Thursday, 2020-01-30

*** igordc has quit IRC00:28
*** igordc has joined #openstack-kolla00:59
*** luksky has quit IRC01:04
*** ricolin has joined #openstack-kolla01:14
*** oyrogerg has joined #openstack-kolla01:15
*** ricolin has quit IRC01:17
*** ricolin_ is now known as ricolin01:18
*** rlljorge has joined #openstack-kolla01:19
*** rlljorge has quit IRC01:32
*** goldyfruit_ has quit IRC01:43
*** ktibi has joined #openstack-kolla02:53
*** ktibi has quit IRC02:57
*** tony31 has joined #openstack-kolla04:05
tony31hi - is there a reference architecture step-by-step setup guide for Kayobe Openstack deployment? The issue which I am facing is that the documentation gives detail about "you can do this" but there's no context to help me understand or learn why such configuration is done. I think what would help a lot is if there was a deployment walk-through, to04:07
tony31give a understanding of it. Then the documentation would make sense to a n00b like me :)04:07
tony31The error I have is `TASK [ssh-known-host : Scan for SSH keys] ***********************************************************************************************************************************************************04:08
tony31"ansible_loop_var": "item", "changed": false, "cmd": ["ssh-keyscan"], "delta": "0:00:00.020674", "end": "2020-01-30 12:01:22.079599", "item": "", "msg": "non-zero return code", "rc": 1, "start": "2020-01-30 12:01:22.058925", "stderr": "usage: ssh-keyscan [-46cHv] [-f file] [-p port] [-T timeout] [-t type]\n\t\t   [host | addrlist namelist] ...",04:08
tony31"stderr_lines": ["usage: ssh-keyscan [-46cHv] [-f file] [-p port] [-T timeout] [-t type]", "\t\t   [host | addrlist namelist] ..."], "stdout": "", "stdout_lines": []}`04:08
*** igordc has quit IRC04:08
tony31not sure where the scan is being done04:08
tony31On this page it should explain the architecture https://docs.openstack.org/kayobe/train/architecture.html - I suggest that this page has a basic diagram of the networks and components. There's little info on that page and the problem I face with that is further into the documentation it is very difficult to reference the components back to the04:10
tony31architecture.04:10
tony31I'm willing to help with this respect once I understand more about it.04:11
*** skramaja has joined #openstack-kolla04:25
*** ricolin has quit IRC04:42
*** pbing19 has joined #openstack-kolla05:41
*** cah_link has joined #openstack-kolla06:59
*** luksky has joined #openstack-kolla07:14
*** waleedm_ has joined #openstack-kolla07:15
mnasiadkaMorning07:29
tony31morning07:40
tony31I'm trying to set up the `seed` for deploying Openstack through `kayobe`. Does anyone know what the task does `TASK [ssh-known-host : Scan for SSH keys]`07:44
tony31I can connect with ssh fine so I am confused about this scanning for ssh keys - where is it scanning?07:44
*** bengates has joined #openstack-kolla08:08
*** bengates has quit IRC08:08
*** bengates has joined #openstack-kolla08:09
openstackgerritMerged openstack/kolla master: Deprecate OpenDaylight  https://review.opendev.org/70483808:13
*** rpittau|afk is now known as rpittau08:14
hrwyay!08:16
openstackgerritMarcin Juszkiewicz proposed openstack/kolla master: Get rid of Python 2 support  https://review.opendev.org/69131608:16
hrwrebased on HEAD08:16
osmanlicilegimorning08:17
*** vedup19 has joined #openstack-kolla08:18
*** tonythomas has joined #openstack-kolla08:27
openstackgerritAndreas Jaeger proposed openstack/kayobe master: Revert "Use OpenStack Train release"  https://review.opendev.org/70493608:34
yoctozeptomgoddard, mnasiadka, hrw: posted random mumblings on https://review.opendev.org/69354408:37
patchbotpatch 693544 - kolla-ansible - WIP: CentOS 8: Add deploy jobs in CI - 37 patch sets08:37
*** tumble has quit IRC08:39
mnasiadkayoctozepto: ekhm, seems I need to invest time in Fernet patch? :D08:39
yoctozeptomnasiadka: yeah, weird08:39
yoctozeptomnasiadka: I thought py3 compat merged?08:39
yoctozeptowhy suddenly such outbreak ;p08:39
mnasiadkait did08:39
yoctozeptoalso, why this and not ubuntu :D08:39
yoctozeptoso many question08:40
yoctozeptoso few answers ;D08:40
