| *** spatel has joined #openstack-ansible | 00:08 | |
| *** ianychoi has quit IRC | 00:12 | |
| *** ianychoi has joined #openstack-ansible | 00:14 | |
| *** ansmith has quit IRC | 00:19 | |
| *** ianychoi has quit IRC | 00:20 | |
| *** ianychoi has joined #openstack-ansible | 00:21 | |
| *** ansmith has joined #openstack-ansible | 00:23 | |
| *** hamzy has joined #openstack-ansible | 00:27 | |
| *** ianychoi has quit IRC | 00:30 | |
| *** ianychoi has joined #openstack-ansible | 00:31 | |
| *** spatel has quit IRC | 00:34 | |
| *** ianychoi has quit IRC | 01:03 | |
| *** tureba has quit IRC | 01:04 | |
| *** ianychoi has joined #openstack-ansible | 01:04 | |
| *** ansmith has quit IRC | 01:07 | |
| logan- | cyberpear: i don't think anyone implemented the ubuntu stig for ansible-hardening yet. | 01:10 |
|---|---|---|
| *** spatel has joined #openstack-ansible | 01:30 | |
| *** spatel has quit IRC | 01:30 | |
| *** kplant has quit IRC | 01:52 | |
| *** ianychoi has quit IRC | 02:14 | |
| *** ianychoi has joined #openstack-ansible | 02:16 | |
| *** ianychoi has quit IRC | 02:38 | |
| *** ianychoi has joined #openstack-ansible | 02:40 | |
| *** udesale has joined #openstack-ansible | 03:51 | |
| *** rgogunskiy has joined #openstack-ansible | 03:52 | |
| *** raukadah is now known as chandankumar | 03:52 | |
| *** rgogunskiy has quit IRC | 03:56 | |
| *** joshualyle has joined #openstack-ansible | 04:09 | |
| *** viks___ has joined #openstack-ansible | 04:17 | |
| *** jhesketh has quit IRC | 04:19 | |
| *** jhesketh has joined #openstack-ansible | 04:19 | |
| *** dave-mccowan has quit IRC | 04:23 | |
| *** gshippey has quit IRC | 04:26 | |
| *** ivve has joined #openstack-ansible | 05:36 | |
| *** Adri2000 has quit IRC | 06:17 | |
| *** pcaruana has joined #openstack-ansible | 06:27 | |
| jrosser | Morning | 06:29 |
| *** ianychoi has quit IRC | 06:34 | |
| *** cshen has joined #openstack-ansible | 06:35 | |
| *** ianychoi has joined #openstack-ansible | 06:41 | |
| *** kopecmartin has joined #openstack-ansible | 06:43 | |
| *** luksky has joined #openstack-ansible | 06:45 | |
| *** rgogunskiy has joined #openstack-ansible | 07:12 | |
| *** rgogunskiy has quit IRC | 07:12 | |
| *** rgogunskiy has joined #openstack-ansible | 07:12 | |
| *** rgogunskiy has joined #openstack-ansible | 07:13 | |
| *** tosky has joined #openstack-ansible | 07:16 | |
| evrardjp | morning | 07:23 |
| *** jsquare has quit IRC | 07:26 | |
| *** jsquare has joined #openstack-ansible | 07:28 | |
| *** ianychoi has quit IRC | 07:59 | |
| *** ianychoi has joined #openstack-ansible | 08:01 | |
| noonedeadpunk | hi folks:) | 08:02 |
| *** key-networks has joined #openstack-ansible | 08:09 | |
| *** rpittau|afk is now known as rpittau|mtg | 08:41 | |
| gillesMo | Hello ! Is there any plan to correct this bug https://bugs.launchpad.net/openstack-ansible/+bug/1807346 ? Ansible upgrade or not using os_* modules ? | 08:42 |
| openstack | Launchpad bug 1807346 in openstack-ansible "[heat] Installations fails during Update Q->R" [Undecided,New] - Assigned to Guilherme Steinmuller Pimentel (guilhermesp) | 08:42 |
| *** fnpanic has joined #openstack-ansible | 08:59 | |
| *** j^2 has joined #openstack-ansible | 09:04 | |
| *** ianychoi has quit IRC | 09:07 | |
| *** ianychoi has joined #openstack-ansible | 09:10 | |
| *** openstackgerrit has joined #openstack-ansible | 09:28 | |
| openstackgerrit | JJ Asghar proposed openstack/ansible-hardening master: Updated the Vagrantfile with more support https://review.opendev.org/667563 | 09:28 |
| *** hamzaachi has joined #openstack-ansible | 09:35 | |
| *** fnpanic has quit IRC | 09:43 | |
| chandankumar | logan-: noonedeadpunk jrosser https://review.opendev.org/#/c/661994/11 it works fine on rhel-8 | 09:48 |
| chandankumar | it is good to go | 09:48 |
| jrosser | gillesMo: i'm not sure what you mean by "correct this bug", we fix the ansible version to whatever is current during the development cycle of a particular release of OSA | 09:54 |
| jrosser | it is very difficult to retrospectively upgrade the version as it has large implications for deployers who integrate with other systems | 09:54 |
| jrosser | it is really an ansible bug and the relevant fixes need to be backported to the required version of ansible | 09:55 |
| *** vnogin has joined #openstack-ansible | 09:55 | |
| jrosser | gillesMo: are you able to propose a backport to ansible 2.6 as described here? https://github.com/ansible/ansible/pull/42899#issuecomment-445787226 | 10:00 |
| gillesMo | jrosser: I understand, so it'sd perhaps easier to revert the changes in os_heat to not use ansible modules until openstack-ansible use a corrected version of Ansible | 10:06 |
| jrosser | gillesMo: that would also be possible, yes | 10:07 |
| jrosser | if you are able to test such a change that would be very helpful | 10:07 |
| gillesMo | backport to Ansible 2.6 ? My current Rocky deployment use Ansible 2.5.14 | 10:08 |
| jrosser | oh right of course :/ | 10:08 |
| jrosser | hmm | 10:08 |
| *** ansmith has joined #openstack-ansible | 10:13 | |
| *** ansmith has quit IRC | 10:27 | |
| *** luksky has quit IRC | 10:28 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-openstack_hosts master: Install user supplied CA certificates into system trust store https://review.opendev.org/641445 | 10:35 |
