| *** markvoelker has joined #openstack-ansible | 00:11 | |
| *** markvoelker has quit IRC | 00:15 | |
| *** zirpu has joined #openstack-ansible | 00:24 | |
| *** markvoelker has joined #openstack-ansible | 00:26 | |
| poopcat | A little late but @allenb, can you login to the Galera container and check the '/etc/xinetd.d/mysqlchk' config and validate the IPs allowed. Perhaps adjusting 'only_from' to be 'only_from = 0.0.0.0/0' .... then restarting xinetd... | 00:26 |
|---|---|---|
| poopcat | If you run 'hatop -s /var/run/haproxy.stat' from the haproxy node and all the Galera VIPs are down, this is likely your problem | 00:27 |
| zirpu | cool. thanks. | 00:27 |
| poopcat | whoops @allanb | 00:27 |
| *** gyee has quit IRC | 00:41 | |
| *** ianychoi has quit IRC | 00:57 | |
| *** ianychoi has joined #openstack-ansible | 00:59 | |
| *** spsurya has joined #openstack-ansible | 00:59 | |
| *** ThiagoCMC has joined #openstack-ansible | 02:54 | |
| ThiagoCMC | Guys! The command `openstack-ansible setup-hosts.yml` is failing on Ubuntu 18.04 with HWE (Linux 5.0), error: "modprobe: FATAL: Module nf_conntrack_ipv4 not found in directory /lib/modules/5.0.0-25-generic" | 02:54 |
| ThiagoCMC | Any idea? | 02:54 |
| ThiagoCMC | The documented step `apt install linux-image-extra-$(uname -r)` also fails on newer Ubuntu systems! O_O | 02:55 |
| poopcat | I'm not sure if the 5.0 kernel is supported at the moment | 02:56 |
| ThiagoCMC | :-O | 02:56 |
| ThiagoCMC | The command `find /lib/modules* -iname "*nf_conntrack_ipv4*"` only returns: "/lib/modules/4.15.0-58-generic/kernel/net/ipv4/netfilter/nf_conntrack_ipv4.ko" | 02:56 |
| poopcat | https://docs.openstack.org/project-deploy-guide/openstack-ansible/stein/overview.html | 02:57 |
| ThiagoCMC | "Linux kernel version 3.13.0-34-generic or later is required." | 02:57 |
| poopcat | sure, I understand | 02:58 |
| poopcat | 5.0 kernel is pretty fresh on Ubuntu 18.04 | 02:58 |
| poopcat | try rolling back and re-running the setup-hosts | 02:58 |
| ThiagoCMC | Sure, but 4.18 isn't that fresh and it also doesn't have the module either. | 02:58 |
| ThiagoCMC | I'll rollback the kernel and try again, no problem... | 02:59 |
| poopcat | I've never tested with 4.18. Most of the clusters I'm on are 4.4 | 02:59 |
| ThiagoCMC | I was hoping to find a workaround on that Ansible task... | 02:59 |
| ThiagoCMC | The "openstack_hosts : Load kernel module(s)" | 03:00 |
| poopcat | you could work around it by commenting out the necessary checks... but that's probably not the way you'd want to go | 03:00 |
| ThiagoCMC | Sure, I understand that... =P | 03:01 |
| ThiagoCMC | Just curious... Looks like that the "nf_conntrack" does the job on 5.0? | 03:02 |
| *** ThiagoCMC has quit IRC | 03:07 | |
| *** BjoernT has joined #openstack-ansible | 03:12 | |
| *** BjoernT has quit IRC | 03:17 | |
| *** BjoernT has joined #openstack-ansible | 03:17 | |
| *** rohit02 has joined #openstack-ansible | 03:41 | |
| *** gkadam has joined #openstack-ansible | 03:54 | |
| *** gkadam has quit IRC | 03:54 | |
| *** BjoernT has quit IRC | 05:06 | |
| *** shyamb has joined #openstack-ansible | 05:06 | |
| *** kopecmartin|off is now known as kopecmartin | 05:08 | |
| *** udesale has joined #openstack-ansible | 05:10 | |
| *** BjoernT has joined #openstack-ansible | 05:12 | |
| *** udesale has quit IRC | 05:14 | |
| *** raukadah is now known as chkumar|rover | 05:17 | |
| *** dave-mccowan has quit IRC | 05:21 | |
| *** mathlin has joined #openstack-ansible | 05:29 | |
| *** BjoernT has quit IRC | 05:29 | |
| *** mathlin has left #openstack-ansible | 05:30 | |
| *** miloa has joined #openstack-ansible | 06:01 | |
| miloa | Morning | 06:02 |
| jrosser | morning | 06:02 |
| CeeMac | Morning | 06:23 |
| *** ianychoi has quit IRC | 06:34 | |
| *** ianychoi has joined #openstack-ansible | 06:34 | |
| miloa | Sorry to bother you on that : https://review.opendev.org/#/c/671762/ if someone can give me some hints to progress on it, many thanks (Zuul build fails on openstack-ansible-deploy-aio_ceph-ubuntu-bionic).. | 06:39 |
| *** arbrandes has quit IRC | 07:00 | |
| *** ianychoi has quit IRC | 07:00 | |
| *** shyamb has quit IRC | 07:01 | |
| *** arbrandes has joined #openstack-ansible | 07:01 | |
| *** ianychoi has joined #openstack-ansible | 07:01 | |
| jrosser | miloa: here is the error https://logs.opendev.org/62/671762/5/check/openstack-ansible-deploy-aio_ceph-ubuntu-bionic/77b48c0/logs/ara-report/result/1d6a3693-949d-428b-8840-dc5fd8a2a27d/ | 07:07 |
| jrosser | This has been fixed as far as I know | 07:08 |
| *** trident has quit IRC | 07:10 | |
| miloa | jrosser: Thanks:) so if i do a recheck, is it engough ? or may i have to do something more ? | 07:15 |
| jrosser | Try a recheck | 07:15 |
| miloa | jrosser: ok thanks :) | 07:15 |
| jrosser | And feisty I’ve not voted on the patch because it’s quite complex so is a difficult review | 07:15 |
| jrosser | Argh autocorrects | 07:15 |
| jrosser | *feisty/fwiw | 07:16 |
| openstackgerrit | Erik Berg proposed openstack/openstack-ansible-os_neutron master: l3 agent on network_hosts do dvr_snat, anywhere else dvr https://review.opendev.org/675274 | 07:17 |
| jrosser | The two code paths for keys provided either by the user or retrieved from the cluster are quite intermingled with overall complex logic | 07:17 |
| *** trident has joined #openstack-ansible | 07:17 | |
| *** shyamb has joined #openstack-ansible | 07:20 | |
| miloa | I tried to not change a lot of code :) and provide for example for nova key an output that can be used by the following task without change. | 07:22 |
| *** udesale has joined #openstack-ansible | 07:24 | |
| jrosser | miloa: you have a couple of other constructs, either block: or including a new task file for conditional “do this for cluster keys or this for user” and so long as both eventually do the same set_fact for the thing you need.... | 07:30 |
| *** ivve has joined #openstack-ansible | 07:39 | |
| ioni | noonedeadpunk, | 07:39 |