openstackgerritAndreas Jaeger proposed openstack/kayobe master: Update docs building  https://review.opendev.org/70493908:41
mnasiadkayoctozepto: https://zuul.opendev.org/t/openstack/build/d887ffc47b874c4fbdce0cd15e782749/log/primary/logs/docker_logs/keystone_fernet.txt#3208:42
mnasiadkayoctozepto: this might be the problem :D08:43
yoctozeptoxDDDD08:43
yoctozeptowe be booting that container lolz08:43
mnasiadkalol of the day08:44
tony31hey can anyone help me with a problem that I have with Kayobe?08:54
yoctozeptomgoddard, mnasiadka: ^08:55
mnasiadkatony31: ok, first things first - what is the kayobe release that you are using?08:56
tony31stable/train08:56
*** luksky has quit IRC08:58
tony31I am following: https://docs.openstack.org/kayobe/train/deployment.html08:58
tony31and trying to configure the seed host OS with `kayobe seed host configure`08:59
mnasiadkayeah, I see the error you are pasting08:59
mnasiadkawhich is a bit weird08:59
*** waleedm_ has quit IRC09:00
tony31mnasiadka - I had some issues with trying to do the seed previously and I did indeed have an error on my part. My error was that in the `inventory` the hostname of the seed was not correct and so I got this error. I did resolve it but I have the error again. Do I need to `kayobe control host bootstrap` if I say, add a host to the inventory or make09:00
tony31some change like that?09:00
mnasiadkayou can try running it, it won't make things worse09:01
tony31another issue I had was the `globals.yml` did not have the correct user for SSH-ing to the seed. that too is resolved, but gave a similar or same error09:01
tony31thanks - will try it09:01
tony31:)09:01
*** waleedm_ has joined #openstack-kolla09:02
*** gfidente has joined #openstack-kolla09:04
*** veduph has joined #openstack-kolla09:15
*** pbing19 has quit IRC09:18
*** ivve has joined #openstack-kolla09:19
mgoddardtony31: hi09:22
mgoddardyou were gone when I joined yesterday09:23
mgoddardhopefully I can help with any kayobe issues09:23
mgoddardjust reading back09:23
tony31hi mgoddard - sorry, it's 17:23 now and I normally wrap up at 5pm :)09:23
mgoddardah, where are you? seems like your day ends when mine begins :)09:23
tony31i tried the above and same error - rebooting the VMs now for no real reason09:24
tony31Perth, Australia09:24
tony31:)09:24
mgoddardoh nice09:24
*** waleedm__ has joined #openstack-kolla09:28
*** waleedm_ has quit IRC09:30
tony31same issue - at a loss :)09:32
tony31whats it trying to do when "scan for ssh keys" ?09:32
*** luksky has joined #openstack-kolla09:35
mgoddardit's trying to scan for the seed's host keys and will add them to known hosts09:36
tony31any idea where it's scanning?09:36
mgoddardit could be that there is no IP defined for the seed09:36
tony31I have the private key in ~/.ssh/id_rsa09:36
tony31I am using host name in the inventory, but I tried using an IP there and it gave the same result09:37
yoctozeptotony31: "scanning" probably means ssh tries to connect to it and grab them keys :D09:37
yoctozeptomgoddard: to confirm ^09:38
tony31it needs to use the SSH key to connect in the first place :)09:38
tony31so some more info on this setup:09:38
mgoddardare you using a seed host you have provisioned yourself?09:38
mgoddardwhich already has an IP?09:38
mgoddardif so, add the IP in network-allocations.yml09:38
tony31mgoddard - yes kind of, it's just a base CentOS 7 vm cloud qcow image09:39
mgoddardnetwork config is quite tricky09:39
tony31i've not installed anything extra in it - from the docs I understood that `kayobe seed host configure` does the set up09:39
mgoddardhave you defined your networks in networks.yml?09:40
tony31no i havent at all yet09:40
mgoddardok09:40
mgoddardhttps://github.com/stackhpc/a-universe-from-nothing/blob/stable/train/etc/kayobe/networks.yml09:40
mgoddardthat's the global network config for a 'flat' all in one network09:40
tony31ah top man09:41
mgoddardthen IPs are allocated here: https://github.com/stackhpc/a-universe-from-nothing/blob/stable/train/etc/kayobe/network-allocation.yml09:41
mgoddardif left empty, kayobe will allocate IPs for you. If the seed already has an IP you'll need to populate that file for it though09:41