| *** vnogin has quit IRC | 10:42 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-openstack_hosts master: Install user supplied CA certificates into system trust store https://review.opendev.org/641445 | 10:42 |
| CeeMac | good day | 10:55 |
| *** luksky has joined #openstack-ansible | 10:58 | |
| *** cshen has quit IRC | 11:05 | |
| *** tosky has quit IRC | 11:15 | |
| *** tosky has joined #openstack-ansible | 11:16 | |
| *** kplant has joined #openstack-ansible | 11:16 | |
| *** cshen has joined #openstack-ansible | 11:21 | |
| *** fnpanic has joined #openstack-ansible | 11:22 | |
| *** ianychoi has quit IRC | 11:40 | |
| *** ianychoi has joined #openstack-ansible | 11:45 | |
| *** ivve has quit IRC | 11:51 | |
| *** udesale has quit IRC | 11:51 | |
| *** udesale has joined #openstack-ansible | 11:52 | |
| guilhermesp | mornings! | 11:53 |
| *** tuxjohnson has joined #openstack-ansible | 11:58 | |
| *** vnogin has joined #openstack-ansible | 11:58 | |
| *** luksky has quit IRC | 12:02 | |
| *** luksky has joined #openstack-ansible | 12:03 | |
| *** vnogin has quit IRC | 12:03 | |
| *** tuxjohnson has quit IRC | 12:10 | |
| *** ianychoi has quit IRC | 12:14 | |
| *** ianychoi has joined #openstack-ansible | 12:22 | |
| *** martinkennelly has joined #openstack-ansible | 12:23 | |
| kaiokmo | morning guilhermesp | 12:31 |
| jrosser | anyone around who understands nested virt / kvm vs. qemu better than i do? mnaser? there is something wierd like that breaking our octavia tests | 12:35 |
| mnaser | jrosser: linksies? | 12:35 |
| jrosser | sure, one moment | 12:35 |
| *** tuxjohnson has joined #openstack-ansible | 12:36 | |
| *** tuxjohnson has left #openstack-ansible | 12:36 | |
| jrosser | mnaser: http://logs.openstack.org/51/664551/1/check/openstack-ansible-deploy-aio_metal_octavia-ubuntu-bionic/e374421/logs/openstack/aio1-nova/nova-compute.log.txt.gz#_2019-06-11_22_28_21_361 | 12:36 |
| *** vnogin has joined #openstack-ansible | 12:37 | |
| *** ansmith has joined #openstack-ansible | 12:38 | |
| *** fnpanic has quit IRC | 12:39 | |
| jrosser | we've been running AIO+octavia here on bionic bare metal this week, thats all working fine | 12:40 |
| jrosser | also any reviews on this - https://review.opendev.org/#/c/667150/ horizon is br0k on master right now, but passing tests :/ | 12:42 |
| *** vnogin has quit IRC | 12:42 | |
| *** gkadam has joined #openstack-ansible | 12:43 | |
| kplant | jrosser: not a resources issue? 2019-06-11T22:28:20.871535Z qemu-system-x86_64: cannot set up guest memory 'pc.ram': Cannot allocate memory | 12:48 |
| jrosser | kplant: ooooh! | 12:48 |
| jrosser | kplant: i may have paid too much attention to qemu-system-x86_64: warning: TCG doesn't support requested feature: CPUID.01H:ECX.vmx [bit 5] | 12:49 |
| kplant | yeah that is a lot scarier looking for sure | 12:50 |
| kplant | but usually that can be safely ignored, especially on nested virt | 12:50 |
| *** dave-mccowan has joined #openstack-ansible | 12:51 | |
| kplant | had to look it up but it means you don't support LZCNT opcode | 12:52 |
| johnsom | jrosser: that vmx warning is normal, it is the “cannot allocate memory” error that is the fatal error. | 13:14 |
| jrosser | johnsom: i was just looking at reducing the amount of ram we use | 13:15 |
| jrosser | johnsom: is there a smaller value we can pick here? https://github.com/openstack/openstack-ansible-os_octavia/blob/master/defaults/main.yml#L264 | 13:15 |
| johnsom | qemu does some probes to see if certain cpu features are present. That is the vmx message. | 13:15 |
| johnsom | Technically yes, if you aren’t loading the lb or adding a lot of listeners, you can cut that in half. But for practical use, you really should have at least 1gb | 13:16 |
| johnsom | I.e. if you cut that, only do it for your gate job, not that default. | 13:17 |
| jrosser | johnsom: sure - i can override that just for the AIO / gate jobs | 13:19 |
| johnsom | Also, we are switching ubuntu kernels real soon now that should save some ram. It saves ~200MB in the qcow2 alone. | 13:19 |
| johnsom | It removes a ton of kernel modules we don’t need. | 13:20 |
| johnsom | FYI, https://review.opendev.org/#/c/665861 | 13:27 |
| *** key-networks has quit IRC | 13:27 | |
| johnsom | and https://review.opendev.org/666063 | 13:28 |
| *** spatel has joined #openstack-ansible | 13:31 | |
| *** spatel has quit IRC | 13:36 | |
| CeeMac | quick question, the os_<project> roles, they're kept in their own repos outside of openstack-ansible? So I needed to change the content of a file in one of the roles in line with a patch im writing for osa how does one acheive this? | 13:43 |
| CeeMac | tl;dr - halp! | 13:43 |
| guilhermesp | CeeMac: I think you can just go to /etc/ansible/roles/os_<project> and checkout your patch | 13:44 |
| CeeMac | guilhermesp: I'm not sure I follow. If you look at etherpad I shared with you, the first step requires modifying the os_glance default/main.yml | 13:47 |
| CeeMac | but atm i'm checked out branch in openstack-ansible, which is where the rest of the changes need to be made. Do I need to submit seperate patches? One for the role and one for osa? | 13:48 |
