| *** sdake has quit IRC | 00:01 | |
| *** errr has joined #openstack-ansible | 00:09 | |
| *** markvoelker has quit IRC | 00:11 | |
| *** errr_ has quit IRC | 00:12 | |
| *** sdake has joined #openstack-ansible | 00:44 | |
| *** sdake has quit IRC | 01:06 | |
| *** markvoelker has joined #openstack-ansible | 01:09 | |
| *** markvoelker has quit IRC | 01:41 | |
| *** DanyC_ has quit IRC | 01:44 | |
| *** DanyC has joined #openstack-ansible | 02:00 | |
| *** DanyC has quit IRC | 02:04 | |
| *** sdake has joined #openstack-ansible | 02:07 | |
| *** sdake has quit IRC | 02:13 | |
| *** vollman has quit IRC | 02:19 | |
| *** markvoelker has joined #openstack-ansible | 02:38 | |
| openstackgerrit | Chandan Kumar proposed openstack/openstack-ansible-os_heat master: Fixed the egg name of heat to openstack_heat https://review.openstack.org/635518 | 02:45 |
|---|---|---|
| chandankumar | cloudnull: please have a look on this https://review.openstack.org/633883 and https://review.openstack.org/634377 | 02:46 |
| *** ArchiFleKs has quit IRC | 02:46 | |
| *** ArchiFleKs has joined #openstack-ansible | 02:57 | |
| *** hwoarang has quit IRC | 03:00 | |
| *** hwoarang has joined #openstack-ansible | 03:02 | |
| *** udesale has joined #openstack-ansible | 03:11 | |
| *** markvoelker has quit IRC | 03:12 | |
| *** dave-mccowan has joined #openstack-ansible | 03:13 | |
| *** gkadam has joined #openstack-ansible | 03:18 | |
| *** chhagarw has joined #openstack-ansible | 03:19 | |
| *** gkadam_ has joined #openstack-ansible | 03:28 | |
| *** gkadam has quit IRC | 03:31 | |
| *** ianychoi has quit IRC | 03:49 | |
| *** ianychoi has joined #openstack-ansible | 03:55 | |
| *** markvoelker has joined #openstack-ansible | 04:09 | |
| *** hwoarang has quit IRC | 04:19 | |
| *** hwoarang has joined #openstack-ansible | 04:20 | |
| *** vnogin has joined #openstack-ansible | 04:23 | |
| *** vnogin has quit IRC | 04:27 | |
| *** markvoelker has quit IRC | 04:41 | |
| *** komish has joined #openstack-ansible | 04:58 | |
| *** komish has quit IRC | 05:07 | |
| *** dave-mccowan has quit IRC | 05:08 | |
| *** gyee has quit IRC | 05:36 | |
| *** markvoelker has joined #openstack-ansible | 05:38 | |
| *** chandankumar is now known as chkumar|ruck | 05:41 | |
| openstackgerrit | Chandan Kumar proposed openstack/openstack-ansible-os_heat master: Fixed the egg name of heat to openstack_heat https://review.openstack.org/635518 | 05:43 |
| *** markvoelker has quit IRC | 06:12 | |
| *** vnogin has joined #openstack-ansible | 06:32 | |
| *** vnogin has quit IRC | 06:33 | |
| *** chkumar|ruck has quit IRC | 06:51 | |
| *** chandankumar has joined #openstack-ansible | 06:52 | |
| *** chandankumar is now known as chkumar|ruck | 06:52 | |
| *** kopecmartin|off is now known as kopecmartin | 06:58 | |
| *** markvoelker has joined #openstack-ansible | 07:09 | |
| *** DanyC has joined #openstack-ansible | 07:18 | |
| *** DanyC has quit IRC | 07:22 | |
| *** pcaruana has joined #openstack-ansible | 07:23 | |
| *** rgogunskiy has joined #openstack-ansible | 07:39 | |
| *** markvoelker has quit IRC | 07:41 | |
| *** fnpanic has quit IRC | 07:56 | |
| *** Emine has joined #openstack-ansible | 08:12 | |
| *** vnogin has joined #openstack-ansible | 08:25 | |
| *** markvoelker has joined #openstack-ansible | 08:38 | |
| *** sc has joined #openstack-ansible | 08:40 | |
| *** ianychoi has quit IRC | 08:50 | |
| *** Emine has quit IRC | 09:05 | |
| *** markvoelker has quit IRC | 09:12 | |
| *** fnpanic has joined #openstack-ansible | 09:22 | |
| *** DanyC has joined #openstack-ansible | 09:28 | |
| fnpanic | good morning | 09:29 |
| *** vnogin has quit IRC | 09:30 | |