tony31I was trying to figure out the networks this morning and was going around in circles. im lacking the context to figure it out. I was hoping to work forward to get some context and then take some jumps back to do the network part09:41
tony31the seed is a VM with a single IP at the moment09:41
mgoddardthat's fine09:42
tony31thanks - taking a look at the git link now09:42
mgoddardthe SSH keyscan problem occurs because there is no IP, so the command gets executed without an argument09:42
mgoddardwe could do a better job of detecting that and giving a suitable error09:42
mgoddardas you mentioned in your ML post, our docs are quite a good reference, but not much for beginners atm09:43
mgoddardI want to change that, and add a configuration walkthrough09:43
mgoddardbut also as you mention we need to improve our architecture overview09:43
mgoddardare you new to openstack or just kayobe?09:44
*** vedup19 has quit IRC09:46
mgoddardfull network config is documented here: https://docs.openstack.org/kayobe/latest/configuration/network.html09:48
tony31it now fails at a much earlier point09:53
tony31There's no host IP for the seed there, it's just to define the allocation pools and network names. I'm not yet at the stage of deploying, which is where I think that this networks.yml comes into play. I'm at the point of installing the seed09:54
tony31I deployed RDO's tripleo openstack about 3 years ago, the current version was Pike at that time.09:54
tony31so I'm not a complete newbie but I'm not an expert either09:55
tony31im brand new to kayobe09:55
mgoddardok, good to know for context09:55
mgoddardwhere is it failing now?09:58
tony31`TASK [Write host config to file] ********************************************************************************************************************************************************************10:01
tony31controller_extra_network_interfaces +\n    (controller_network_host_network_interfaces\n     if inventory_hostname in groups['network'] else [])) | reject('none') | unique | list }}\n'. Error was a <class 'ansible.errors.AnsibleError'>, original message: An unhandled exception occurred while templating '{{ ([public_net_name,\n     tunnel_net_name]10:01
tony31+\n    external_net_names) | reject('none') | unique | list }}\n'. Error was a <class 'ansible.errors.AnsibleError'>, original message: Unexpected templating type error occurred on ({{ ([public_net_name,\n     tunnel_net_name] +\n    external_net_names) | reject('none') | unique | list }}\n): can only concatenate list (not \"AnsibleUnicode\") to10:01
tony31list"}`10:01
tony31the section from the git that defined the ranges was not in the yml file i have at all, so I pasted it in and updated the ip ranges. I think it was not right to do that.10:02
tony31if I revert back to what I had previously and re-run, I can see that it determines that the VM is a CentOS VM when it prints this :10:03
tony311TASK [ip-allocation : Include RedHat family-specific variables] *************************************************************************************************************************************10:03
tony31*************************************************************************************************************************************10:03
tony31so I mean to say, it appears that it can SSH into the instance OK10:03
mnasiadkamgoddard: seems we need to run "rm /run/nologin" in keystone-ssh container on centos8 - https://zuul.opendev.org/t/openstack/build/d887ffc47b874c4fbdce0cd15e782749/log/primary/logs/docker_logs/keystone_fernet.txt#3210:06
mnasiadkaand any other ssh containers :)10:06
mgoddardmnasiadka: yeah, there is a patch for it10:07
mgoddardhttps://review.opendev.org/70484810:07
patchbotpatch 704848 - kolla - CentOS 8: Allow SSH access to keystone_ssh container - 1 patch set10:07
hrw+210:08
mnasiadkaso I'll set +w as well ;)10:08
hrwdid already10:08
mnasiadkaah, hrw was faster :)10:08
mgoddardtony31: possibly you have set external_net_names to a string rather than a list?10:08
openstackgerritMichal Nasiadka proposed openstack/kolla-ansible master: doc: external Ceph - add manila ceph caps note  https://review.opendev.org/70495310:09