| CeeMac | I'm still trying to get to grips with the development side of things.... | 13:48 |
| *** af_tim has joined #openstack-ansible | 13:48 | |
| guilhermesp | well CeeMac if you need to make changes in roles, just submit patches for these roles. the main osa repo, which we call integrated repo, is only repsonsible to evoke the roles and do all the deployment. So if you made a patch for os_glance and want to test it in an aio locally, you can just checkout your patch in the role and you're ready to go | 13:51 |
| guilhermesp | not sure if is clear for you now, let me know | 13:51 |
| CeeMac | guilhermesp: i think that makes sense. So, If I checkout the os_glance role, make the necessary changes and submit that, linked to the bug. then I can checkout osa, make the required modifications to complete the fix and submit that, linked to the same bug? does that make sense? | 13:54 |
| guilhermesp | oh wait... are you willing to go in the glance role and develop things there, submit a patch and etc? If that's the case, you can do it anywhere, just clone the role anywhere, configure your git review and submit the patch so we can review it :) | 13:57 |
| *** vnogin has joined #openstack-ansible | 14:00 | |
| CeeMac | guilhermesp: tbh I'm not sure what i'm doing at all, haha. I'm trying to convert that etherpad into "make it work in master" | 14:04 |
| guilhermesp | so CeeMac you want to submit patches fixing the things for the involved roles right? like os_glance | 14:05 |
| CeeMac | btw this for the glance cinder store issue we were working on. olso.rootwrap needs deploying as part of the glance role deployment so I need to tweak a couple of files in there, then everything else is deployment changes | 14:06 |
| CeeMac | guilhermesp: correct | 14:06 |
| guilhermesp | so basically just clone the os_<project> anywhere. Configure your git to be able to submit the patch ( I think spotz has a pdf showing how to do it ) | 14:07 |
| * jrosser fixes his magnum deploy, *again* https://media.giphy.com/media/jqOiGF0SHLI9G/giphy.gif | 14:07 | |
| guilhermesp | hahahah jrosser what is the issue now? | 14:08 |
| jrosser | some new brokennes this time in magnum for S | 14:08 |
| jrosser | usual internal vs. external keystone endpoint confusion | 14:08 |
| guilhermesp | oh boy.... I think we have magnum deployed with S, do we mnaser ? | 14:09 |
| spotz | guilhermesp: I have a whole pdf slide deck:) | 14:10 |
| CeeMac | guilhermesp: presumably I only need to configure git once for gerrit? I did that from https://docs.openstack.org/infra/manual/developers.html | 14:10 |
| CeeMac | spotz: Hi! Would appreciate any help/info/brain uploads anyone can provide :D | 14:11 |
| guilhermesp | basically CeeMac but I think spotz 's pdfs are more comprehensive | 14:11 |
| CeeMac | i like comprehensive | 14:11 |
| CeeMac | comprehensive is good | 14:11 |
| guilhermesp | :P | 14:12 |
| CeeMac | dont want to break anything :S | 14:12 |
| guilhermesp | i think this is a nice opportunity for you to start contributing | 14:12 |
| jrosser | guilhermesp: all this stuff is very confusing https://review.opendev.org/#/c/583745/ | 14:12 |
| guilhermesp | it is nice, you'll see | 14:12 |
| jrosser | theres an internal<>external endpoint swap in there | 14:13 |
| spotz | CeeMac: this was the last version of it https://github.com/spotz/berlin18/blob/master/gitandgerrit.pdf. The session wasn't held in Denver | 14:13 |
| guilhermesp | huum jrosser I think mnaser can explain better then me the reason for that swap | 14:14 |
| CeeMac | thanks spotz :D | 14:14 |
| spotz | CeeMac: NP it was a fun session to present, I miss it | 14:15 |
| guilhermesp | is magnum having trouble to access the internal endpoint jrosser ? | 14:15 |
| jrosser | it appears to be using the external one, when it shouldnt | 14:15 |
| jrosser | *trying to use | 14:15 |
| jrosser | time to dig in the code and see whats going on | 14:16 |
| jrosser | this feels very similar to our troubles with heat | 14:16 |
| guilhermesp | yeah smells like | 14:16 |
| CeeMac | spotz: seems file is corrupt? | 14:18 |
| spotz | CeeMac: might be too big, the vancouver one works https://github.com/spotz/vancouver18/blob/master/gitandgerrit.pdf | 14:19 |
| CeeMac | spotz: thats better, thanks | 14:21 |
| spotz | CeeMac: no worries:) | 14:22 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Reduce the RAM allocation for the octavia amphora flavor in AIO/CI https://review.opendev.org/667633 | 14:26 |
| *** Original_ElGuapo has quit IRC | 14:33 | |
| guilhermesp | jrosser: is this related to the latest failures with ocatavia role? ^ | 14:33 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_magnum master: Remove deprecated and incorrect config from magnum.conf https://review.opendev.org/667636 | 14:36 |
| jrosser | guilhermesp: yes I think that is why octavia fails on bionic CI | 14:37 |
| jrosser | and also there seems to be wierdness in our magnum.conf too https://review.opendev.org/667636 | 14:37 |
| *** ansmith_ has joined #openstack-ansible | 14:38 | |
| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_octavia master: db_setup: refactor database setup to a common file https://review.opendev.org/664888 | 14:40 |
| *** ansmith has quit IRC | 14:40 | |
| *** luksky has quit IRC | 14:43 | |