| ioni | noonedeadpunk, i found the issue related to cecph! | 07:39 |
| ioni | rbd create --size 1 --user openstack vms/test | 07:40 |
| ioni | 2019-08-21 10:37:26.384 7fac65535700 0 -- 172.29.100.200:0/3784294304 >> 172.29.100.200:6822/1328282 conn(0x56371bf1a690 :-1 s=STATE_CONNECTING_WAIT_CONNECT_REPLY_AUTH pgs=0 cs=0 l=1).handle_connect_reply connect got BADAUTHORIZER | 07:40 |
| ioni | noonedeadpunk, it's due to https://github.com/ceph/ceph-ansible/commit/a4f4dd75352411eae8509d32d5b73d1c1ff9b718 | 07:40 |
| *** rpittau|afk is now known as rpittau | 07:40 | |
| ioni | let me find the review and add a comment :D | 07:40 |
| *** shyamb has quit IRC | 07:54 | |
| *** zbr is now known as zbr|ooo | 07:56 | |
| *** shyamb has joined #openstack-ansible | 08:32 | |
| *** ianychoi has quit IRC | 08:32 | |
| *** ianychoi has joined #openstack-ansible | 08:33 | |
| *** shyam89 has joined #openstack-ansible | 08:45 | |
| jrosser | ioni: interesting, drop a link to the review when you've done that :) | 08:46 |
| ioni | jrosser, done that | 08:46 |
| *** shyamb has quit IRC | 08:46 | |
| ioni | jrosser, here: https://review.opendev.org/#/c/675785/1 | 08:46 |
| *** rohit02 has quit IRC | 08:46 | |
| ioni | jrosser, i think ceph guys messed up | 08:47 |
| ioni | https://github.com/ceph/ceph-ansible/issues/2296 | 08:47 |
| jrosser | do you understand what they tried to do? | 08:47 |
| ioni | # ceph auth caps client.<ID> mon 'allow r, allow command "osd blacklist"' osd '<existing OSD caps for user>' | 08:47 |
| ioni | https://docs.ceph.com/docs/master/releases/luminous/#upgrade-from-jewel-or-kraken | 08:47 |
| ioni | this | 08:47 |
| jrosser | they leave an unhelpful commit message waying what they did (which is obvious from the code!) but not why :/ | 08:47 |
| jrosser | *saying | 08:48 |
| ioni | it seems that they implemented a staging for upgrade | 08:48 |
| ioni | i think | 08:48 |
| ioni | but they added the permission to osd instead of mon | 08:48 |
| ioni | that's what i'm getting | 08:48 |
| ioni | maybe i'm understading poorly | 08:48 |
| ioni | i'm new to ceph | 08:48 |
| jrosser | so it would be possible to override those vars in OSA | 08:50 |
| ioni | jrosser, i did lost 6 days of my life trying to figuring out why the hell doesn't work | 08:50 |
| ioni | jrosser, i did that :) | 08:50 |
| jrosser | i can believe it :/ | 08:51 |
| jrosser | so is this breaking out gates anywhere? | 08:51 |
| jrosser | there are broken ceph jobs i think currently | 08:51 |
| ioni | the sha bump i believe is uses ceph for cinder volume | 08:51 |
| ioni | and cinder breaks to run migrations because cinder-volume is down | 08:52 |
| ioni | and is down due to failling to connect on ceph | 08:52 |
| ioni | but the logs don't tell you anything | 08:52 |
| ioni | i had to redefine the rados timeout to 10 secunds instead of -1 (doesn't timeout) | 08:52 |
| ioni | to see timeouts | 08:52 |
| ioni | then i tried to create manually the volumes | 08:53 |
| ioni | admin worked | 08:53 |
| ioni | cinder didn't work | 08:53 |
| ioni | fun times | 08:53 |
| ioni | i'm strating to hate ceph | 08:53 |
| ioni | *starting | 08:54 |
| jrosser | so if i've understood correctly we could apply the fix you have already done to here https://opendev.org/openstack/openstack-ansible/src/branch/master/tests/roles/bootstrap-host/templates/user_variables_ceph.yml.j2 | 08:54 |
| jrosser | and that would unblock our ceph jobs? | 08:54 |
| *** gkadam has joined #openstack-ansible | 08:54 | |
| ioni | https://paste.xinu.at/Mch/ | 08:54 |
| ioni | this | 08:54 |
| ioni | jrosser, works pretty ok | 08:56 |
| jrosser | great work digging into this btw | 08:56 |
| jrosser | would you like to make a patch for OSA? or I can take a look? | 08:57 |
| *** gkadam is now known as gkadam-brb | 08:57 | |
| ioni | jrosser, if is ok with you, can you do it? | 08:57 |
| jrosser | is this affexting master too, interestingly the ceph jobs appear to pass there | 08:58 |
| jrosser | the same patch is in stable-4.0 https://github.com/ceph/ceph-ansible/commit/642851fa5d4199a15cca019419afacb73541556e | 09:00 |
| *** gkadam-brb is now known as gkadam | 09:00 | |
| ioni | maybe ceph 14 handles this differently ? | 09:01 |
| ioni | no idee | 09:01 |
| openstackgerrit | Maksim Malchuk proposed openstack/openstack-ansible-os_nova master: Fix cell_v2 discover_hosts when Ironic enabled https://review.opendev.org/677681 | 09:05 |
| *** jawad_axd has joined #openstack-ansible | 09:20 | |
| jrosser | ioni: so i'm a bit confused - here https://paste.xinu.at/Mch/ you didnt add the blacklist part to the mon section? | 09:28 |
| ioni | jrosser, i removed it | 09:28 |
| ioni | just to work | 09:28 |
| ioni | i removed it from osd and that's all | 09:28 |
| jrosser | right - but the upgrade notes say to add it to the mon section | 09:28 |
| ioni | i didn't add anything else | 09:28 |
| ioni | i just wanted to get over it and let them fix it | 09:29 |
| jrosser | hmm ok then i think i'll hold off on a patch - i can't test it here | 09:29 |
| ioni | jrosser, i'll test it, hold on | 09:29 |
| ioni | client.cinder | 09:34 |
| ioni | key: AQCo/FxdAAAAABAAu23GYtVFWPfb3NJtdo88ug== | 09:34 |
| ioni | caps: [mon] profile rbd, allow r, allow command "osd blacklist" | 09:34 |
| ioni | caps: [osd] profile rbd pool=volumes, profile rbd pool=vms, profile rbd pool=images | 09:34 |
| ioni | i think this? | 09:34 |
| ioni | no idee :D | 09:35 |
| *** Namrata has joined #openstack-ansible | 09:42 | |
| Namrata | Hi Folks,I am trying to install openstack ansible rocky release on ubuntu 18.04 Bionic beaver but Rocky Release using Ubuntu 18.04 with netplan so what do you suggest moving back to ifdown or netplan only? | 09:46 |
| admin0 | does anyone have a tenant cleanup script or something equivalent ? | 10:03 |
| *** Namrata has quit IRC | 10:05 | |
| *** shyam89 has quit IRC | 10:05 | |
| *** Namrata has joined #openstack-ansible | 10:08 | |