| fnpanic | i have a issue with setup-openstack and tempest .... | 09:30 |
| fnpanic | from my understanding it should not run by default because of the condition in setup-openstack.yml | 09:31 |
| fnpanic | error message | 09:32 |
| fnpanic | http://paste.openstack.org/show/744730/ | 09:32 |
| fnpanic | https://github.com/openstack/openstack-ansible/commit/485100fbd36adbacd6f300a34a6acd9df3fd5932#diff-1e43c10bdb781c87ce1f9c1f71c2b4f0 | 09:33 |
| fnpanic | tempest_run: no was set by me | 09:33 |
| fnpanic | but it seems to be deftault :-( | 09:34 |
| fnpanic | https://review.openstack.org/382211 | 09:34 |
| fnpanic | this was the change of the install method | 09:34 |
| *** slaweq has joined #openstack-ansible | 09:42 | |
| *** tosky has joined #openstack-ansible | 09:45 | |
| *** markvoelker has joined #openstack-ansible | 10:10 | |
| *** vnogin has joined #openstack-ansible | 10:12 | |
| *** vnogin has quit IRC | 10:38 | |
| *** vollman has joined #openstack-ansible | 10:39 | |
| *** markvoelker has quit IRC | 10:42 | |
| *** vnogin has joined #openstack-ansible | 10:58 | |
| *** Emine has joined #openstack-ansible | 11:02 | |
| *** vnogin has quit IRC | 11:08 | |
| fnpanic | any idea on this topic? | 11:09 |
| fnpanic | :-( | 11:09 |
| *** udesale has quit IRC | 11:10 | |
| *** udesale has joined #openstack-ansible | 11:10 | |
| openstackgerrit | Chandan Kumar proposed openstack/openstack-ansible-os_heat master: Fixed the egg name of heat to openstack_heat https://review.openstack.org/635518 | 11:13 |
| openstackgerrit | Chandan Kumar proposed openstack/openstack-ansible-os_heat master: [DNM] heat tempest tests finding https://review.openstack.org/630694 | 11:32 |
| *** sdake has joined #openstack-ansible | 11:34 | |
| *** markvoelker has joined #openstack-ansible | 11:39 | |
| *** sdake has quit IRC | 11:41 | |
| *** sdake has joined #openstack-ansible | 11:44 | |
| *** Emine has quit IRC | 11:51 | |
| CeeMac | any iscsi experts in the house? | 11:56 |
| *** sdake has quit IRC | 12:11 | |
| *** markvoelker has quit IRC | 12:12 | |
| *** sdake has joined #openstack-ansible | 12:14 | |
| *** slaweq has quit IRC | 12:15 | |
| *** slaweq has joined #openstack-ansible | 12:18 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Define OSA clone dir in the openstack-ansible.sh script https://review.openstack.org/635821 | 12:18 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible stable/rocky: Define OSA clone dir in the openstack-ansible.sh script https://review.openstack.org/635821 | 12:20 |
| *** Adri2000 has quit IRC | 12:27 | |
| openstackgerrit | Michael Vollman proposed openstack/openstack-ansible-lxc_hosts master: Enable forwarding for all ipv4 interfaces https://review.openstack.org/635823 | 12:28 |
| *** Adri2000 has joined #openstack-ansible | 12:32 | |
| *** Emine has joined #openstack-ansible | 12:47 | |
| *** sdake has quit IRC | 12:48 | |
| fnpanic | hi | 13:01 |
| CeeMac | hi | 13:03 |
| *** partlycloudy has quit IRC | 13:04 | |
| *** vnogin has joined #openstack-ansible | 13:07 | |
| *** markvoelker has joined #openstack-ansible | 13:09 | |
| *** sawblade6 has quit IRC | 13:15 | |
| *** gary_perkins has quit IRC | 13:18 | |
| *** sawblade6 has joined #openstack-ansible | 13:19 | |
| *** gary_perkins has joined #openstack-ansible | 13:19 | |
| *** vnogin has quit IRC | 13:25 | |
| *** ianychoi has joined #openstack-ansible | 13:27 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-plugins master: Remove the config_template module https://review.openstack.org/635838 | 13:31 |
| *** fnpanic has quit IRC | 13:33 | |
| *** ArchiFleKs has quit IRC | 13:36 | |
| *** chkumar|ruck is now known as chkumar|out | 13:39 | |
| *** kmadac has joined #openstack-ansible | 13:39 | |