| guilhermesp | I think I might add as depends-on of one of my octavia patches | 14:44 |
| guilhermesp | yeah I was just looking at your magnum pr | 14:44 |
| jrosser | guilhermesp: bad news :( undoing this https://review.opendev.org/#/c/583745/ makes it all work again | 14:46 |
| guilhermesp | uoww... should it be bad or good news? :P | 14:47 |
| jrosser | well good news in that it works again! | 14:47 |
| jrosser | bad news in that it's down the wormhole again with how all this keystone stuff is handled | 14:48 |
| guilhermesp | heheheh I'm not sure if mnaser 's patch was something that worked for us then we submitted the patch... | 14:48 |
| jrosser | i am just totally not liking how www_authenticate_uri works across all of this stuff | 14:50 |
| jrosser | the documentation for that from keystone middleware is ambiguous, because it talks about "client" without specifying the context of the client, i.e internal vs. external | 14:52 |
| jrosser | then the same boilerplate text is in all of the service config files "Complete "public" Identity API endpoint. This endpoint should not be an "admin" endpoint, as it should be accessible by all end users." | 14:53 |
| jrosser | and yet it just doesnt seem to work that way | 14:53 |
| *** cshen has quit IRC | 14:54 | |
| johnsom | jrosser FYI, if you have a minute, I have this open: https://review.opendev.org/#/c/648744/ but I don't have the cycles to track down how it needs to be adjusted. If you are motivated, I would be happy to have your help in getting that cleaned up. | 15:00 |
| * jrosser looks | 15:00 | |
| *** asettle is now known as asettle-PTO | 15:03 | |
| johnsom | It's a straight forward fix, but the context switching to how OSA needs it is a bit rough for me sometimes. I'm sure Jesse is right in his comment, it's just the amount of time it would take me to figure it out and the bugs I have in flight in Octavia itself keep dropping it's priority on my list. | 15:03 |
| jrosser | johnsom: no worries i will try to look, we're in the middle of doing an octaiva deployment on the lab which is why i'm fiddling with all this now | 15:05 |
| johnsom | Ok, thanks for the help. | 15:05 |
| *** hamzaachi_ has joined #openstack-ansible | 15:14 | |
| *** hamzaachi has quit IRC | 15:16 | |
| *** ansmith has joined #openstack-ansible | 15:17 | |
| *** ansmith_ has quit IRC | 15:19 | |
| *** jamesdenton has joined #openstack-ansible | 15:29 | |
| *** rgogunskiy has quit IRC | 15:31 | |
| *** tosky has quit IRC | 15:37 | |
| *** hamzaachi_ has quit IRC | 15:37 | |
| *** vnogin has quit IRC | 15:52 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible-os_octavia master: Ensure that admin endpoint is used to configure octavia quota https://review.opendev.org/664551 | 16:42 |
| *** jsquare has quit IRC | 16:53 | |
| dirk | mnaser: is it okay to drop ubuntu-xenial in OSA? | 16:53 |
| *** jsquare has joined #openstack-ansible | 16:53 | |
| dirk | https://review.opendev.org/#/c/666272/2 is failing for many weeks already due to xenial-ssl, which doesn't seem that anyone is still fixing | 16:54 |
| openstackgerrit | Dirk Mueller proposed openstack/openstack-ansible-galera_server master: Fix galera_server installation for openSUSE Leap 15.1 https://review.opendev.org/666272 | 17:00 |
| openstackgerrit | Dirk Mueller proposed openstack/openstack-ansible-galera_server master: Remove xenial-ssl https://review.opendev.org/667682 | 17:00 |
| openstackgerrit | Dirk Mueller proposed openstack/openstack-ansible-galera_server master: Replace opensuse 42.3 functional testing with opensuse-15 https://review.opendev.org/667683 | 17:00 |
| *** martinkennelly has quit IRC | 17:07 | |
| jrosser | dirk: anything xenial on OSA master should be switched to bionic | 17:08 |
| dirk | there are already passing + voting bionic jobs | 17:09 |
| dirk | its just that the xenial job is broken and prevents anything from merging | 17:09 |
| dirk | I suggested https://review.opendev.org/#/c/667682 above to get rid of it | 17:09 |
| *** zbr|ruck is now known as zbr | 17:18 | |
| *** weifan has joined #openstack-ansible | 17:19 | |
| *** spatel has joined #openstack-ansible | 17:21 | |
| *** hamzaachi has joined #openstack-ansible | 17:22 | |
| *** kopecmartin is now known as kopecmartin|off | 17:24 | |
| *** luksky has joined #openstack-ansible | 17:29 | |
| *** udesale has quit IRC | 17:29 | |
| *** gkadam has quit IRC | 17:30 | |
| openstackgerrit | Merged openstack/openstack-ansible-os_horizon master: Only install py3 version of libapache2-mod-wsgi for ubuntu distro installs https://review.opendev.org/667150 | 17:33 |
| *** gkadam has joined #openstack-ansible | 17:35 | |
| spatel | I have Stupid question, I have two openstack cloud isolated, and both pointing to same ceph storage with same account and key password. in this case if i upload image to A cloud does that will be visible in cloud B? | 17:35 |
| mgariepy | sounds like a really bad plan IMO. | 17:41 |
| mgariepy | the answer would be no since glance won't scan the backend for images, the image needs to be pushed by glance. | 17:42 |
| mgariepy | spatel, why not creating pools for each deployment instead of sharing the sames ? | 17:43 |
| spatel | mgariepy: got you point.. i was just trying to understand how does that work, why do we need to create two pool, why not share? | 17:44 |