openstackgerritSaravanan KR proposed openstack/kolla master: Add support to build netcontrold binary image  https://review.opendev.org/70472310:09
tony31@mgod10:10
tony31mgoddard - yes I had it wrong.10:11
tony31apologies10:11
tony31good point10:11
mgoddardnp10:11
tony31ok - it's a success! Thanks a lot for sticking with me!10:13
mgoddardgreat :)10:16
mgoddardalways a bit of a learning curve with openstack10:16
mgoddardthanks for feedback also, I'm making notes10:16
tony31The only problem now is that it's trying to connect to the seed OS via the first usable address I had inserted into `aio_allocation_pool_start:` which was not the seed IP10:17
tony31is it possible to make it use the inventory for `seed` ?10:18
yoctozepto11:16:26 <mgoddard> always a bit of a learning curve with openstack10:23
yoctozepto"a bit" is euphemistic10:23
openstackgerritMichal Nasiadka proposed openstack/kolla-ansible master: doc: external Ceph - add manila ceph caps note  https://review.opendev.org/70495310:23
mgoddardtony31: you need to add the seed's IP to network-allocation.yml10:24
yoctozeptomgoddard, mnasiadka: hmm, but still, that c8 ci job had depends-on on ssh fix yet it fails10:24
tony31:)  understood - I spent about 2 weeks deep diving into RDO/tripleo10:24
mgoddardyoctozepto: fails on keystone fernet?10:25
tony31mgoddard - ah I see that network-allocation.yml now - thats why it keeps trying the old IP. I'm sure that was not there earlier10:25
mgoddardhaven't checked yet10:25
yoctozeptomgoddard: yeah, the morning one10:25
yoctozeptosame issue10:25
mgoddarddoh10:25
yoctozeptoI thought you meant we need that patch10:25
yoctozeptoyet it depends-on it already10:25
yoctozeptoand failing10:25
mgoddardyoctozepto: replied to your c8 comments10:25
yoctozeptomgoddard: merging your c8 proposals ;-)10:26
mgoddardwe could get a trio10:26
mgoddardthanks :)10:26
mnasiadkayoctozepto: well, we have some deps on the way to merge, so you can do w-1 if you wish :)10:26
yoctozeptomnasiadka: right, done10:28
mnasiadkayoctozepto: I think I found the culprit :)10:31
tony31mgoddard thanks so much for your help today mate. I'll try and carry on tomorrow10:31
tony31heading off now. thanks again. respect10:32
mgoddardtony31: no problem. Mailing list is a good place for overnight questions - I'll reply to your message for the benefit of others10:32
tony31thank you10:32
openstackgerritMark Goddard proposed openstack/kolla master: CentOS 8: Use upstream Ceph/master  https://review.opendev.org/70433710:37
openstackgerritMark Goddard proposed openstack/kolla master: CentOS 8: Allow SSH access to keystone_ssh container  https://review.opendev.org/70484810:37
*** tony31 has quit IRC10:38
hrwmgoddard: thanks for doing fresh releases10:44
*** pcaruana has quit IRC10:46
mnasiadkaok, with https://review.opendev.org/#/c/704953/ - I think the external ceph has been simplified as much as possible10:48
patchbotpatch 704953 - kolla-ansible - doc: external Ceph - add manila ceph caps note - 2 patch sets10:48
mnasiadkaI mean that's the last patch I plan :)10:48
*** bengates has quit IRC10:49
*** bengates has joined #openstack-kolla10:51
openstackgerritMark Goddard proposed openstack/kolla-ansible master: WIP: CentOS 8: Add deploy jobs in CI  https://review.opendev.org/69354410:59
openstackgerritMark Goddard proposed openstack/kolla-ansible master: CI: build swift images for swift scenario  https://review.opendev.org/70496010:59
*** numans has joined #openstack-kolla11:05
mgoddardhrw: np11:12
*** rpittau is now known as rpittau|bbl11:14
yoctozeptomgoddard, mnasiadka: what about nova-ssh?11:14
mgoddardyoctozepto: good q11:15
yoctozeptowhat is it for anyway?11:15
yoctozepto(prime time for such question lol)11:15
mgoddardyoctozepto: is it for migration?11:16
yoctozeptomgoddard: I'm not sure, libvirt is open for remote I think11:17
yoctozeptomgoddard: can't find ref to its port except for itself11:17
yoctozeptomgoddard: yeah it has some mention of migration there11:18
yoctozeptomgoddard: but hardly believable used this way due to nobody else knowing its port11:18
hrwcan you check something for me on x86-64 openstack?11:19