| *** markvoelker has quit IRC | 13:42 | |
| *** ArchiFleKs has joined #openstack-ansible | 13:42 | |
| *** hamzaachi has joined #openstack-ansible | 13:45 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests master: Use the config_template module from the dedicated repo https://review.openstack.org/635841 | 13:46 |
| *** vnogin has joined #openstack-ansible | 13:49 | |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-tests master: Use the config_template module from the dedicated repo https://review.openstack.org/635841 | 13:53 |
| openstackgerrit | Jesse Pretorius (odyssey4me) proposed openstack/openstack-ansible-plugins master: Remove the config_template module https://review.openstack.org/635838 | 13:54 |
| kmadac | Hello, I'm trying to modify a bit deployment of bird daemons for calico networking plugin in OSA. I need to get IP address of br-external bridge of haproxy nodes and write them to bird template. Unfortunately, i'm getting error during deployment that ansible_br_external is undefined. I can see that variable with correct valur in /etc/openstack_depl | 13:56 |
| kmadac | oy/ansible_facts/haproxy1. What could be the reason that it cannot be found during neutron deployment? | 13:56 |
| odyssey4me | logan- ^ | 13:57 |
| mnaser | https://review.openstack.org/#/c/634991/ appreciate if anyone can have a look, it's passing CI and tempest runs :> | 13:58 |
| openstackgerrit | Andy Smith proposed openstack/openstack-ansible master: Add qdrouterd role for rpc messaging backend deployment https://review.openstack.org/624184 | 13:59 |
| *** vnogin has quit IRC | 14:00 | |
| jamesdenton | QQ for those familiar with the functional tests. Running it in os_nova, I'm finding the VMs aren't able to hit the network. I see a task that blows out iptables on localhost, but nothing that comes back and implements a MASQUERADE. Any idea how the NAT is meant to occur? https://github.com/openstack/openstack-ansible-tests/blob/master/test-prepare-host.yml#L31-L35 | 14:11 |
| noonedeadpunk | folks, I've got a question regarding this release note https://github.com/openstack/openstack-ansible/blob/master/releasenotes/notes/haproxy-package-cache-backend-da096228387bc1f4.yaml | 14:13 |
| noonedeadpunk | it seems, that haproxy is still configured and expects repo_cache port 3142 to be up on repo container | 14:14 |
| mnaser | uh you seem to be right about that jamesdenton | 14:14 |
| noonedeadpunk | however, i don't see 3142 to be up on repo_container | 14:14 |
| noonedeadpunk | so does anybody know, if usage of this port is still relevat, and should be configured, as I didn't found any mentioning of this port except this release note and haproxy configuration https://github.com/openstack/openstack-ansible/blob/master/playbooks/common-tasks/package-cache-proxy.yml | 14:16 |
| noonedeadpunk | It seems to be relevant only for Centos, but I have it configured only with ubuntu everywhere as well | 14:17 |
| odyssey4me | noonedeadpunk that is relevant for releases prior to rocky where there is an apt/rpm cache in the repo container - the tasks there remove the config for it because we removed the cache in rocky | 14:25 |
| odyssey4me | going afk for the afternoon/evening, have a great w/end all! | 14:26 |
| jamesdenton | see ya odyssey4me | 14:28 |
| noonedeadpunk | odyssey4me: ok, thanks. It seems, that during Q->R upgrade configuration left in haproxy... | 14:28 |
| noonedeadpunk | I just didn't know, wheter it is safe to remove it now | 14:28 |
| noonedeadpunk | thanks) | 14:29 |
| *** dave-mccowan has joined #openstack-ansible | 14:29 | |
| mnaser | hey odyssey4me -- is it possible for us to get a job inside openstack/nova which runs our role functional tests? | 14:31 |