| mgariepy | because you are not sure the uuids won't be the same. | 17:45 |
| mgariepy | and just to save you some trouble donw the line. | 17:46 |
| spatel | hmm got it.. i will create second pool now | 17:46 |
| mgariepy | would you install 2 pc on the same rootfs ? | 17:46 |
| spatel | mgariepy: never.. just trying to understand what could go wrong | 17:46 |
| *** jamesdenton has quit IRC | 17:49 | |
| *** miouge36 has joined #openstack-ansible | 17:49 | |
| *** zerick_ has joined #openstack-ansible | 17:50 | |
| CeeMac | quick question. technically is 'master' still equating with stein from a release perspective? | 17:53 |
| *** miouge3 has quit IRC | 17:53 | |
| *** zerick has quit IRC | 17:53 | |
| jrosser | CeeMac: master will become Train at the point that branch is created | 17:53 |
| CeeMac | but from a 'current' release perspective, specifically in the os_* roles, which version is master 'now'? I'm trying to pull the correct oslo.rootwrap for glance and I'm not sure whether to use the stein or train one.... | 17:53 |
| *** altlogbot_0 has quit IRC | 17:55 | |
| jrosser | CeeMac: there is no train version, just master and the stable/<blah> branches | 17:56 |
| *** altlogbot_0 has joined #openstack-ansible | 17:56 | |
| *** jamesdenton has joined #openstack-ansible | 17:56 | |
| CeeMac | jrosser: thats what I'm struggling with. | 17:56 |
| jrosser | explain a bit more? | 17:57 |
| *** altlogbot_0 has quit IRC | 17:57 | |
| CeeMac | I'm trying to write a patch for glance. It was easy for rocky when I was doing in my env. But to submit a patch i need to do it against master. However, if master isn't Stein, and it isnt Train, which olso.rootwrap version do I code into the requirements.txt? | 17:57 |
| CeeMac | there is a specific version difference between the 2 | 17:58 |
| *** altlogbot_2 has joined #openstack-ansible | 17:58 | |
| logan- | CeeMac: the version of requirements we use in master will determine that | 17:58 |
| logan- | https://github.com/openstack/openstack-ansible/blob/82e853ce73973db6b7b1eb8df94107950105da8a/playbooks/defaults/repo_packages/openstack_services.yml#L34 | 17:58 |
| CeeMac | logan-: you mean based on the SHA of the role? | 17:59 |
| logan- | no, i mean if you install master right now, you will get the version of oslo.policy that is determined by the upper-constraints in the requirements repo | 18:00 |
| logan- | and requirements is pinned to that sha | 18:00 |
| logan- | so the version of oslo.policy you get is https://opendev.org/openstack/requirements/src/commit/47979b8d9080b529152a985820d3615e754e9c4a/upper-constraints.txt#L549 | 18:00 |
| CeeMac | its oslo.rootwrap i'm after though | 18:00 |
| logan- | ok, then https://opendev.org/openstack/requirements/src/commit/47979b8d9080b529152a985820d3615e754e9c4a/upper-constraints.txt#L494 | 18:01 |
| logan- | same thing | 18:01 |
| CeeMac | got it | 18:01 |
| logan- | you're still pulling your constraints from requirements | 18:01 |
| *** weifan has quit IRC | 18:01 | |
| CeeMac | great | 18:01 |
| CeeMac | i couldn't find it in there for rocky | 18:01 |
| *** weifan has joined #openstack-ansible | 18:01 | |
| *** altlogbot_2 has quit IRC | 18:01 | |
| CeeMac | or i was looking in the wrong place, so i wrote it into doc/requirements.tx of os_glance | 18:01 |
| *** weifan has quit IRC | 18:01 | |
| *** weifan has joined #openstack-ansible | 18:02 | |
| CeeMac | if its in the upper-constraints then I don't need to do that bit, just add it to the pip packages | 18:02 |
| CeeMac | thanks :) | 18:02 |
| *** altlogbot_3 has joined #openstack-ansible | 18:02 | |
| logan- | no prob | 18:02 |
| CeeMac | as a random aside, how do i stop opendev.org displaying in chinese by default? :S | 18:02 |
| CeeMac | point of clarity on convention, after I've modified some files should i manually 'git add' them, or just git commit -a at the end? the developer guide seems to indicate the latter, but just wanted to double check. logan- jrosser ? | 18:09 |
| jrosser | It’s a stylistic thing to a degree | 18:11 |
| jrosser | If you’re confident you are not adding a ton of other crap on top of your actual change then -a | 18:11 |
| CeeMac | well, confident seems like quite a strong word at this point | 18:11 |
| * CeeMac stoked beard thoughtfully | 18:12 | |
| CeeMac | i'll go with 'add' and see how i get on | 18:12 |
| jrosser | But I pretty much always git add <blah>, or even git add -p <blah> for partially adding changed things | 18:12 |
| jrosser | -p is very cool | 18:12 |
| CeeMac | tell me more about this -p | 18:12 |
| jrosser | You can add changes to the index chunk by chunk in an interactive way | 18:13 |
| jrosser | Per file | 18:13 |
| jrosser | So if you’ve done a big edit it’s easy to make a few commits out of it | 18:13 |
| CeeMac | ooohhhh kaaaaaay so maybe I'll ask you about that a bit more in the future | 18:14 |
| * CeeMac backs away slowly | 18:14 | |
| jrosser | git diff —cached is your friend too to look at the current state of the index | 18:14 |
| CeeMac | i don't think I'll be needing that for a while, lol | 18:14 |
| *** weifan has quit IRC | 18:15 | |
| *** weifan has joined #openstack-ansible | 18:16 | |
| *** weifan has quit IRC | 18:20 | |