hrwcirros image (0.4.0 or newer) with hw_firmware_type=uefi11:19
hrwstein+ is enough11:19
mgoddardyoctozepto: who knows11:22
mnasiadkayoctozepto, mgoddard: seems migration - https://docs.openstack.org/nova/pike/admin/ssh-configuration.html11:23
mnasiadkaups, an oldish doc :)11:23
mnasiadkahttps://docs.openstack.org/nova/train/admin/ssh-configuration.html#cli-os-migrate-cfg-ssh11:24
hrwdoes someone care about influxdb/telegraf images?11:26
openstackgerritMichal Nasiadka proposed openstack/kolla-ansible master: OVN Support  https://review.opendev.org/69684111:26
openstackgerritMark Goddard proposed openstack/kolla master: CentOS 8: Allow SSH access to keystone_ssh and nova_ssh  https://review.opendev.org/70484811:27
openstackgerritMark Goddard proposed openstack/kolla master: CentOS 8: Relax ironic iPXE EFI bootloader location  https://review.opendev.org/70496511:27
mnasiadkahrw: I'm pretty sure dougsz cares about influxdb, but have no clue about telegraf11:27
yoctozeptomnasiadka, mgoddard: yeah, but it seems config was never finished11:27
yoctozeptoI mean, nova can't really guess that port :P11:28
yoctozeptooriginially it seems it was on api interface11:28
yoctozeptothen it was moved to migration11:29
yoctozepto(which is the same by default anyways)11:29
yoctozeptobut no clear signs of usage11:29
yoctozepto"so that Compute service can use SSH to move disks to other nodes"11:29
yoctozeptoI believe we all run centralized storage then11:30
*** csmart has quit IRC11:30
mnasiadkayoctozepto: interesting :D11:32
mnasiadka(that we all run a container that is totally unused)11:33
yoctozeptomnasiadka: yeah, happens11:33
yoctozeptohrw, mnasiadka: based on scenarios etherpad nobody cares about telegraf11:33
yoctozeptodoes tripleo use it?11:33
hrwno idea11:34
hrwI just found out that I can enable both on aarch6411:34
yoctozeptothen no problem :-)11:34
openstackgerritMark Goddard proposed openstack/kolla-ansible master: WIP: CentOS 8: Add deploy jobs in CI  https://review.opendev.org/69354411:35
mgoddardtelegraf is part of the TICK stack :)11:36
mgoddardone of our many monitoring stacks11:36
yoctozeptomgoddard: we did a release?11:37
mnasiadkawell then, last meaningful commit to telegraf Dockerfile was done by berendt11:37
mgoddardyoctozepto: yep11:37
mgoddard6 in fact11:37
yoctozeptomgoddard: cool to know :-)11:37
mgoddardI shared the patch in the channel11:37
mnasiadkamgoddard: I guess it's just a question how many monitoring stacks do we want to deliver :)11:37
mgoddardas many as possible11:37
mgoddardwhy, are there some we are missing?11:38
yoctozepto12:37:38 <mgoddard> I shared the patch in the channel11:38
yoctozeptomissed then11:38
openstackgerritMarcin Juszkiewicz proposed openstack/kolla master: aarch64: enable influxdb and telegram for Debian/Ubuntu  https://review.opendev.org/70497011:38
yoctozeptomgoddard: could use docs that we have 'em11:38
yoctozepto;p11:38
yoctozeptohrw: telegraf11:38
yoctozeptonot gram11:39
yoctozeptomgoddard: btw, tick is... ? telegraf, influxdb, collectd, kibana?11:39
hrw;D11:39
mgoddardTelegraf, InfluxDB, Chronograf, Kapacitor11:40
mgoddardseems we are missing some11:40
openstackgerritMarcin Juszkiewicz proposed openstack/kolla master: aarch64: enable influxdb and telegraf for Debian/Ubuntu  https://review.opendev.org/70497011:40
hrwWho has x86-64 openstack cloud in usable condition?11:41
* hrw out11:42
yoctozeptomgoddard: then drop telegraf ;D11:43
yoctozeptoinfluxdb is for monasca11:44
*** pcaruana has joined #openstack-kolla11:48
openstackgerritMerged openstack/kolla master: CentOS 8: base image fixes for RabbitMQ, fluentd & InfluxDB  https://review.opendev.org/70478211:59
mgoddardwoop11:59
yoctozeptowe be merging ya11:59
*** veduph has quit IRC12:01
mgoddardupdated LP for new releases12:02
openstackgerritMichal Nasiadka proposed openstack/kolla-ansible master: WIP: Ceph-Ansible upgrade jobs  https://review.opendev.org/70498412:08
*** kplant has joined #openstack-kolla12:16
openstackgerritMichal Nasiadka proposed openstack/kolla-ansible master: Ansible lint: lines longer than 160 chars  https://review.opendev.org/70242912:20
