hosts by default?19:55
mgariepyaio deploys cinder and nova on the host.19:55
cloudnullso we'd not suspect to see this on a multi-node deployment?19:56
*** pwnall1337 is now known as zz_pwnall133719:56
mgariepyi guess we should see it unless i don't understand the issue correctly.19:57
cloudnullim confused then. sorry. its been one of those days :)19:57
cloudnullso this was an issue you're seeing on an AIO ?19:58
cloudnullwas this in the gate?19:58
cloudnullor local test.19:58
*** retreved has quit IRC19:59
cloudnulland was it changing anything on the host?19:59
cloudnullI see from the output19:59
cloudnulltask path: /etc/ansible/roles/os_nova/tasks/cinder_install.yml:7919:59
mgariepythe issue as I understand it is : cinder_venv_donwload_url is  set in group_vars/cinder_all.yml which is not a group of nova-* container but nova install cinder19:59
cloudnullit looks like it's trying to execute the nova role but running the cinder_install.yml file20:00
mgariepyin  /etc/ansible/roles/os_nova/tasks/cinder_install.yml20:00
cloudnullhttps://github.com/openstack/openstack-ansible-os_nova/tree/master/tasks20:00
cloudnullbut https://github.com/openstack/openstack-ansible-os_cinder/blob/master/tasks/cinder_install.yml#L7920:00
mgariepyho.20:00
cloudnullwhich from the output looks like the nova role file path is running the cinder role20:01
cloudnullfor some reason20:01
cloudnullmaybe we've broken the role downloader ?20:01
xdfilhmm there seems to be no clean way to do an OVS install20:01
mgariepyha20:01
mgariepylol20:01
mgariepyit's probably me..20:02
mgariepywell. i cannot type as well..20:02
xdfilI think you have to do a first-run with br-vxlan settings then do a second run after manually setting up OVS and changing the IP settings20:02
mgariepycloudnull, i guess i did a checkout of cinder patch in the wrong folder20:03
mgariepyLOL20:03
cloudnullI've done that :)20:03
xdfilgit is hard20:03
cloudnullit happens20:03
xdfilI dont know how to use it either20:03
mgariepythe issue is not git, it was me.20:04
xdfilI suppose id have to say the same20:04
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible stable/newton: Updates to support ignored packages and external indexes  https://review.openstack.org/44671820:05
xdfilHeres and idea:20:05
xdfilI tried setting the repo container as bare metal last night20:05
xdfilso that i could destroy/create containers without cleaning up the apt stuff20:05
xdfilbut the containers on the local metal repo cant talk to the repo over the LB vip20:06
xdfiler20:06
xdfilno it was the host itseld that couldnt access the repo i think20:06
*** Jeffrey4l has quit IRC20:06
xdfileither way i had to ditch that idea20:06
*** Jeffrey4l has joined #openstack-ansible20:07
openstackgerritKevin Carter (cloudnull) proposed openstack/openstack-ansible-repo_build stable/newton: Updates to support ignored packages and external indexes  https://review.openstack.org/44671920:08
xdfilcloudnull are there changes in the roadmap for inventory?20:08
mrdaMorning OSA20:11
xdfilhi, but its afternoon20:11
xdfilunless you are in hawaii or something20:12
cloudnullxdfil: mrda is even farther away :)20:13
xdfiland if your in asia, then its tomorrow20:13
*** muxdaemon has joined #openstack-ansible20:13
mrdacloudnull: or more distant20:13
cloudnullxdfil: palendae: has some changes in mind20:13
cloudnullmrda: is from the future20:13
xdfiltomorrow hasnt happend yet so your just in my imagination?20:13
mrdaJust please don't ask for sports results you can bet on, I'm tired of providing those20:13
xdfilcloudnull is there an etherpad or something?20:14
cloudnullwhat if i told you there is no tomorrow ?20:14
mgariepyarf:: /usr/local/bin/openstack-ansible: line 64: 32706 Segmentation fault      /opt/ansible-runtime/bin/ansible-playbook "${@}" ${VAR1}20:14
mgariepytime to go home. cya tomorrow all20:14
palendaexdfil: What kind of inventory changes? I do have reorg of the code in mind, I'm not sure if that'd help you20:14
mrdaI can tell you will certainty that there is no spoon20:14
xdfilo/20:14
xdfilI'd just like to be aware20:14
xdfilI think it would help me learn20:15
cloudnullmgariepy: seg fault ? !20:15