| CeeMac | jrosser: running unit-testing is a pre-req to submitting a review? | 18:30 |
| jrosser | CeeMac: no | 18:31 |
| CeeMac | oh, good, because i don't have that setup on my laptop where I've writing the patch | 18:32 |
| openstackgerrit | Dirk Mueller proposed openstack/openstack-ansible master: Add aio_metal and aio_distro_metal openSUSE 15 jobs https://review.opendev.org/666058 | 18:36 |
| openstackgerrit | Craig McIntyre proposed openstack/openstack-ansible-os_glance master: Add oslo.rootwrap to glance_pip_packages https://review.opendev.org/667704 | 18:40 |
| CeeMac | is there a(n easy) way to make it so that you don't have to put your SSH key in multiple times after git review? | 18:40 |
| jrosser | ssh-agent | 18:42 |
| CeeMac | ack | 18:44 |
| *** ivve has joined #openstack-ansible | 18:45 | |
| CeeMac | i missed something, i can just add the other files and git -amend? then 'git review' again? | 18:47 |
| *** hwoarang has quit IRC | 19:03 | |
| jrosser | Sure, git add / git commit —amend / git review will do the trick | 19:03 |
| *** hwoarang has joined #openstack-ansible | 19:04 | |
| openstackgerrit | Craig McIntyre proposed openstack/openstack-ansible-os_glance master: Add missing pre-reqs for glance cinder store https://review.opendev.org/667704 | 19:12 |
| CeeMac | ok, think I'm done | 19:12 |
| CeeMac | ^that should match with my notes here https://etherpad.openstack.org/p/LEnPABy7Um | 19:13 |
| CeeMac | guilhermesp: fyi | 19:13 |
| *** hamzaachi has quit IRC | 19:20 | |
| mgariepy | anyone here is using keystone endpoint group filter ? | 19:21 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Reduce the RAM allocation for the octavia amphora flavor in AIO/CI https://review.opendev.org/667633 | 19:22 |
| jrosser | mgariepy: no - but i would like to know how to | 19:23 |
| spatel | How do i disable SSL option here? http://paste.openstack.org/show/753464/ | 19:24 |
| chandankumar | noonedeadpunk: jrosser here is the logs from rhel-8 run https://logs.rdoproject.org/17/171117/24/check/tripleo-ci-rhel-8-standalone-rhos-15/c8690fb/job-output.txt.gz | 19:24 |
| spatel | mgariepy: do you know how i can turn off SSL during image upload https://github.com/openstack/openstack-ansible-os_octavia/blame/dcdeb98468161402a3585990704749b40cac8ed2/tasks/octavia_amp_image.yml#L71 | 19:28 |
| mgariepy | spatel, https://github.com/openstack/openstack-ansible-openstack_openrc/blob/master/templates/clouds.yaml.j2#L19 set openrc_insecure: true | 19:31 |
| spatel | I do have that setting already.. | 19:32 |
| jrosser | chandankumar: that is excellent | 19:33 |
| mgariepy | spatel, is the correct setting in the cloud.yaml ? | 19:33 |
| spatel | mgariepy: this is what it looks like on my machine http://paste.openstack.org/show/753465/ | 19:34 |
| mgariepy | spatel, in /root/ what's the content of the actual file and not the ansible template | 19:34 |
| *** weifan has joined #openstack-ansible | 19:35 | |
| *** weifan has quit IRC | 19:35 | |
| *** weifan has joined #openstack-ansible | 19:35 | |
| spatel | mgariepy: this is what it looks in /root/openrc http://paste.openstack.org/show/753466/ | 19:36 |
| spatel | its pointing to http:// not SSL | 19:37 |
| *** kplant has quit IRC | 19:37 | |
| mgariepy | spatel, openstack endpoint list ? | 19:37 |
| mgariepy | if your catalog refer to ssl endpoint that's what it's going to contact | 19:37 |
| *** hamzaachi has joined #openstack-ansible | 19:38 | |
| spatel | mgariepy: http://paste.openstack.org/show/753467/ | 19:39 |
| *** martinkennelly has joined #openstack-ansible | 19:39 | |
| mgariepy | you have a bunch of https:10.30.xx in there. | 19:39 |
| martinkennelly | jrosser, you were right my bug was coming ansible_default_ipv4.interface. thank you. | 19:40 |
| spatel | octavia uploading image to glance which is https:// | 19:40 |
| jrosser | martinkennelly: oh great - i'm sure someone else had the same | 19:40 |
| spatel | 10.30.x.x is my public facing VIP | 19:40 |
| mgariepy | your public interface are all https. | 19:40 |
| jrosser | martinkennelly: do you know why it happened? | 19:40 |
| spatel | mgariepy: is there any temporary way i can say ignore SSL handshake validation or upload image? | 19:41 |
| martinkennelly | jrosser: Just going about findout out why. Will update bug report if I find something. | 19:41 |
| mgariepy | from the openrc you pasted the openrc_insecure: true is not set. | 19:41 |
| spatel | currently i am stuck at that point :( | 19:41 |
| mgariepy | or was set after the file has been generated | 19:41 |
| jrosser | martinkennelly: great thanks, we may be able to fix it, or insert an asset that fails early if something is wrong. i was wondering if there is a situation where the ansible facts have not been gathered | 19:42 |
| spatel | you are saying if i set openrc_insecure: true in /root/openrc that will fix this issue right? | 19:42 |
| spatel | mgariepy: ^ | 19:42 |
| martinkennelly | jrosser: It maybe a bug in how I setup my system, and maybe not something on OS ansible's end, but good to know anyway. | 19:43 |
| martinkennelly | jrosser: it was a clean ubuntu 18.04 both times fyi. | 19:43 |
| jrosser | martinkennelly: sure - i'm quite sure you are not the first with this issue though | 19:43 |
| jrosser | i'm always suspicious of netplan in these situations too :) | 19:44 |