| mnaser | the reason is that they want a way to run a full deployment and test that it works with the built-in placement | 14:32 |
| prometheanfire | mnaser: cross job? | 14:34 |
| mnaser | prometheanfire: yeah | 14:34 |
| prometheanfire | mnaser: you could take a look at how we do them and just copypasta | 14:35 |
| mnaser | is there somewhere that we do them right now prometheanfire ? | 14:36 |
| prometheanfire | mnaser: requirements does them a bit | 14:36 |
| prometheanfire | .zuul.d/cross-jobs.yaml in the reqs repo | 14:36 |
| * mnaser checks | 14:37 | |
| mnaser | prometheanfire: looks like openstack ansible doesnt have a cross job listed tehre | 14:37 |
| mnaser | i remember one of our projects have a cross repo job | 14:38 |
| mnaser | s/projects/roles/ | 14:38 |
| *** sdake has joined #openstack-ansible | 14:39 | |
| *** markvoelker has joined #openstack-ansible | 14:39 | |
| *** kopecmartin is now known as kopecmartin|off | 14:39 | |
| prometheanfire | mnaser: oh, I just meant it as a way to template out creation of one for osa | 14:39 |
| prometheanfire | that's good though :D | 14:39 |
| mnaser | prometheanfire: cause we do some tricky magic to make sure we use the zuul cloned thingamajig | 14:40 |
| mnaser | instead of the one we clone | 14:40 |
| mnaser | openstack-ansible-cross-repo-functional | 14:40 |
| *** sdake has quit IRC | 14:42 | |
| jamesdenton | mnaser lxc-dnsmasq appears to setup the NAT. Then that task blows it out. May be worth restarting lxc-dnsmasq right after that task to get he NAT back | 14:42 |
| jamesdenton | unless this isn't the way to be running the functional test | 14:43 |
| vollman | jamesdenton: are you running os_nova functional on centos? | 14:47 |
| jamesdenton | yes. i saw your patch for ip forwarding this morning, if that's what you're alluding to | 14:47 |
| openstackgerrit | Merged openstack/openstack-ansible-os_placement master: Add initial placement role https://review.openstack.org/618820 | 14:48 |
| jamesdenton | vollman Besides some stupid virtualenv issue i also ran into, if things are left alone the tests bail here: http://paste.openstack.org/show/744742/ | 14:49 |
| jamesdenton | And from the looks of it, the NAT setup by lxc-dnsmasq is no longer there. | 14:50 |
| *** sdake has joined #openstack-ansible | 14:52 | |
| jrosser | thankyou for spending time on this btw, I am out of the office travelling today so haven’t been able to follow it up | 14:52 |
| jrosser | What you describe is the same as I saw with the NAT missing | 14:52 |
| jamesdenton | i will compare with bionic and see if there's a difference | 14:53 |
| *** rgogunskiy has quit IRC | 14:53 | |
| vollman | jamesdenton: that is. I have only been testing with distro_install because it runs to completion with just the ssh timeout | 14:55 |
| *** gary_perkins has quit IRC | 14:55 | |
| vollman | with that scenario the lxc nat is setup fine. only the br-vlan forwarding is missing | 14:56 |
| *** gary_perkins has joined #openstack-ansible | 14:56 | |
| *** gary_perkins has quit IRC | 14:56 | |
| jamesdenton | gotcha | 14:56 |
| vollman | and | 14:56 |
| *** gary_perkins has joined #openstack-ansible | 14:56 | |
| cloudnull | fnpanic: that issue (http://paste.openstack.org/show/744730/) was likely caused by missing facts. | 14:56 |
| vollman | I have to update that change because enabling forwarding for all didnt work. only explicitly enabling forwarding for br-vlan | 14:57 |
| cloudnull | anyone around to review - https://review.openstack.org/#/q/topic:bug/1794320+(status:open+OR+status:open) | 14:57 |
| jamesdenton | vollman interesting. maybe because the interface was already existing? | 14:57 |