| miloa | jrosser: this time i've got Zuul RETRY_LIMIT Error on openstack-ansible-deploy-aio_ceph-ubuntu-bionic | 10:11 |
| *** rpittau is now known as rpittau|bbl | 10:14 | |
| noonedeadpunk | jrosser: ioni btw, I have PR to ceph-ansible regarding common topic https://github.com/ceph/ceph-ansible/pull/4329 | 10:20 |
| noonedeadpunk | So I guess we may ask folks why this patch has been done, and, probably rollback it there | 10:20 |
| noonedeadpunk | admin0: do you mean smth like https://github.com/openstack/ospurge ? | 10:26 |
| noonedeadpunk | Namrata: I'd suggest running either netplan or ifupdown, not both | 10:27 |
| Namrata | noonedeadpunk so any of them would work fine | 10:28 |
| Namrata | Thanks | 10:28 |
| noonedeadpunk | yep, there shouldn't be problems, since OSA normally do not configure networking (unless you're running script for CI) | 10:29 |
| *** shyamb has joined #openstack-ansible | 10:30 | |
| Namrata | noonedeadpunk got it Thanks | 10:34 |
| jrosser | noonedeadpunk: oh goodness feels like theres a lot of mess right now in the ceph ansible stuff | 10:39 |
| noonedeadpunk | ++ | 10:39 |
| jawad_axd | jrosser: how to use ceph rbd namespaces along with pool names in OSA ? Is it supported yet or not? I tried with poolname/namespace but no success. | 10:52 |
| jrosser | miloa: looks like you hit a CI bug https://review.opendev.org/#/c/677669/ | 10:55 |
| jrosser | i've rechecked your job now that is merged | 10:55 |
| miloa | jrosser: Many thanks :) | 10:56 |
| *** owalsh is now known as owalsh|away | 11:08 | |
| ioni | noonedeadpunk, so it seems that you were on the right patch regarding cinder volume and ceph | 11:19 |
| ioni | noonedeadpunk, but the osd blacklist is the problem here | 11:19 |
| *** shyamb has quit IRC | 11:20 | |
| noonedeadpunk | there just more than one problem in these 4 lines:) | 11:20 |
| noonedeadpunk | ioni: As I understood from yours conversation, we're waiting for jrosser to test things out? | 11:21 |
| jrosser | well i'm running ceph natilus here | 11:21 |
| jrosser | and OSA stein | 11:21 |
| ioni | noonedeadpunk, he asked me if i added osd blacklist into mon and I said, no | 11:21 |
| jrosser | the gate breaks with stein/mimic so i can't reproduce | 11:21 |
| ioni | i just removed it completly | 11:21 |
| jrosser | and the gate works with master/nautilus, so i'm a bit confused right now about what is going on | 11:22 |
| noonedeadpunk | ioni: and you have mimic on your deployment? | 11:23 |
| jrosser | noonedeadpunk: i was thinking we could temporarily add a override into our user_variables_ceph to unblock the gate | 11:23 |
| ioni | noonedeadpunk, yes | 11:23 |
| jrosser | but i'm just not really sure what "right" looks like there in light of ioni bug and your PR | 11:23 |
| noonedeadpunk | jrosser: yeah, since it affects only mimic I think it would be faster and easier | 11:23 |
| ioni | noonedeadpunk, i just used whatever is default and stein | 11:23 |
| ioni | noonedeadpunk, and natilus seems to new for me | 11:24 |
| *** udesale has quit IRC | 11:26 | |
| *** udesale has joined #openstack-ansible | 11:27 | |
| *** shyamb has joined #openstack-ansible | 11:42 | |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Temporary override of ceph openstack_keys https://review.opendev.org/677707 | 11:47 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible stable/stein: Temporary override of ceph openstack_keys https://review.opendev.org/677709 | 11:47 |
| jrosser | ioni: noonedeadpunk ^ lets see how they go. Gross templating a template :( | 11:48 |
| noonedeadpunk | jrosser: do we rly need this for master? Can we just place a patch for stein only? | 11:50 |
| * jrosser shrugs, there’s the same errors in ceph ansible for both branches | 11:51 | |
| noonedeadpunk | but we have broken gates ony for stein...? | 11:52 |
| jrosser | Correct, but would you also like to validate the pool permission changes in master? | 11:52 |
| noonedeadpunk | I hoped to get PR to ceph-ansible merged:)) | 11:53 |
| noonedeadpunk | but yeah, lets leave it as for now:) | 11:53 |
| jrosser | perhaps the commit mesage should be clearer | 11:55 |
| jrosser | i've combined the osd blacklist and your PR changes into one there | 11:55 |
| *** udesale has quit IRC | 12:02 | |
| admin0 | noonedeadpunk, thanks .i will check ospurge out | 12:05 |
| *** rpittau|bbl is now known as rpittau | 12:10 | |
| openstackgerrit | Merged openstack/openstack-ansible-apt_package_pinning master: Update invalid link for README https://review.opendev.org/677419 | 12:14 |
| openstackgerrit | Merged openstack/openstack-ansible-plugins master: Update invalid link for README https://review.opendev.org/677416 | 12:16 |
| openstackgerrit | Merged openstack/openstack-ansible-galera_client master: Update invalid link for README https://review.opendev.org/677421 | 12:26 |
| *** udesale has joined #openstack-ansible | 12:32 | |
| openstackgerrit | Maksim Malchuk proposed openstack/openstack-ansible-os_nova stable/stein: Fix cell_v2 discover_hosts when Ironic enabled https://review.opendev.org/677714 | 12:35 |
| openstackgerrit | Maksim Malchuk proposed openstack/openstack-ansible-os_nova stable/rocky: Fix cell_v2 discover_hosts when Ironic enabled https://review.opendev.org/677716 | 12:37 |
| *** shyam89 has joined #openstack-ansible | 12:37 | |
| *** shyamb has quit IRC | 12:37 | |
| *** dr_feelgood has joined #openstack-ansible | 12:38 | |
| openstackgerrit | Maksim Malchuk proposed openstack/openstack-ansible-os_nova stable/queens: Fix cell_v2 discover_hosts when Ironic enabled https://review.opendev.org/677717 | 12:40 |
| *** gkadam has quit IRC | 12:42 | |
| *** shyam89 has quit IRC | 12:55 | |
| openstackgerrit | Merged openstack/openstack-ansible-nspawn_hosts master: Update openSUSE job https://review.opendev.org/677165 | 13:04 |
| *** dr_feelgood has quit IRC | 13:12 | |
| *** ThiagoCMC has joined #openstack-ansible | 13:13 | |
| ThiagoCMC | Good morning guys! | 13:15 |