| mgariepy | spatel, i think the client uses cloud.yaml instead. | 19:45 |
| *** altlogbot_3 has quit IRC | 19:45 | |
| mgariepy | but you will need to set the correct args in the file so it's not validating the cert. | 19:46 |
| *** altlogbot_3 has joined #openstack-ansible | 19:46 | |
| spatel | mgariepy: what option i have to hack it and move forward ? | 19:46 |
| martinkennelly | jrosser: I executed ansible all -i localhost, -m setup -c local, and I can see ansible_default_ipv4 is populated and has an interface | 19:46 |
| mgariepy | edit the file i guess | 19:46 |
| spatel | on deployment node right? | 19:47 |
| mgariepy | i think it's in this container: ostack-infra-2-3_octavia_server_container-e804ce0b | 19:47 |
| spatel | in that container there is no openrc file in /root director | 19:48 |
| mgariepy | is there a clouds.yaml file ? | 19:48 |
| spatel | in that container /root is empty | 19:49 |
| spatel | I believe that TASK is running on deployment node to load Octavia image.. | 19:49 |
| jrosser | martinkennelly: and before you did that is was not populated? | 19:50 |
| martinkennelly | jrosser: I did it after bootstrapping-aio.sh was run ONLY | 19:51 |
| jrosser | bootstrap-ansible.sh before that? | 19:51 |
| martinkennelly | jrosser: however if i rerun bootstrapping-aio.sh it fails with same error | 19:51 |
| martinkennelly | jrosser: yes | 19:51 |
| martinkennelly | jrosser: remember it works if i revert back a previous openstack version. | 19:52 |
| jrosser | martinkennelly: right - but that will change the major release of ansible which may be a factor | 19:53 |
| martinkennelly | jrosser: so maybe the difference how they get the interface between the two releases will help guide me. | 19:53 |
| spatel | mgariepy: hmm its good question where that playbook is running? | 19:55 |
| jrosser | martinkennelly: so here is where we gather facts after all the networking is set up on an AIO https://github.com/openstack/openstack-ansible/blob/master/tests/roles/bootstrap-host/tasks/prepare_networking.yml#L197 | 19:55 |
| jrosser | not sure i can see anywhere that is explicitly done up front though | 19:56 |
| chandankumar | cloudnull: https://review.opendev.org/#/c/661994/ please have a look when free logs attached in comments | 19:56 |
| martinkennelly | jrosser: I do not follow. How is this related? Thank you. | 19:57 |
| jrosser | well, the place that you get the error is here https://github.com/openstack/openstack-ansible/blob/master/tests/roles/bootstrap-host/tasks/prepare_networking.yml#L173 | 19:58 |
| jrosser | which is just before that fact gathering | 19:58 |
| chandankumar | cloudnull: thanks! | 19:59 |
| jrosser | so it is possible that in your configuration we should be also doing the same facts gathering before the place your error occurs, to ensure that ansible_default_ipv4 is always populated before it is used | 19:59 |
| martinkennelly | jrosser: Want me to try? | 20:00 |
| jrosser | sure, you could copy those lines earlier in that file, and give it a whirl | 20:00 |
| martinkennelly | jrosser: no luck | 20:04 |
| jrosser | martinkennelly: do you have a default route? | 20:05 |
| martinkennelly | jrosser: yes, ip r confirms this for me. | 20:06 |
| guilhermesp | nice patach CeeMac ! Can you please fix our comments? | 20:06 |
| martinkennelly | jrosser: I will ask the ansible people and see if they know anything. thank you again | 20:08 |
| jrosser | martinkennelly: i did find this https://github.com/ansible/ansible/issues/20051 | 20:08 |
| jrosser | but if you have a default route then it can't be that | 20:08 |
| jrosser | martinkennelly: omg https://github.com/ansible/ansible/blob/devel/lib/ansible/module_utils/facts/network/linux.py#L71-L72 | 20:10 |
| martinkennelly | jrosser: :) | 20:11 |
| *** af_tim has quit IRC | 20:13 | |
| martinkennelly | jrosser: can you run that comment ip -4 route get 8.8.8.8 on your machine and show me your output | 20:13 |
| *** altlogbot_3 has quit IRC | 20:15 | |
| jrosser | martinkennelly: http://paste.openstack.org/show/753468/ | 20:17 |
| *** altlogbot_0 has joined #openstack-ansible | 20:18 | |
| spatel | export LC_ALL=C.UTF-8 its throwing error like "sh: warning: setlocale: LC_ALL: cannot change locale (C.UTF-8)" | 20:21 |
| spatel | Do i need that option in openrc file? | 20:22 |
| martinkennelly | jrosser: Nothing to take from this. All normal. | 20:22 |
| spatel | any idea what is going on here http://paste.openstack.org/show/753469/ | 20:28 |
| jrosser | martinkennelly: how very odd indeed | 20:29 |
| martinkennelly | jrosser: As I said earlier when I run "ansible all -i localhost, -m setup -c local" I can see the correct default ipv4 interface. How can I run this command with openstack-ansible | 20:32 |
| *** hamzaachi has quit IRC | 20:36 | |
| jrosser | martinkennelly: that is exactly what is happening here https://github.com/openstack/openstack-ansible/blob/82e853ce73973db6b7b1eb8df94107950105da8a/tests/roles/bootstrap-host/tasks/prepare_networking.yml#L197 | 20:36 |
| CeeMac | guilhermesp: thanks, sure I'll take a look at doing that tomorrow | 20:37 |
| jrosser | -m setup is the 'ah-hoc' way of running the setup module directly from the command line | 20:37 |