| *** hamzaachi has quit IRC | 14:57 | |
| jamesdenton | although you'd think 'default' would be responsible for that | 14:58 |
| vollman | yea it certainly wasnt what I was expecting | 14:58 |
| *** Emine has quit IRC | 15:05 | |
| logan- | kmadac: hello, still around? | 15:10 |
| *** markvoelker has quit IRC | 15:12 | |
| kmadac | @logan: yes i'm here, but in the meantime I found the solution. I have to run setup task on all haproxy nodes to gather facts and it works now. | 15:15 |
| logan- | great | 15:15 |
| jamesdenton | jrosser I think i know what's happening. On a clean system, if you run the func tests, the 'iptables clearing' operation is executed before apt/yum is run and lxc-dnsmasq is installed. If something bombs during the run and you re-run the tests, the clearing still happens but lxc-dnsmasq is not restarted, and we lose the NAT. It's safe to assume in the gate if the functional test fails it's never rerun and this isn't | 15:16 |
| jamesdenton | an issue. | 15:16 |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/ansible-role-systemd_service master: Build out the PrivateNetwork function for services https://review.openstack.org/635396 | 15:22 |
| *** pcaruana has quit IRC | 15:31 | |
| *** spatel has joined #openstack-ansible | 15:33 | |
| spatel | jamesdenton: i think you might be right person to explain this http://paste.openstack.org/show/744746/ | 15:33 |
| spatel | why total hypervisor count is different? | 15:34 |
| spatel | stats vs list ? | 15:34 |
| *** vollman has quit IRC | 15:39 | |
| *** hamzaachi has joined #openstack-ansible | 15:41 | |
| jamesdenton | not sure. Are you getting unexpected info from the list command, even with the grep? | 15:42 |
| spatel | if you see this output http://paste.openstack.org/show/744746/ | 15:42 |
| jamesdenton | how many do you really have? | 15:43 |
| spatel | counts are differ in first output 179 compute and in second output when i list compute node it says 193 | 15:43 |
| spatel | i think 193 is correct number because i can see them... in my inventory | 15:43 |
| jamesdenton | k | 15:44 |
| spatel | does stats command doing something more interesting like working one or non-working on? | 15:44 |
| jamesdenton | right, maybe? Do you have any in a down or disabled state? | 15:44 |
| spatel | only 3 nodes are down | 15:46 |
| spatel | let me drill and see | 15:46 |
| jrosser | jamesdenton: so you think there is something different failing in the actual gate test? | 15:50 |
| *** priteau has joined #openstack-ansible | 15:50 | |
| *** sdake has quit IRC | 15:52 | |
| *** Emine has joined #openstack-ansible | 15:52 | |
| *** sdake has joined #openstack-ansible | 15:52 | |
| jamesdenton | Hard to say, I haven't seen those logs. Just observing the funkiness locally. As for the lxc-dnsmasq thing, i can see the trigger to restart it here, https://github.com/openstack/openstack-ansible-tests/blob/master/test-prepare-host.yml#L209-L215, but can't find the handler 'Restart dnsmasq' | 15:53 |
| jamesdenton | So i guess what i meant by my earlier comment is that rerunning the functional test locally results in behaviors that may never been seen in the gate (due to the lack(?) of dnsmasq restart). | 15:54 |
| *** sdake has quit IRC | 15:56 | |
| *** sdake has joined #openstack-ansible | 15:58 | |
| *** priteau has quit IRC | 16:00 | |
| *** vnogin has joined #openstack-ansible | 16:01 | |
| jrosser | this is where the gate check fails http://logs.openstack.org/74/633474/4/check/openstack-ansible-functional-centos-7/6bdccc6/logs/openstack/infra1/stestr_results.html | 16:02 |
| jrosser | "Failed to establish authenticated ssh connection to cirros@10.1.3.161" | 16:02 |
| *** vnogin has quit IRC | 16:05 | |