| ThiagoCMC | I'm trying to deploy Stein on Ubuntu 18.04 with Linux 4.15 (since with 4.18 and 5.0 is failing) and now, 3 containers doesn't start, error: "networkd-dispatcher[1514]: ERROR:Failed to get interface "a3730f1a_eth0" status: Command '['/bin/networkctl', 'status', '--no-pager', '--no-legend', '--', 'a3730f1a_eth0']' returned non-zero exit status 1." | 13:16 |
| ThiagoCMC | Any idea? | 13:16 |
| admin0 | ThiagoCMC, ovs, lb ? | 13:33 |
| ThiagoCMC | lb | 13:33 |
| admin0 | with or without netplan ? | 13:34 |
| ThiagoCMC | Only Glance, Cinder API and Vol are failing, all the other containers are up and running. | 13:34 |
| ThiagoCMC | Netplan | 13:34 |
| ThiagoCMC | Cinder vol is a container because I'm not using LVM/iSCSI, it'll be Ceph-based. | 13:34 |
| ThiagoCMC | My openstack_deploy is basically this one: https://github.com/tmartinx/openstack_deploy | 13:35 |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova master: Fix cell_v2 discover_hosts when Ironic enabled https://review.opendev.org/677681 | 13:38 |
| openstackgerrit | Merged openstack/openstack-ansible-os_nova master: Update invalid link for README https://review.opendev.org/677138 | 13:38 |
| *** zbr|ooo is now known as zbr | 13:47 | |
| *** udesale has quit IRC | 13:47 | |
| *** jawad_axd has quit IRC | 13:48 | |
| *** jawad_axd has joined #openstack-ansible | 13:49 | |
| *** jawad_axd has quit IRC | 13:49 | |
| *** BjoernT_ has joined #openstack-ansible | 13:50 | |
| *** jawad_axd has joined #openstack-ansible | 13:50 | |
| *** jawad_ax_ has joined #openstack-ansible | 13:52 | |
| *** jawad_axd has quit IRC | 13:55 | |
| *** jawad_ax_ has quit IRC | 13:56 | |
| noonedeadpunk | mnaser: seems that I've missed smth alse, since patches to uwsgi are not handled by openstackgerrit IRC bot | 14:00 |
| *** pvradu has joined #openstack-ansible | 14:00 | |
| *** pvradu has quit IRC | 14:05 | |
| mnaser | noonedeadpunk: yeah that's another thing | 14:07 |
| mnaser | noonedeadpunk: https://github.com/openstack/project-config/blob/master/gerritbot/channels.yaml#L315-L333 | 14:07 |
| noonedeadpunk | mnaser: great, thanks | 14:08 |
| *** gregwork has quit IRC | 14:14 | |
| *** gundalow has quit IRC | 14:15 | |
| *** tiffanie has quit IRC | 14:15 | |
| *** strattao has quit IRC | 14:16 | |
| *** mgagne has quit IRC | 14:16 | |
| *** strattao has joined #openstack-ansible | 14:16 | |
| *** logan- has quit IRC | 14:16 | |
| *** irclogbot_1 has quit IRC | 14:17 | |
| *** irclogbot_1 has joined #openstack-ansible | 14:17 | |
| *** logan_ has joined #openstack-ansible | 14:17 | |
| *** mgagne has joined #openstack-ansible | 14:17 | |
| *** ebbex has quit IRC | 14:18 | |
| *** ebbex has joined #openstack-ansible | 14:18 | |
| *** logan_ is now known as logan- | 14:18 | |
| *** tiffanie has joined #openstack-ansible | 14:18 | |
| *** gregwork has joined #openstack-ansible | 14:18 | |
| *** mhayden has quit IRC | 14:19 | |
| *** Jeffrey4l_ has quit IRC | 14:19 | |
| *** gundalow has joined #openstack-ansible | 14:19 | |
| *** Jeffrey4l has joined #openstack-ansible | 14:21 | |
| *** mhayden has joined #openstack-ansible | 14:26 | |
| *** pvradu has joined #openstack-ansible | 14:36 | |
| *** Namrata has quit IRC | 14:37 | |
| *** chkumar|rover is now known as raukadah | 14:40 | |
| noonedeadpunk | argh, erlang broke our gates... | 14:47 |
| *** pvradu has quit IRC | 14:51 | |
| admin0 | noonedeadpunk, best i thought was to install in one of the utility containers ( as it has admin rc ) .. created a venv .. but cannot install because the packages are not in the osa repo | 14:58 |
| noonedeadpunk | admin0: are you running on rocky? | 14:59 |
| admin0 | yes | 14:59 |
| noonedeadpunk | Just add --isolated flag to pip. It's not the best workaround, but the easiest one | 15:00 |
| allanb | thanks @poopcat that worked. next bug involves the cinder migrations. searching for the logs. | 15:18 |
| *** ivve has quit IRC | 15:21 | |
| *** gyee has joined #openstack-ansible | 15:22 | |
| *** pvradu has joined #openstack-ansible | 15:25 | |
| poopcat | make sure you add the override 'galera_monitoring_allowed_source' to make it persistent | 15:26 |
| poopcat | @allanb https://docs.openstack.org/openstack-ansible-galera_server/latest/ | 15:26 |
| allanb | cool. was just looking for that. | 15:26 |
| poopcat | ThiagoCMC: Unrelated what you're trying to solve for but I was looking for this for you last night but you had logged off by the time I got it: https://github.com/openstack/openstack-ansible-openstack_hosts/blob/master/vars/ubuntu-16.04.yml | 15:31 |
| poopcat | errr https://github.com/openstack/openstack-ansible-openstack_hosts/blob/master/vars/ubuntu-18.04.yml | 15:31 |
| ThiagoCMC | Cool! I downgraded to 4.15 and it's all good now! | 15:31 |
| ThiagoCMC | I guess that for Linux 5.0, we could just remove those modules, since they don't exists anymore. | 15:32 |
| ThiagoCMC | Also, I'm planning to go OpenvSwitch with DPDK, so, no need to conntract modules at all... | 15:32 |
| allanb | where would i find the cinder migration logs? that my latest error. | 15:44 |
| allanb | "some migrations failed unexpectedly. see log." container /var/log has nothing. | 15:44 |
| noonedeadpunk | allanb: ioni was the last to debug this problem, since it's pretty repeatable | 15:48 |
| allanb | thanks. it's weird because it says there are none to do. this is a fresh install. | 15:49 |
| ioni | the migration failing on cinder is due to cinder volume being down | 15:49 |
| ioni | openstack volume service list | 15:49 |
| allanb | ah. | 15:49 |
| ioni | and was down due to cinder user(ceph) not able to connect on volumes pool | 15:50 |
| ioni | that was my case and the case of gates tests | 15:50 |
| ioni | i have to go now | 15:50 |
| allanb | thanks. cheers. | 15:51 |
| *** mgagne has quit IRC | 15:54 | |
| *** mgagne has joined #openstack-ansible | 15:55 | |
| *** pvradu has quit IRC | 15:58 | |
| *** pvradu has joined #openstack-ansible | 16:02 | |
| *** KeithMnemonic1 has joined #openstack-ansible | 16:02 | |