*** pcaruana has quit IRC12:21
openstackgerritMark Goddard proposed openstack/kolla master: CI: Fix tag suffix logic for CentOS 8  https://review.opendev.org/70498712:22
mgoddardyoctozepto, hrw, mnasiadka, osmanlicilegi: ^12:22
mnasiadkamgoddard: nice bug12:23
mgoddardthanks12:23
mgoddardI made it myself12:23
mnasiadkamgoddard: if I would be yoctozepto-picky, I would ask for a launchpad bug and a release note :D12:23
mgoddardmnasiadka: but you are not :)12:23
* yoctozepto triggered12:23
mgoddardquick, someone +2A!12:24
yoctozeptomgoddard: did12:24
yoctozeptomnasiadka: those are for stuff that we have released12:25
yoctozeptono point in here12:25
mnasiadkayoctozepto: just laughing, I know it's CI :)12:25
openstackgerritMichal Nasiadka proposed openstack/kolla-ansible master: WIP: Ceph-Ansible upgrade jobs  https://review.opendev.org/70498412:30
mnasiadkalol, in https://review.opendev.org/#/c/699414/ upgrade-ceph job timed out, and it was just sending logs to swift when it timed out12:35
patchbotpatch 699414 - kolla - CentOS 8: Update packages in images - 10 patch sets12:35
mgoddarddamn, grafana has the DNF GPG issue too12:43
mgoddardah, repo_gpgcheck=112:44
yoctozeptodrop the conf <bass music ensues>12:46
yoctozeptobtw, folks, mnasiadka, mgoddard: I have this issue - I want to configure haproxy timeouts for e.g. nova_novncproxy - is there a clean way? must I copy all nova services definition to override one bit?12:47
mnasiadkayoctozepto: welcome to hell :)12:48
yoctozeptomnasiadka: I was expeciting something along the lines: "no, yoctozepto, you are just tired, here is how to do it"12:49
yoctozeptomnasiadka: and not hell visit12:49
openstackgerritMark Goddard proposed openstack/kolla master: CentOS 8: disable repo_gpgcheck for grafana image  https://review.opendev.org/70499912:49
mgoddard^ should unblock grafana builds12:50
mnasiadkayoctozepto: I was thinking one day how to overcome it, and we would need to rework this a bit, to allow the user to supply custom dict with just one changed variable and use combine()12:50
yoctozeptomgoddard: who cares about grafana when haproxy timeouts are not configurable :D12:50
mgoddardmnasiadka: +1, similar to volumes12:51
mnasiadkayoctozepto: so now you know how to fix it - go go go :)12:51
openstackgerritMark Goddard proposed openstack/kayobe master: WIP: Use python3 for local kolla-ansible execution  https://review.opendev.org/70500012:52
yoctozeptomnasiadka: yeah, I had a similar idea, though not volume-based - we need to override this stuff a bit12:53
mnasiadkayoctozepto: if we dislike the current approach - imagine what the users say ;)12:53
yoctozeptomnasiadka: policy prohibits any cursing12:54
mnasiadkayoctozepto: your university policy?12:54
yoctozeptomnasiadka: no, irc netiquette :-)12:54
*** goldyfruit_ has joined #openstack-kolla12:59
mnasiadkamgoddard: I just noticed we are running the ceph jobs always, while I marked ceph-ansible jobs to be ran only when roles using ceph are changed (and deploy-ceph-ansible.sh but it's irrelevant to this)13:06
mnasiadkamgoddard: are we okay with this behaviour? or should c-a jobs run always?13:07
mnasiadkayoctozepto: ^^13:07
yoctozeptomnasiadka: hmm13:07
yoctozeptoI think we would like to regularly test multinode deploys13:07
yoctozeptothough it's fine if they run periodics13:08
yoctozeptocould save ci cycles13:08
yoctozeptoI have no problem with that13:08
yoctozeptothe most fragile mariadb has its own tests13:08
yoctozeptoso we are covered13:08
mnasiadkawell upgrade-ceph is run in periodics13:08
mnasiadkabut not -ceph13:09
yoctozeptodoh, then one needs to run it13:09
mnasiadkawell, upgrade-ceph is longer, does an upgrade from previous release and is multinode13:09
mnasiadkaso we could add upgrade-ceph-ansible to periodics13:09
mnasiadkaand stick to that13:09
*** scottsol has quit IRC13:10
yoctozeptomnasiadka: +113:10
yoctozeptothough one might like to see deployment of current as multinode13:10
yoctozeptoI can see how we could break it :-)13:11
yoctozeptoand periodics run daily so it's not a big deal13:11
yoctozeptoespecially when we have rather solid stability compared to other projects13:11
*** vmixor has joined #openstack-kolla13:12
openstackgerritMichal Nasiadka proposed openstack/kolla-ansible master: OVN Support  https://review.opendev.org/69684113:16