| *** markvoelker has joined #openstack-ansible | 16:09 | |
| CeeMac | any of you guys working with custom themes in horizon? | 16:11 |
| *** TxGirlGeek has joined #openstack-ansible | 16:13 | |
| openstackgerrit | Merged openstack/ansible-role-python_venv_build master: Add tripleo-ci-centos-7-standalone-os-tempest job https://review.openstack.org/634377 | 16:18 |
| *** vollman has joined #openstack-ansible | 16:21 | |
| jamesdenton | jrosser thanks for the log. I'll get my local instance up and see if i can poke at it | 16:29 |
| *** shardy has quit IRC | 16:30 | |
| jamesdenton | Also, the 'Restart dnsmasq' handler is in lxc_hosts role. The notify listed here (https://github.com/openstack/openstack-ansible-tests/blob/master/test-prepare-host.yml#L209-L215) doesn't appear to be effective. I copied that 'post_tasks' block to Line 101, reran the test, and the iptables rules that were wiped appear to be re-implemented. Just verified on centos and ubuntu. But, I think this situation only comes up | 16:34 |
| jamesdenton | when the functional test is re-run on a host, but good to know | 16:34 |
| openstackgerrit | Merged openstack/openstack-ansible-os_tempest master: Ping router once it is created https://review.openstack.org/633883 | 16:36 |
| *** gyee has joined #openstack-ansible | 16:38 | |
| *** markvoelker has quit IRC | 16:43 | |
| *** hamzy has quit IRC | 16:47 | |
| *** aedc has joined #openstack-ansible | 16:48 | |
| *** aedc has quit IRC | 16:53 | |
| *** Emine has quit IRC | 17:03 | |
| *** vnogin has joined #openstack-ansible | 17:15 | |
| *** vnogin has quit IRC | 17:15 | |
| prometheanfire | I think some copypasta was not done right for the docs | 17:18 |
| prometheanfire | # nova_pci_passthrough_whitelist: '{ "physical_network": "<ml2 network name>", "devname": "<physical nic>" }' | 17:18 |
| prometheanfire | https://docs.openstack.org/openstack-ansible-os_nova/latest/ | 17:18 |
| *** hamzaachi has quit IRC | 17:30 | |
| *** gkadam_ has quit IRC | 17:32 | |
| *** DanyC has quit IRC | 17:39 | |
| *** DanyC has joined #openstack-ansible | 17:40 | |
| *** Emine has joined #openstack-ansible | 17:40 | |
| *** markvoelker has joined #openstack-ansible | 17:40 | |
| *** DanyC has quit IRC | 17:44 | |
| *** udesale has quit IRC | 17:44 | |
| *** sdake has quit IRC | 17:47 | |
| mnaser | someone say PASTA? | 18:01 |
| prometheanfire | mnaser: yep | 18:03 |
| prometheanfire | it also looks like the alias thing in os_nova just takes a list of strings | 18:03 |
| prometheanfire | feels odd that each string would be '{ "vendor_id":"8086", "product_id":"154d", "device_type":"type-PF", "name":"a1" }' | 18:04 |
| prometheanfire | I guess freeform stings make some sense | 18:04 |
| mnaser | it does | 18:07 |
| guilhermesp | arxcruz chkumar|out that is odd http://logs.openstack.org/82/627782/6/check/openstack-ansible-upgrade-aio_lxc-ubuntu-bionic/c034aa4/logs/ara-report/result/2c2866d8-62e8-43aa-b1ef-d0b8bd375783/ | 18:09 |
| guilhermesp | I have a hold with the failing job | 18:09 |
| guilhermesp | inside the vm, when I run tempest run -l using the stein workspace, it works, but the opposite of the error above happens in the old venv ( rocky venv ) | 18:10 |
| guilhermesp | apparently, after failing, works, but breaks in the upgrade process | 18:10 |
| cloudnull | guilhermesp that's odd... https://pasted.tech/pastes/d0e58eccd53b315c7a4f0fb00e844710063ef59d | 18:10 |
| cloudnull | module not instlled? | 18:11 |
| cloudnull | maybe an implicit dependency changed? | 18:11 |
| guilhermesp | could be...but I'm trying to find explanations too to see why in the hold of the job I'm logged in it works! :P | 18:11 |
| guilhermesp | lemme paste here | 18:12 |