| *** KeithMnemonic has quit IRC | 16:03 | |
| *** rpittau is now known as rpittau|afk | 16:05 | |
| *** pvradu has quit IRC | 16:05 | |
| *** markvoelker has quit IRC | 16:08 | |
| allanb | seems to be a rabbitmq issue. progress 1 step at a time. :-) | 16:08 |
| ThiagoCMC | Is there any docs about how to deploy OSA with Manila and Ceph? | 16:12 |
| ThiagoCMC | I have the `openstack_deploy/conf.d/manila.yml` but, I don't know what the `manila-data_hosts` means... lol | 16:13 |
| noonedeadpunk | ThiagoCMC: manila-data_hosts it's kinda storage hosts iirc | 16:14 |
| ThiagoCMC | I already deployed OSA Rocky with Ceph but, only for Glance and Cinder. | 16:14 |
| ThiagoCMC | noonedeadpunk, right... =) | 16:14 |
| ThiagoCMC | Is it required even when planning to use Ceph only? | 16:14 |
| noonedeadpunk | unfortunatelly there's no docs yet, and manila role is still not really solid, since we're just working on it's CI | 16:14 |
| ThiagoCMC | That's okay, I would love to give it a try and give some feedback | 16:15 |
| noonedeadpunk | ThiagoCMC: yep, since it's the hosts where manila-share will be instaled | 16:17 |
| noonedeadpunk | https://opendev.org/openstack/openstack-ansible-os_manila/src/branch/master/defaults/main.yml#L248 | 16:18 |
| ThiagoCMC | Got it... Is it a container as well? | 16:18 |
| ThiagoCMC | I'm thinking about just deploying it on the shared controllers too | 16:19 |
| noonedeadpunk | By default yes, but you may set is_metal: true as usual | 16:20 |
| ThiagoCMC | Awesome, thanks! | 16:20 |
| ThiagoCMC | I'm facing another problem now... lol | 16:20 |
| ThiagoCMC | The task: "python_venv_build : Upgrade pip/setuptools/wheel to the versions we want" is failing. Playbook: utility-install.yml | 16:21 |
| ThiagoCMC | Part of the error is: "NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7f343e340690>: Failed to establish a new connection: [Errno 113] No route to host" | 16:21 |
| ThiagoCMC | Weird, I can ssh into the host and `ping google.com` normally... | 16:21 |
| ThiagoCMC | "FAILED - RETRYING: Upgrade pip/setuptools/wheel to the versions we want (1 retries left)." | 16:22 |
| ThiagoCMC | :-( | 16:22 |
| *** markvoelker has joined #openstack-ansible | 16:22 | |
| ThiagoCMC | I think that I found the problem... It's trying to reach my internal IP: "http://10.224.241.250:8181/os-releases/19.0.2.dev22/ubuntu-18.04-x86_64" but, might not be ready yet (should the haproxy, I guess) | 16:22 |
| ThiagoCMC | But haproxy-install.yml worked | 16:23 |
| ThiagoCMC | :-/ | 16:23 |
| ThiagoCMC | Haproxy is only listening on "haproxy_keepalived_external_vip_cidr", not the "internal" one | 16:26 |
| guilhermesp | arxcruz raukadah what would be the reason for tempest init not create content on the workspace dir? http://paste.openstack.org/show/761099/ | 16:26 |
| arxcruz | checking | 16:27 |
| ThiagoCMC | Typo... lol | 16:27 |
| guilhermesp | I have a hold on the bionic distro job that is failing after aug 5 https://review.opendev.org/#/c/672948/ | 16:27 |
| guilhermesp | digging into the logs | 16:28 |
| guilhermesp | i see that breaks with rc 3 in the temepst workspace init task | 16:28 |
| guilhermesp | inside the hold, that paste above provides me this output | 16:28 |
| guilhermesp | while in a local vm, running the gate-check-commit script it executes normally and populates /root/workspace | 16:29 |
| arxcruz | guilhermesp: which version of tempest ? from git ? | 16:29 |
| guilhermesp | not sure if I can find tempest git version as it is a distro job | 16:32 |
| guilhermesp | http://paste.openstack.org/show/761101/ | 16:33 |
| guilhermesp | that's the version of the packages | 16:33 |
| noonedeadpunk | guilhermesp: I'm wondering if https://review.opendev.org/#/c/675353/ can be related to the mentioned problem | 16:34 |
| guilhermesp | I was taking a look at this noonedeadpunk earlier | 16:34 |
| guilhermesp | bth I'm not sure, but it is the only change merged after the failures | 16:34 |
| guilhermesp | only related change merged* | 16:35 |
| noonedeadpunk | Since it added extra tempest checks, and I'm worried that it could interfer with build-in inside ubuntu pckages tests | 16:35 |
| guilhermesp | maybe we can revert that change and see what happens | 16:37 |
| noonedeadpunk | I guess rebase on the SHA before merge should do the same I guess | 16:38 |
| raukadah | guilhermesp: can you try tempest init -d | 16:38 |
| raukadah | ? | 16:38 |
| guilhermesp | sure raukadah | 16:38 |
| raukadah | tempest workspace list also? | 16:38 |
| guilhermesp | is it a valid parameter for that version? http://paste.openstack.org/show/761102/ | 16:40 |
| guilhermesp | will try to list the workspace | 16:40 |
| guilhermesp | "tempest workspace list" doenst provice me output | 16:40 |
| raukadah | guilhermesp: can you try tempest init --debug | 16:40 |
| guilhermesp | but in fact tempest workspace is listed in my local tests but not in the hold | 16:41 |
| guilhermesp | will try raukadah | 16:41 |
| raukadah | sorry it was --debug not -d | 16:41 |
| guilhermesp | no output | 16:41 |
| *** spsurya has quit IRC | 16:43 | |
| raukadah | guilhermesp: http://paste.openstack.org/show/761103/ can add me to that box | 16:43 |
| raukadah | ? | 16:43 |
| guilhermesp | sure | 16:43 |
| guilhermesp | a sec | 16:43 |
| guilhermesp | raukadah: ssh root@104.130.127.203 | 16:44 |
| raukadah | guilhermesp: http://paste.openstack.org/show/761104/ | 16:46 |
| raukadah | check foobar and tempest.log | 16:47 |
| raukadah | something wrong with plugin registration | 16:48 |
| raukadah | that's why workspace dir is not getting created | 16:48 |
| noonedeadpunk | and this feels like related thing to mixed setup... | 16:48 |
| guilhermesp | I've seen that shortly earlier... yeah maybe revert noonedeadpunk 's patch to see what happens | 16:48 |
| raukadah | yes might be | 16:49 |
| raukadah | let me try to reproduce locally with tempest and heat-tempest-plugin | 16:49 |
| *** Garyx has joined #openstack-ansible | 16:49 | |