openstackgerritMerged openstack/kolla-ansible master: Use listen port for Placement precheck  https://review.opendev.org/70459513:17
*** rpittau|bbl is now known as rpittau13:29
*** veduph has joined #openstack-kolla13:29
*** mixor has joined #openstack-kolla13:30
*** vmixor has quit IRC13:30
mgoddardmnasiadka, yoctozepto: I'd suggest that when we remove internal ceph tests we run ceph-ansible unconditionally13:43
mgoddardit's our multinode test13:43
mnasiadkamgoddard: ok, I’ll post a depends-on change to not forget :)13:43
mgoddardk13:43
mgoddardall centos8 deploy jobs (that have all required images) passing \o/13:45
bersacelol, searching kolla-ansible's github for openstack roles code triggers all ansible roles references :D13:49
bersaces/triggers/matches/13:49
*** veduph has quit IRC13:53
yoctozeptomgoddard: woop, gr8!13:57
mnasiadkamgoddard: so I guess it’s time for c8 c-a?13:59
mgoddardplease!14:00
*** pcaruana has joined #openstack-kolla14:03
*** veduph has joined #openstack-kolla14:05
openstackgerritAntony Messerli proposed openstack/kolla master: Adds python3-systemd for ansible deploy interface  https://review.opendev.org/70502114:21
openstackgerritMark Goddard proposed openstack/kolla-ansible master: Python 3: Use distro_python_version for dev mode  https://review.opendev.org/70502214:21
openstackgerritMark Goddard proposed openstack/kolla-ansible master: Python 3: Use distro_python_version for WSGI python_path  https://review.opendev.org/70502314:21
openstackgerritMark Goddard proposed openstack/kolla-ansible master: Python 3: Use distro_python_version for monasca agent CA file  https://review.opendev.org/70502414:21
openstackgerritMark Goddard proposed openstack/kolla-ansible master: Remove unused python path calculation from vmtp  https://review.opendev.org/70502514:21
*** heikkine has joined #openstack-kolla14:22
hrwmgoddard: one day someone will move that python path into vars.yml14:27
mgoddardvars.yml?14:28
openstackgerritMark Goddard proposed openstack/kayobe master: WIP: Use python3 for local kolla-ansible execution  https://review.opendev.org/70500014:28
hrwmgoddard: to whatever k-a uses for common vars14:29
mgoddardglobals.yml14:29
*** rlljorge has joined #openstack-kolla14:29
mgoddardor group_vars/all.yml14:29
mgoddardit's already there14:29
hrwsecond14:29
hrwah14:29
mgoddardor at least distro_python_path it14:29
mgoddardis14:29
hrwWho has x86-64 openstack cloud in usable condition?14:30
*** pcaruana has quit IRC14:34
*** cah_link has quit IRC14:43
*** veduph has quit IRC14:44
*** veduph has joined #openstack-kolla14:44
openstackgerritMark Goddard proposed openstack/kayobe master: WIP: Allow OVS bridges to connect directly to interface  https://review.opendev.org/70503714:56
*** waleedm__ has quit IRC14:58
*** ricolin has joined #openstack-kolla14:58
yoctozeptohrw: me15:00
yoctozeptohrw: which python path?15:00
hrwyoctozepto: priv15:00
openstackgerritMark Goddard proposed openstack/kayobe master: WIP: Use python3 for local kolla-ansible execution  https://review.opendev.org/70500015:01
*** pcaruana has joined #openstack-kolla15:02
*** veduph has quit IRC15:05
*** veduph has joined #openstack-kolla15:05
*** veduph has quit IRC15:22
openstackgerritMarcin Juszkiewicz proposed openstack/kolla master: nova-libvirt: add uefi package to support UEFI instances  https://review.opendev.org/68228015:28
openstackgerritMarcin Juszkiewicz proposed openstack/kolla master: nova: use proper uefi package names under CentOS/RHEL 8  https://review.opendev.org/70504015:28
*** dswebb has joined #openstack-kolla15:29
*** k_mouza has joined #openstack-kolla15:31
*** k_mouza has quit IRC15:36
*** TrevorV has joined #openstack-kolla15:44
*** ricolin has quit IRC15:45
*** priteau has joined #openstack-kolla15:45
*** Trevor_V has joined #openstack-kolla15:47
openstackgerritMerged openstack/kolla master: CI: Fix tag suffix logic for CentOS 8  https://review.opendev.org/70498715:48
*** TrevorV has quit IRC15:51
hrwmeh.. k-a tries /usr/bin/python ;(15:53
hrwERROR: Ansible version should be between 2.8 and 2.9. Current version is 9.34482 which is not supported.15:53