xdfilbecause often I have ideas that are usually things you guys have already tried or whatnot20:15
palendaexdfil: Ah, ok20:15
mgariepyyep, it does happen quite often on my centos install.20:15
palendaeYeah, I need to write more of the ideas down...20:16
palendaeThe pike work will mostly be about implementing a spec20:16
mgariepyhappened here: task path: /etc/ansible/roles/os_nova/tasks/nova_install.yml:5920:17
mgariepybut might not be related.20:17
cloudnullwow. i wonder what is causing that ...20:18
palendaexdfil: https://review.openstack.org/#/c/410342/ is the spec I hope to get done in Pike. Might just be making the plugin system and moving our current file IO to that20:18
cloudnullansible-playbook to segfault...20:19
cloudnulli'll dig more into that20:19
cloudnullmgariepy: which kernel if you dont mind me asking?20:19
xdfilpalendae I'll check it out20:20
xdfillooks promising20:20
xdfilI've been meaning to ask yall: I cant seem to figure out how to set enable_distributed_router: True (horizon local_settings.py)20:21
xdfilits the only option that doesn't seem to have a variable lol20:21
xdfilI tried with config override but I think it works differently in non-ini file20:22
xdfilis it just me or is horizon kind of junky?20:24
xdfilits using angular and such and you'd expect it to be more modern interface. But it has those dumb modal windows20:25
dmsimardcloudnull, odyssey4me, mgariepy: o/ I'm in the process of drafting a project manifesto for ARA, would appreciate your input if you have any :) https://review.openstack.org/44670920:25
*** crushil has quit IRC20:35
DimGRwhats the difference if you ssh in a container or if you lxc-attach ?20:37
*** FrankZhang has quit IRC20:38
*** crushil has joined #openstack-ansible20:40
*** FrankZhang has joined #openstack-ansible20:41
openstackgerritNolan Brubaker proposed openstack/openstack-ansible-ops master: Add a basic glance API tests  https://review.openstack.org/44613120:41
*** cjloader_ has joined #openstack-ansible20:41
*** cjloader_ has quit IRC20:41
palendaeAnyone seen an error like this? I'm using 14.0.0 https://gist.github.com/nrb/c2e3ac5ee2080f4b34787df0fc98147e20:43
*** cjloader has quit IRC20:45
*** jwitko__ has joined #openstack-ansible20:46
cloudnullDimGR: ssh is opening a ssh connection to the instance, lxc-attach is connecting to the containers namespace.20:48
xdfilpalendae seems like the plays really want config overides but there arent any20:48
cloudnullthey both achive roughly the same goal.20:48
DimGRok thanks20:48
cloudnulldmsimard: running around  for the next little while.20:48
cloudnullbut i will look at that soon-ish20:48
xdfilwhy you using 14.0.0 ? why not 14.0.8 or whatever is latest?20:48
dmsimardcloudnull: not an emergency at all, thanks for your time20:49
cloudnullthanks for making ara20:49
*** acormier has quit IRC20:49
cloudnull:)20:49
dmsimard¯\_(ツ)_/¯20:49
*** crushil has quit IRC20:50
palendaexdfil: Yeah, though they should be in swift.yml, and someone else has this same config but working :(20:53
*** acormier has joined #openstack-ansible20:54
odyssey4mexdfil unfortunately due to horizon's config format, we have to  implement a var for each setting, so if it doesn't have one then please push a patch :)20:58
odyssey4meif you haven't figured out how to do that yet, then feel free to raise a bug to ask for the widget to turn that setting on20:58
xdfilpalendae if the config is the same and OSA version is the same, then maybe clear the facts?20:58
odyssey4mepalendae FYI 14.0.0 was using OpenStack RC code so there be dragons - I'd advise using a later tag if possible20:59
xdfilodyssey4me I'll get into that in a week or 2 hopefully20:59
palendaeodyssey4me: Roger, thanks20:59
odyssey4mewe also pushed a ton of fixes between 14.0.0 and 14.0.120:59
odyssey4meIIRC20:59
palendaeodyssey4me: I was using 14.0.0 because then I'd have space to minor upgrade, but yeah21:00
palendaeMight just bump it21:00
*** phalmos has quit IRC21:00
odyssey4mepalendae 10 tags released since then ;)21:00
palendaeHaha, yeah, fair21:01
xdfilodyssey4me I think i remember hearing you say OSA was trying not to have a variable for /everything/21:01
xdfilbut having the variables sure makes it easier...21:02