| raukadah | noonedeadpunk: when you say about mix setup means mixing deb pkgs and git for tempest and its plugins? | 16:51 |
| noonedeadpunk | yep. so tempest is installed from repo packages, but all plugins come from git | 16:51 |
| *** stuartgr has joined #openstack-ansible | 16:52 | |
| noonedeadpunk | So I guess we might be just missing heat check before | 16:52 |
| raukadah | noonedeadpunk: please never do that | 16:52 |
| noonedeadpunk | guilhermesp: actually, we have empty whitelist... https://storage.gra1.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/logs_48/672948/3/check/openstack-ansible-deploy-aio_distro_metal-ubuntu-bionic/df07761/logs/etc/host/tempest/whitelist.yaml.txt.gz | 16:52 |
| raukadah | noonedeadpunk: I have tried to solve the same puppet side i think this year | 16:53 |
| raukadah | install all stuff from git | 16:53 |
| noonedeadpunk | raukadah: so I'd say there just no other way to get tests for ubuntu | 16:53 |
| raukadah | noonedeadpunk: ok, will look into that on friday may be, I am on pto tomorrow | 16:54 |
| noonedeadpunk | for example barbican was failing due to no tests https://review.opendev.org/#/c/673491/ | 16:54 |
| noonedeadpunk | so for some services this seems necessary thing... | 16:55 |
| *** ivve has joined #openstack-ansible | 16:56 | |
| noonedeadpunk | But probably we need to have separate list of plugins to be installed for mixed setups (or do excludes from existing one) | 16:56 |
| noonedeadpunk | But I'm worried about empty whitelist actually as well, since it should have some tests in it | 16:57 |
| raukadah | may be we need to think what we want to test there and then install? | 16:58 |
| noonedeadpunk | I guess that's exactly what happens... | 16:59 |
| noonedeadpunk | Like https://opendev.org/openstack/openstack-ansible-os_tempest/src/branch/master/vars/main.yml#L79-L83 | 16:59 |
| ThiagoCMC | What can cause connectivity issues between containers on the same host? | 17:00 |
| ThiagoCMC | I mean, I have the "br-mgmt" at the host, IP X, the utility container have its eth1 on same subnet as br-mgmt, utility can ping host, but utility can't ping other containers... | 17:01 |
| ThiagoCMC | ...On same bridge. | 17:01 |
| jrosser | ThiagoCMC: we saw something maybe similar on centos with forwarding not enabled on the bridge | 17:02 |
| jrosser | but never on ubuntu | 17:03 |
| noonedeadpunk | I had common things with infiniband, but I don't thing you've tried to build obsoleted eIPoIB :) | 17:03 |
| ThiagoCMC | jrosser, right, have you fixed it on centos? | 17:03 |
| allanb | i'm having a bridge issue, but between the controller and storage nodes. i can ping the controller from storage on its bridge ip, but not connect to the container (rabbitmq) on the controller from the storage. | 17:03 |
| guilhermesp | so noonedeadpunk raukadah the action now is to hold it? Maybe we dont need the hold from infra anymore as the issue was spotted right? | 17:03 |
| allanb | also: ubuntu1804, stable/stein. | 17:03 |
| ThiagoCMC | allanb, same versions here, Linux 4.15 | 17:04 |
| jrosser | ThiagoCMC: we did fix it - you'd have to look back through the history i'm afraid as it was a while ago | 17:04 |
| jrosser | are you using netplan? | 17:04 |
| ThiagoCMC | yes, netplan | 17:04 |
| jrosser | hmmm well my strictly personal opinion is that i stuck with ifupdown on bionic | 17:05 |
| ThiagoCMC | Damn... lol | 17:05 |
| ThiagoCMC | ok | 17:05 |
| noonedeadpunk | ifupdown works on bionic for me pretty well:))) | 17:05 |
| allanb | +1 ifupdown. keep it simple. :-) | 17:05 |
| ThiagoCMC | I have the same Netplan for my KVM hosts, they can talk normally | 17:05 |
| noonedeadpunk | allanb: if only ifupdown on ubuntu could restart networking..... | 17:05 |
| allanb | yeah. there is that. | 17:06 |
| raukadah | guilhermesp: yes | 17:07 |
| raukadah | noonedeadpunk: guilhermesp with tempest + heat plugin works fine i install from git in local env | 17:09 |
| guilhermesp | how did you test it raukadah ? I have a local vm which I ran the gate-check-commit and seems to work fine | 17:10 |
| guilhermesp | about the workspace creation | 17:10 |
| raukadah | nope | 17:10 |
| noonedeadpunk | guilhermesp: so you're ok locally with distro scenario? | 17:11 |
| raukadah | just git clone tempest and heat tempest plugin and then install it in venv then temepst list-plugins and temepst init --debug | 17:11 |
| noonedeadpunk | Maybe we should try to rebase patch on master? | 17:11 |
| guilhermesp | noonedeadpunk: that what I ran scripts/gate-check-commit.sh aio_distro_metal_heat deploy distro | 17:11 |
| guilhermesp | in a local bionic server | 17:11 |
| guilhermesp | rebase could be an option, let me try | 17:11 |
| jrosser | allanb: there is a helper script to check that all the veth are connected - someone will have to fire you the name I’m on a mobile so searching for it is tricky | 17:12 |
| noonedeadpunk | loocs good | 17:12 |
| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_heat master: Fix keystone endpoint for heat servers https://review.opendev.org/672948 | 17:13 |
| noonedeadpunk | mnaser: jrosser guilhermesp - I've got uwsgi passing tests https://review.opendev.org/#/c/676410/ - they do not run role itself, since nothing use it yet... but for initial commit I guess it's ok | 17:13 |
| allanb | @jrosser thanks. | 17:13 |
| noonedeadpunk | at least linters and docs are ok:) | 17:14 |
| allanb | i think my problem is that the br-mgmt on storage01 node isn't passing to controller01 node. however, the ip for the br-mgmt interfaces ping between them. just doesn't pass packages to the rabbitmq container. | 17:14 |
| allanb | ip forwarding is turned on. | 17:14 |
| ThiagoCMC | Same here... | 17:15 |
| ThiagoCMC | Fixed! | 17:15 |
| ThiagoCMC | For some reason, I had to `apt purge docker.io ; apt autoremove -y ; reboot` | 17:16 |
| ThiagoCMC | Looks like that it's iptables rules at the containers host. | 17:16 |
| ThiagoCMC | Whew | 17:16 |