| *** Emine has quit IRC | 18:12 | |
| *** markvoelker has quit IRC | 18:13 | |
| *** hamzy has joined #openstack-ansible | 18:14 | |
| guilhermesp | cloudnull: http://paste.openstack.org/show/744756/ exactly the opposite when it runs the first time | 18:16 |
| *** Emine has joined #openstack-ansible | 18:16 | |
| *** DanyC has joined #openstack-ansible | 18:17 | |
| *** hamzy has quit IRC | 18:22 | |
| *** DanyC has quit IRC | 18:22 | |
| *** hamzy has joined #openstack-ansible | 18:25 | |
| *** chhagarw has quit IRC | 18:25 | |
| openstackgerrit | Kevin Carter (cloudnull) proposed openstack/ansible-role-systemd_service master: Build out the PrivateNetwork function for services https://review.openstack.org/635396 | 18:26 |
| cloudnull | guilhermesp the error message is odd. | 18:32 |
| cloudnull | I'm thinking this has to do with the site packages? | 18:32 |
| cloudnull | like the module is installed in the site packages but is not working in the venv ? | 18:33 |
| cloudnull | on that node, can you import the module from the global python? | 18:33 |
| cloudnull | then try importing it from the venv python? | 18:33 |
| *** hamzaachi has joined #openstack-ansible | 18:36 | |
| *** Emine has quit IRC | 18:38 | |
| chkumar|out | cloudnull: guilhermesp I will take a look on weekend | 18:49 |
| chkumar|out | if not reproducable, we can hold the node running the job and debug it | 18:50 |
| *** electrofelix has quit IRC | 18:50 | |
| *** hamzy has quit IRC | 18:50 | |
| guilhermesp | chkumar|out: I already have the hold for that job, unless you are saying to recheck and follow all the execution | 18:51 |
| *** hamzaachi has quit IRC | 18:51 | |
| *** hamzaachi has joined #openstack-ansible | 18:51 | |
| *** kmadac has quit IRC | 18:54 | |
| *** Emine has joined #openstack-ansible | 18:55 | |
| chkumar|out | guilhermesp: it might be some issue with virtualenv | 18:55 |
| *** hamzaachi has quit IRC | 18:56 | |
| chkumar|out | guilhermesp: I will take a look tomorrow | 18:56 |
| chkumar|out | guilhermesp: is the same error coming on suse and centos also? | 18:56 |
| *** hamzy has joined #openstack-ansible | 18:58 | |
| cloudnull | oh that's a good question . | 19:01 |
| cloudnull | maybe a difference in virtualenv version ? | 19:01 |
| chkumar|out | cloudnull: what about rebuilding the venv , /me searching on google | 19:01 |
| cloudnull | if you just run virtualenv with the same path it should rewire it | 19:02 |
| jamesdenton | jrosser vollman So after some testing, it looks like on CentOS (or maybe networkd) systems the following happens: 1) when net.ipv4.conf.all.forwarding = 1, any new bridge that is created (via brctl) inherits 1. The same is not true, however, for bridge interfaces instantiated by networkd files. Adding the 'IPForward=1' to the .network file and restarting networkd enables forwarding [http://paste.openstack.org/show/744 | 19:04 |
| jamesdenton | 758/]. So the approach may be to edit the template here? https://github.com/openstack/ansible-role-systemd_networkd/blob/master/templates/systemd-network.j2. I'll leave that for you to decide | 19:04 |
| jamesdenton | http://paste.openstack.org/show/744758/ | 19:04 |
| guilhermesp | actually the patch just runs on a aio ubuntu bionic https://review.openstack.org/#/c/627782/6 | 19:05 |
| guilhermesp | cloudnull chkumar|out ^ | 19:05 |
| vollman | jamesdenton: working on a patch to systemd-network for that | 19:05 |
| jamesdenton | jrosser Regardless of the method, enabling ip_forward on br-vlan allowed the centos function test to pass, once the virtualenv and NAT issue were worked out | 19:06 |
| jamesdenton | vollman sweet | 19:06 |
| *** Emine has quit IRC | 19:06 | |