| jrosser | when you tried it you did redo boostrap-aio.sh ? | 20:37 |
| martinkennelly | no, i did not redo bootstrap-aio.sh. | 20:38 |
| *** kplant has joined #openstack-ansible | 20:38 | |
| martinkennelly | sorry, I did redo bootstrap-aio.sh, nothing changed, i still got the error. | 20:40 |
| jrosser | i'm quite stumped here really, we do actually gather facts here right at the start of bootstrap-aio https://github.com/openstack/openstack-ansible/blob/master/tests/bootstrap-aio.yml#L18 | 20:42 |
| *** hamzaachi has joined #openstack-ansible | 20:42 | |
| jrosser | logan-: do you have any idea about this ^^ undefined ansible_default_ipv4? | 20:43 |
| *** altlogbot_0 has quit IRC | 20:43 | |
| *** altlogbot_2 has joined #openstack-ansible | 20:44 | |
| logan- | maybe delete cached facts or pastebin the cached fact file for that host | 20:45 |
| logan- | in case that hasn't been suggested yet, haven't read much backscroll | 20:45 |
| logan- | martinkennelly: does your ad-hoc setup command work without '-c local'? | 20:46 |
| logan- | martinkennelly: actually, if you go into the osa playbooks directory and run ansible, it will use the osa inventory, so could you cd to that dir and run "ansible localhost -m setup" | 20:47 |
| *** spatel has quit IRC | 20:52 | |
| *** ansmith has quit IRC | 20:57 | |
| *** altlogbot_2 has quit IRC | 21:00 | |
| *** tosky has joined #openstack-ansible | 21:02 | |
| *** altlogbot_2 has joined #openstack-ansible | 21:04 | |
| CeeMac | guilhermesp: just a quick question, to fix your comments I would update the files / git add / git commit --amend / git review again? | 21:05 |
| CeeMac | jrosser: same question :) | 21:05 |
| *** pcaruana has quit IRC | 21:05 | |
| guilhermesp | basically you just go where you were editing the files and do it the workflow as you mentioned. If you don't have the repo anymore, you just clone it again, go to the top corner of you patch and download your changes | 21:07 |
| *** ivve has quit IRC | 21:07 | |
| *** hamzaachi has quit IRC | 21:08 | |
| *** martinkennelly has quit IRC | 21:09 | |
| CeeMac | Ok thanks, will do in workflow, repo clone and local branch is still on my laptop | 21:18 |
| CeeMac | Should I be concerned about those zuul failures? | 21:18 |
| *** ianychoi has quit IRC | 21:23 | |
| *** ianychoi has joined #openstack-ansible | 21:28 | |
| *** hamzaachi has joined #openstack-ansible | 21:40 | |
| CeeMac | Erm. What do I need to do to submit my replies to comments on the review? /noobquestions | 21:44 |
| CeeMac | guilhermesp: ^ | 21:51 |
| CeeMac | nvm, found the 'reply' button.... | 22:04 |
| mnaser | jrosser: i think we are running out of memory in the octavia gate | 22:09 |
| mnaser | jrosser: http://logs.openstack.org/51/664551/1/check/openstack-ansible-deploy-aio_metal_octavia-ubuntu-bionic/e374421/logs/openstack/instance-info/free_22-28-08.log.txt.gz | 22:09 |
| mnaser | 260mb free.. that will do it | 22:09 |
| mnaser | guilhermesp, jrosser: yeah we do run stein magnum.. idk. | 22:10 |
| johnsom | mnaser He posted a patch to drop the per-amp memory for the gates as your tests don't really stress the amp (not a lot of listeners, etc.). https://review.opendev.org/667633 | 22:11 |
| mnaser | dirk: yes, no xenial jobs should exist at this point in master | 22:11 |
| johnsom | Ah, you found it. | 22:12 |
| mnaser | CeeMac: awesome first contribution!!! let me know if you need help cleaning it up too :) | 22:13 |
| mnaser | :) | 22:14 |
| mnaser | i was going through queue | 22:14 |
| mnaser | lol | 22:14 |
| *** weifan has quit IRC | 22:15 | |
| *** luksky has quit IRC | 22:16 | |
| CeeMac | thanks mnaser , will do. | 22:17 |
| CeeMac | once i get this one through, I think I have a couple more to submit soon :) | 22:18 |
| kplant | mnaser: you run a pretty large os deployment with a ceph backend right? | 22:19 |
| *** ansmith has joined #openstack-ansible | 22:20 | |
| *** weifan has joined #openstack-ansible | 22:29 | |
| *** hamzaachi has quit IRC | 22:38 | |
| mnaser | kplant: yessir | 22:38 |
| mnaser | a few | 22:38 |
| kplant | mind me asking how you manage qos with cinder? front-end qos seems very lack luster to me | 22:43 |
| kplant | once you drop front-end qos on a volume there's no more online retyping, per-gb-qos doesn't react to extending volumes until you detach/reattach | 22:43 |
| *** schwicht has joined #openstack-ansible | 22:48 | |
| openstackgerrit | Merged openstack/openstack-ansible-openstack_hosts stable/stein: Use stein RDO release repo https://review.opendev.org/665189 | 22:48 |
| *** schwicht has quit IRC | 22:50 | |
| openstackgerrit | Jimmy McCrory proposed openstack/openstack-ansible-os_horizon stable/stein: Fix distro installs on Ubuntu https://review.opendev.org/664696 | 23:04 |
| *** tosky has quit IRC | 23:12 | |
| openstackgerrit | Merged openstack/openstack-ansible-galera_server master: Remove xenial-ssl https://review.opendev.org/667682 | 23:42 |
| openstackgerrit | Merged openstack/openstack-ansible-galera_server master: Fix galera_server installation for openSUSE Leap 15.1 https://review.opendev.org/666272 | 23:42 |
| *** weifan has quit IRC | 23:57 | |
| *** hwoarang has quit IRC | 23:57 | |
| *** hwoarang has joined #openstack-ansible | 23:59 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!