| ThiagoCMC | No idea why docker.io is installed. | 17:16 |
| noonedeadpunk | probably you've got magnum installed on the wrong host? | 17:18 |
| jrosser | Magnum itself doesn’t need that? | 17:20 |
| noonedeadpunk | yeah, you're right.... | 17:21 |
| noonedeadpunk | than I also have no idea why docker.io was there :p | 17:21 |
| ThiagoCMC | I don't have magnum enabled... | 17:23 |
| allanb | so walking through my bridge problem, hopefully someone sees something i'm missing. controller01 br-mgmt has 192.168.100.13, storage01 has br-mgmt 192.168.100.15. both can ping each other at those ips. | 17:25 |
| allanb | rabbitmq container has 192.168.100.173, pingable from controller01, not from storage01. 'ip r get 192.168.100.173' on storage01 shows it's routing through the br-mgmt i/f, but no icmp packets show up on controller01. | 17:26 |
| jrosser | allanb: you’ve not got an address collision somewhere have you? | 17:27 |
| ThiagoCMC | allanb, check if you have docker.io installed in you host | 17:27 |
| allanb | i'm tempted to manually add a route for that 1 ip on storage01 to route through controller01. no docker.io already checked. :) | 17:27 |
| jrosser | Check that your used_ips is correct | 17:27 |
| ThiagoCMC | allanb, what about your host's iptables rules? | 17:28 |
| jrosser | And then that the ip allocated in the inventory haven’t accidentally also been used on a host or some other of your hardware | 17:28 |
| allanb | iptables seem correct. btw, this is running on vagrant VMs. so it could be something weird there, but everything else is working up to this point in the last playbook, setup-openstack.yml | 17:28 |
| *** errr has quit IRC | 17:29 | |
| jrosser | allanb: well, people do struggle doing multinode things with VMware in a similar way | 17:30 |
| allanb | i'm really not understanding why the storage01 node isn't seeming to pass packets over the br-mgmt i/f for the rabbitmq ip. | 17:30 |
| *** errr has joined #openstack-ansible | 17:30 | |
| jrosser | Whatever magical networking under all that has to be set up to allow ip/mac that don’t belong to the vagrant box to pass | 17:30 |
| allanb | used ips for each of the bridge networks is setup for 10,15. | 17:31 |
| allanb | 10 thru 15. rest are free. | 17:31 |
| *** miloa has quit IRC | 17:31 | |
| jrosser | Can you ping any containers between vm? | 17:32 |
| allanb | no. but i'll verify. | 17:32 |
| jrosser | If you can’t get any cross vm ping between containers I’d pretty much say this was whatever is providing the inter-vm network getting in the way | 17:33 |
| allanb | yeah. that definitly doesn't work. | 17:33 |
| allanb | thanks. i'll continue digging there. | 17:34 |
| *** macz has joined #openstack-ansible | 17:36 | |
| *** allanb has quit IRC | 17:56 | |
| *** CeeMac has quit IRC | 17:59 | |
| *** allanb has joined #openstack-ansible | 18:04 | |
| ThiagoCMC | noonedeadpunk, hey buddy, just curious, the Manila with Ceph uses CephFS, which depends on Ceph Metadata, right? If yes, does OSA created "ceph-meta" containers as well? | 18:08 |
| *** markvoelker has quit IRC | 18:16 | |
| *** markvoelker has joined #openstack-ansible | 18:22 | |
| NobodyCam | Good Morning OSA folks, Are there entries in openstack_user_config.yml for ceilometer? | 18:25 |
| NobodyCam | doh... I found it in conf.d | 18:28 |
| *** andyfriar has joined #openstack-ansible | 18:30 | |
| *** markvoelker has quit IRC | 18:30 | |
| *** kopecmartin is now known as kopecmartin|off | 18:33 | |
| *** andyfriar has quit IRC | 18:50 | |
| jrosser | ThiagoCMC: nearly https://review.opendev.org/#/c/673369/ | 19:02 |
| ThiagoCMC | Nice! | 19:03 |
| jrosser | getting manila seeing / using that might be another thing entirely, but thats a start | 19:04 |
| ThiagoCMC | Do you think that It'll be ready for OpenStack Train, maybe in about 4 months? | 19:04 |
| jrosser | someone needs to put effort into finishing it | 19:04 |
| ThiagoCMC | Sure! I understand that this is new and I'm planning to dive into it and help. | 19:04 |
| jrosser | pactches welcome :) | 19:04 |
| ThiagoCMC | :-D | 19:04 |
| ThiagoCMC | I'm not a developer but, I can find bugs. lol | 19:05 |
| guilhermesp | noonedeadpunk raukadah seems that the rebased didn't work and now suse job is failing https://review.opendev.org/#/c/672948/ | 19:30 |
| *** markvoelker has joined #openstack-ansible | 19:44 | |
| *** markvoelker has quit IRC | 19:49 | |
| BjoernT_ | so it seems systemd_python==234 package does need libsystemd-dev and pkg-config packages now during keystone install in stein, I'm the only running into this ? | 19:50 |
| *** BjoernT_ is now known as BjoernT | 19:50 | |
| noonedeadpunk | guilhermesp: suse failed earlier but yes:( | 20:02 |
| guilhermesp | yeah noonedeadpunk .. | 20:06 |
| jrosser | BjoernT: if it blows up for you but not in CI then i'd be wondering why? | 20:10 |
| BjoernT | yeah I'm still debugging why this is a problem, the venv built on one of the nodes but not others | 20:11 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible master: Temporary override of ceph openstack_keys https://review.opendev.org/677707 | 20:17 |
| openstackgerrit | Jonathan Rosser proposed openstack/openstack-ansible stable/stein: Temporary override of ceph openstack_keys https://review.opendev.org/677709 | 20:19 |
| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_murano master: Initial commit to os_murano role https://review.opendev.org/677830 | 20:58 |
| NobodyCam | Any tricks to adding a package to a existing repo container with rebuilding all? want to add ceilometer | 20:59 |
| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_murano master: Initial commit to os_murano role https://review.opendev.org/677830 | 21:08 |
| BjoernT | @jrosser never mind, found it. lsyncd was not able to synchronize other repo hosts | 21:11 |
| *** altlogbot_2 has quit IRC | 21:16 | |
| *** BjoernT has quit IRC | 21:22 | |
| noonedeadpunk | guilhermesp so I'll try to take a closer look on tempest tomorrow | 21:25 |