| vollman | Unfortunately the IPForward setting only sets all.forwarding and br-vlan.forwarding=0 | 19:07 |
| chkumar|out | guilhermesp: so here we are creating a new venv with new version , but in ideal scenario, the venv would be same? | 19:07 |
| jamesdenton | not that i've seen | 19:07 |
| chkumar|out | during uograde, just the package should be updated? | 19:07 |
| jamesdenton | with IPForward=1 in the .network file for br-vlan, restarting systemd-networkd enabled forwarding on br-vlan | 19:07 |
| *** markvoelker has joined #openstack-ansible | 19:10 | |
| *** hamzy has quit IRC | 19:10 | |
| vollman | jamesdenton: fantastic. That makes the change much simpler. When I tested it I thought it didn't work, but I just retested and it did apply forwarding | 19:10 |
| vollman | not sure what I did the first time | 19:10 |
| *** hamzy has joined #openstack-ansible | 19:11 | |
| jamesdenton | :) | 19:11 |
| jamesdenton | looks like it survived a reboot, too (as it should) | 19:12 |
| *** aedc has joined #openstack-ansible | 19:13 | |
| guilhermesp | chkumar|out: I don't think the ideal scenario would be keep the same versions between the venvs | 19:20 |
| *** hamzaachi has joined #openstack-ansible | 19:27 | |
| *** sdake has joined #openstack-ansible | 19:28 | |
| *** aedc has quit IRC | 19:35 | |
| *** hamzy has quit IRC | 19:37 | |
| *** hamzy has joined #openstack-ansible | 19:38 | |
| *** markvoelker has quit IRC | 19:44 | |
| *** sdake has quit IRC | 19:59 | |
| *** sdake has joined #openstack-ansible | 20:00 | |
| *** sdake has quit IRC | 20:09 | |
| *** sdake has joined #openstack-ansible | 20:11 | |
| *** sdake has quit IRC | 20:13 | |
| *** sdake has joined #openstack-ansible | 20:14 | |
| *** hamzy has quit IRC | 20:21 | |
| *** trident has quit IRC | 20:21 | |
| *** trident has joined #openstack-ansible | 20:22 | |
| *** hamzaachi has quit IRC | 20:25 | |
| *** trident has quit IRC | 20:34 | |
| *** trident has joined #openstack-ansible | 20:35 | |
| *** markvoelker has joined #openstack-ansible | 20:40 | |
| *** DanyC has joined #openstack-ansible | 20:42 | |
| *** DanyC has quit IRC | 20:43 | |
| *** sdake has quit IRC | 20:44 | |
| *** DanyC has joined #openstack-ansible | 20:45 | |
| *** sdake has joined #openstack-ansible | 20:46 | |
| openstackgerrit | Michael Vollman proposed openstack/ansible-role-systemd_networkd master: Add ipforward option to configure IPForward https://review.openstack.org/635929 | 20:50 |
| *** dave-mccowan has quit IRC | 20:50 | |
| *** dave-mccowan has joined #openstack-ansible | 20:52 | |
| *** ArchiFleKs has quit IRC | 20:54 | |
| *** hamzy has joined #openstack-ansible | 20:59 | |
| *** ArchiFleKs has joined #openstack-ansible | 21:02 | |
| *** markvoelker has quit IRC | 21:12 | |
| *** dave-mccowan has quit IRC | 21:18 | |
| *** DanyC has quit IRC | 21:39 | |
| *** DanyC has joined #openstack-ansible | 21:39 | |
| *** sdake has quit IRC | 21:58 | |
| *** markvoelker has joined #openstack-ansible | 22:10 | |
| *** TxGirlGeek has quit IRC | 22:11 | |
| *** hwoarang has quit IRC | 22:31 | |
| *** hwoarang has joined #openstack-ansible | 22:32 | |
| *** markvoelker has quit IRC | 22:42 | |
| *** aedc has joined #openstack-ansible | 23:02 | |
| *** spatel has quit IRC | 23:03 | |
| *** KeithMnemonic has quit IRC | 23:03 | |
| *** DanyC_ has joined #openstack-ansible | 23:11 | |
| *** DanyC has quit IRC | 23:15 | |
| *** aedc has quit IRC | 23:18 | |
| *** slaweq has quit IRC | 23:24 | |
| *** sdake has joined #openstack-ansible | 23:33 | |
| *** DanyC_ has quit IRC | 23:35 | |
| *** DanyC has joined #openstack-ansible | 23:51 | |
| *** DanyC has quit IRC | 23:55 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!