odyssey4mexdfil we really do, but horizon is just one of those nonsense config implementations21:02
palendaeRE: Horizon nonsense - https://review.openstack.org/#/c/404735/21:02
odyssey4mefor anything json, ini or yaml we have config overrides21:02
palendaeThey should have used the local_settings.py convention to read in an ini file from the start21:02
palendaeBut they're working on it now21:02
*** LinStatSDR has joined #openstack-ansible21:02
palendaeodyssey4me, xdfil ^^^21:02
*** LinStatSDR has left #openstack-ansible21:02
odyssey4methanks palendae - starred21:02
xdfil^^21:03
odyssey4metime for me to go back to my hidey hole21:03
palendaeI threw my thoughts in there a few montsh ago21:03
odyssey4metime to !computer21:03
palendaeodyssey4me: Yeah, it's late for you21:03
palendaeGet gone21:03
cloudnullif any cores are around to give this a push through it21:08
cloudnullit'd be appreciated https://review.openstack.org/#/c/446719/21:08
*** acormier has quit IRC21:11
*** smatzek has joined #openstack-ansible21:15
*** agrebennikov has quit IRC21:20
*** muxdaemon has quit IRC21:24
*** muxdaemon has joined #openstack-ansible21:25
*** sukesh has joined #openstack-ansible21:33
*** Jeffrey4l has quit IRC21:34
*** sukesh has left #openstack-ansible21:35
*** sukesh has joined #openstack-ansible21:35
*** retreved has joined #openstack-ansible21:37
*** schwicht has quit IRC21:39
*** jwitko__ has quit IRC21:39
*** gouthamr has quit IRC21:40
*** smatzek has quit IRC21:40
*** jwitko_ has quit IRC21:40
*** karimb has joined #openstack-ansible21:41
*** jrobinson has joined #openstack-ansible21:42
openstackgerritAndrey Grebennikov proposed openstack/openstack-ansible-galera_server master: Added Galera SSL support  https://review.openstack.org/42599721:42
*** Jeffrey4l has joined #openstack-ansible21:48
openstackgerritAndrey Grebennikov proposed openstack/openstack-ansible-haproxy_server stable/newton: Added Haproxy global tunables  https://review.openstack.org/44675421:50
*** Jeffrey4l has quit IRC21:50
*** Jeffrey4l has joined #openstack-ansible21:50
openstackgerritAndrey Grebennikov proposed openstack/openstack-ansible-haproxy_server stable/ocata: Added Haproxy global tunables  https://review.openstack.org/44675521:53
cloudnullalright all, im off. take care.22:11
DimGRbye master22:11
cloudnulllol22:11
DimGR:)22:11
*** agrebennikov has joined #openstack-ansible22:13
agrebennikovtnx cloudnull22:14
*** yolanda has quit IRC22:16
*** zz_pwnall1337 is now known as pwnall133722:18
*** kstev has quit IRC22:25
*** cjloader has joined #openstack-ansible22:29
*** tonytan4ever has quit IRC22:30
*** furlongm has joined #openstack-ansible22:36
*** esberglu_ has quit IRC22:43
xdfildang22:45
xdfilubuntu doesn't randomize your IQN22:45
xdfilautmatically on instlal22:46
*** karimb has quit IRC22:46
xdfilanyone know if I need to clear out any values after changing the IQN? Like cinder/nova DB. ansible facts, anything?22:47
*** manheim has joined #openstack-ansible22:51
*** cjloader has quit IRC22:53
*** galstrom is now known as galstrom_zzz22:53
openstackgerritOpenStack Proposal Bot proposed openstack/openstack-ansible master: Updated from global requirements  https://review.openstack.org/44677822:57
*** askb has joined #openstack-ansible22:59
*** chris_hultin is now known as chris_hultin|AWA23:02
*** jamielennox is now known as jamielennox|away23:03
*** jamielennox|away is now known as jamielennox23:05
*** pmannidi has joined #openstack-ansible23:14
*** Jeffrey4l has quit IRC23:14
*** jrobinson is now known as jrobinson-afk23:19
*** agrebennikov has quit IRC23:27
*** Jeffrey4l has joined #openstack-ansible23:29
*** FrankZhang has quit IRC23:29
*** pwnall1337 is now known as zz_pwnall133723:31
*** klamath has quit IRC23:33
*** jrobinson-afk is now known as jrobinson23:42
*** manheim has quit IRC23:45
*** marst_ has quit IRC23:45
*** DimGR2 has joined #openstack-ansible23:49
openstackgerritMerged openstack/openstack-ansible stable/ocata: Update all SHAs for 15.0.1  https://review.openstack.org/44644523:49
*** DimGR has quit IRC23:51
*** marst has joined #openstack-ansible23:55
*** DimGR2 has quit IRC23:58

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!