| guilhermesp | right noonedeadpunk keep me updated too and let me know if I can help. | 21:28 |
| guilhermesp | i asked infra to rm the hold btw | 21:28 |
| guilhermesp | btw, am I missing something here https://review.opendev.org/#/c/677830/? | 21:28 |
| noonedeadpunk | Looking | 21:30 |
| goldenfri | Hi, I need to add a compute host to rocky running ubuntu 16.04, but only have the venvs for 18.04 is there a simple way to add the 16.04 ones? | 21:30 |
| goldenfri | (in my repo) | 21:30 |
| jrosser | i'm not sure what state the mixed distro support is right now | 21:31 |
| jrosser | "in the past" you needed one infra host (or at least on repo container) of each distro | 21:31 |
| goldenfri | awww thats not good | 21:31 |
| goldenfri | crap, I'm having a major issue with hosts with 2080TIs taking forever to spawn and don't have that issue using 16.04 with queens | 21:32 |
| jrosser | this may be less of an issue now the way that the venvs are built, but i don't know of anyone who's tinkered with this recently | 21:33 |
| jrosser | thing is the wheels are built on the repo container so thats a ton of C compiling which will be distro specific | 21:34 |
| *** ThiagoCMC has quit IRC | 21:35 | |
| goldenfri | :-/ thanks, I should have known this wouldn't be easy | 21:35 |
| jrosser | fwiw i don't see that with a T4 but obv. thats not the same kind of beast | 21:36 |
| *** altlogbot_1 has joined #openstack-ansible | 21:36 | |
| noonedeadpunk | guilhermesp: so I guess you'll need run_tests.sh just to basicly check linters | 21:36 |
| *** altlogbot_1 has quit IRC | 21:38 | |
| jrosser | noonedeadpunk: \o/ the ceph job on stein gets past cinder http://zuul.openstack.org/stream/8c95d6c4173b447fa10c3a46a1fd5039?logfile=console.log | 21:40 |
| *** altlogbot_2 has joined #openstack-ansible | 21:40 | |
| *** altlogbot_2 has quit IRC | 21:42 | |
| noonedeadpunk | jrosser: wow, great news | 21:44 |
| openstackgerrit | Guilherme Steinmuller Pimentel proposed openstack/openstack-ansible-os_murano master: Initial commit to os_murano role https://review.opendev.org/677830 | 21:45 |
| noonedeadpunk | jrosser: regarding your comment for tox.ini - we still have these functional local tests everywhere, and I can imagine that people may still test things that way | 21:48 |
| jrosser | yes i did wonder about that, what you'd do for a local test | 21:49 |
| noonedeadpunk | but you're absolutely right that they are not needed | 21:49 |
| jrosser | its not an issue really - so long as we know why its there | 21:50 |
| jrosser | the in-role functional tests are now basically abandonware | 21:50 |
| noonedeadpunk | And I feel like we didn't even update docs regarding the new way of testing - https://docs.openstack.org/openstack-ansible/latest/contributor/testing.html#roles-functional-or-scenario-testing | 21:51 |
| jrosser | good chance that some of those just don't work any more | 21:51 |
| jrosser | anyway, late here time to !computer, but thanks for the pointer to your ceph-ansible PR that looks to have really helped unlock things | 21:52 |
| noonedeadpunk | So I'd vote for updating docs at the first place and after providing cleanup | 21:53 |
| noonedeadpunk | (unfortunatelly I have pretty poor english so I don't feel myself as a technical writer at least in terms of grammar) | 21:54 |
| noonedeadpunk | yep, I also won't be online by the time when zuul report the result of testing, so will vote for this patch in the morning | 21:55 |
| *** ivve has quit IRC | 21:57 | |
| *** markvoelker has joined #openstack-ansible | 21:59 | |
| noonedeadpunk | guilhermesp: from what I see from tempest logs I'd say we don't test heat with tempest at all for source ubuntu https://storage.gra1.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/logs_48/672948/4/check/openstack-ansible-deploy-aio_metal-ubuntu-bionic/048c685/logs/openstack/aio1-utility/test_list.txt.txt.gz | 22:16 |
| *** devsum has joined #openstack-ansible | 22:20 | |
| noonedeadpunk | ok, that's time I had some rest:) | 22:20 |
| *** markvoelker has quit IRC | 22:24 | |
| *** threestrands has joined #openstack-ansible | 22:34 | |
| *** markvoelker has joined #openstack-ansible | 22:35 | |
| *** markvoelker has quit IRC | 22:40 | |
| *** macz has quit IRC | 23:00 | |
| *** ThiagoCMC has joined #openstack-ansible | 23:36 | |
| ThiagoCMC | Hey guys, I just installed Stein! However, at the Horizon login, the field "Domain" is empty and I can't type anything there... What can be? =P | 23:36 |
| ThiagoCMC | I have "horizon_keystone_multidomain_support" and "horizon_keystone_multidomain_dropdown" set to true on user_vars... | 23:37 |
| poopcat | ThiagoCMC: https://osaguide.shutdowndashr.com:8443/w/index.php/Horizon_/_Keystone_multi-domain_login | 23:38 |
| ThiagoCMC | Thanks! I'll give it a try. | 23:39 |
| ThiagoCMC | Still empty... :-( | 23:46 |
| poopcat | Darn --- the last I tried that on was Pike | 23:49 |
| poopcat | Using the Openstack client do you have multiple domains 'openstack domain list' | 23:50 |
| ThiagoCMC | I have the following domains: magnum, Default and heat. | 23:52 |
| ThiagoCMC | Since this is just a lab, I'll try to keep the defaults and comment out multidomain... | 23:52 |
| ThiagoCMC | But I'll keep your tip there, for next lab deployment. | 23:53 |
| ThiagoCMC | ^_^ | 23:53 |
| ThiagoCMC | Thanks anyway! | 23:53 |
| poopcat | ThiagoCMC: https://osaguide.info ... It's not very pretty but it's just some overrides mostly from Kilo>Pike ... a lot of them apply through to Stein but have really only been tested on those | 23:56 |
| poopcat | It's been a while since I updated it | 23:57 |
| ThiagoCMC | Nice! | 23:57 |
| ThiagoCMC | It's working now! Without the dropdown. | 23:57 |
| ThiagoCMC | Just a feedback, Horizon prints an error at the corner: "Error: Unable to retrieve limits information." | 23:59 |
| ThiagoCMC | But I'll check this later... | 23:59 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!