hrw;D15:53
osmanlicilegiah my lovely patch15:54
osmanlicilegiwhat's the real version?15:55
hrwon host it is 2.9.215:56
hrwon target there is no ansible15:56
hrw(venv3) 16:54 (1s) hrw@puchatek:kolla-ansible$ ANSIBLE_DEBUG=1 kolla-ansible $KAOPTS prechecks15:56
hrwException ignored in: <_io.TextIOWrapper name='<stdout>' mode='w' encoding='UTF-8'>15:56
hrwBrokenPipeError: [Errno 32] Broken pipe15:56
hrwERROR: Ansible version should be between 2.8 and 2.9. Current version is 7.07277 which is not supported.15:56
osmanlicilegiweird. let me check.15:57
hrwosmanlicilegi: https://paste.centos.org/view/d711e1b3 without ANSIBLE_DEBUG=115:57
openstackgerritMerged openstack/kayobe master: Update docs building  https://review.opendev.org/70493915:57
osmanlicilegihrw: venv is python 2 or 3?15:57
hrw315:58
hrwedited bashrc on target to not spit keychain stuff...15:58
hrwERROR: Ansible version should be between 2.8 and 2.9. Current version is 2.17443 which is not supported.15:58
osmanlicilegican you try without venv?15:59
hrwosmanlicilegi: can15:59
hrwosmanlicilegi: but no idea do I have requirements installed for k-a run15:59
osmanlicilegiseems venv causes something weird16:00
hrwosmanlicilegi: it works without ANSIBLE_DEBUG=116:00
hrwERROR: Ansible version should be between 2.8 and 2.9. Current version is 2.74531 which is not supported.16:00
hrwno venv16:00
hrw17:00 (0s) hrw@puchatek:kolla-ansible$ export KAOPTS=" -i hrw/all-in-one --configdir $PWD/hrw --passwords hrw/passwords.yml "16:01
hrw17:01 (0s) hrw@puchatek:kolla-ansible$ ANSIBLE_DEBUG=1 ./tools/kolla-ansible $KAOPTS prechecks16:01
hrw7.51044 7.19623 6.87138 6.55343 6.24087 5.91051 5.52886 5.06969 4.5836716:03
hrwquite random values16:03
hrwosmanlicilegi: to make it funnier: target is c816:04
osmanlicilegiawesome :)16:05
*** skramaja has quit IRC16:06
*** ivve has quit IRC16:10
hrwok. will have to create c7 vm to play16:13
hrwbut not today16:17
hrwI will be back on Tuesday - FOSDEM calls ;D16:17
hrwbye16:21
openstackgerritMark Goddard proposed openstack/kayobe master: WIP: Use python3 for local kolla-ansible execution  https://review.opendev.org/70500016:23
*** oyrogerg has quit IRC16:24
osmanlicilegihrw: enjoy fosdem16:26
*** admin0 has quit IRC16:28
*** bengates has quit IRC16:42
openstackgerritAntony Messerli proposed openstack/kolla master: Adds python3-systemd for ansible deploy interface  https://review.opendev.org/70502116:44
*** oyrogerg has joined #openstack-kolla17:02
yoctozeptohrw, osmanlicilegi: c8 broke bash?17:02
osmanlicilegiyoctozepto: seems so17:03
*** factor has quit IRC17:08
*** priteau has quit IRC17:21
*** gfidente is now known as gfidente|afk17:31
*** evrardjp has joined #openstack-kolla17:33
*** rpittau is now known as rpittau|afk18:02
*** igordc has joined #openstack-kolla18:09
*** luksky has quit IRC18:25
*** tonythomas has quit IRC18:27
*** k_mouza has joined #openstack-kolla18:34
*** k_mouza has quit IRC18:38
*** k_mouza_ has joined #openstack-kolla18:39
*** gfidente|afk is now known as gfidente18:43
*** ivve has joined #openstack-kolla18:47
*** luksky has joined #openstack-kolla19:14
*** gfidente has quit IRC19:35
openstackgerritMerged openstack/kolla master: CentOS 8: disable repo_gpgcheck for grafana image  https://review.opendev.org/70499919:41
*** hamzy_ has joined #openstack-kolla20:11
openstackgerritMerged openstack/kolla master: CentOS 8: Update packages in images  https://review.opendev.org/69941420:13
*** hamzy has quit IRC20:14
*** rlljorge has quit IRC20:39
*** kplant has quit IRC20:44
*** Trevor_V has quit IRC20:55
openstackgerritMichal Nasiadka proposed openstack/kolla-ansible master: WIP: Ceph-Ansible upgrade jobs  https://review.opendev.org/70498421:05
*** dking_desktop has joined #openstack-kolla21:05
openstackgerritMichal Nasiadka proposed openstack/kolla-ansible master: OVN Support  https://review.opendev.org/69684121:11
*** rlljorge has joined #openstack-kolla22:24
rlljorgeHi there, someone has a train release with ceph ?22:25
*** mixor has quit IRC22:49
*** ivve has quit IRC23:07
*** luksky has quit IRC23:21

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