| *** k_mouza has joined #openstack-kolla | 00:01 | |
| *** k_mouza has quit IRC | 00:05 | |
| *** igordc has quit IRC | 00:08 | |
| *** noxoid_ has joined #openstack-kolla | 00:14 | |
| *** dmellado_ has joined #openstack-kolla | 00:16 | |
| *** dcapone2004 has quit IRC | 00:17 | |
| *** noxoid has quit IRC | 00:21 | |
| *** sm806 has quit IRC | 00:21 | |
| *** mgkwill has quit IRC | 00:21 | |
| *** dmellado has quit IRC | 00:21 | |
| *** dmellado_ is now known as dmellado | 00:22 | |
| *** sm806 has joined #openstack-kolla | 00:22 | |
| *** chrizl has joined #openstack-kolla | 00:23 | |
| *** egonzalez has quit IRC | 00:25 | |
| *** cz3 has quit IRC | 00:25 | |
| *** mnasiadka has quit IRC | 00:25 | |
| *** mwhahaha has quit IRC | 00:25 | |
| *** r3ap3r has quit IRC | 00:25 | |
| *** ab-a has quit IRC | 00:25 | |
| *** logan- has quit IRC | 00:25 | |
| *** mnasiadka has joined #openstack-kolla | 00:25 | |
| *** cz3 has joined #openstack-kolla | 00:25 | |
| *** r3ap3r has joined #openstack-kolla | 00:25 | |
| *** mwhahaha has joined #openstack-kolla | 00:25 | |
| *** egonzalez has joined #openstack-kolla | 00:25 | |
| *** rm_work has quit IRC | 00:27 | |
| *** logan- has joined #openstack-kolla | 00:27 | |
| *** openstackstatus has quit IRC | 00:28 | |
| *** Jeffrey4l has quit IRC | 00:29 | |
| *** rm_work has joined #openstack-kolla | 00:30 | |
| *** ab-a has joined #openstack-kolla | 00:30 | |
| *** Jeffrey4l has joined #openstack-kolla | 00:31 | |
| *** skramaja has quit IRC | 01:11 | |
| *** dking_desktop has left #openstack-kolla | 01:34 | |
| *** zhurong has joined #openstack-kolla | 03:03 | |
| openstackgerrit | Xinliang Liu proposed openstack/kolla-ansible stable/rocky: Use become for Cinder LVM precheck https://review.opendev.org/706172 | 03:48 |
|---|---|---|
| *** k_mouza has joined #openstack-kolla | 04:00 | |
| *** k_mouza has quit IRC | 04:05 | |
| *** skramaja has joined #openstack-kolla | 04:20 | |
| *** zhurong has quit IRC | 04:41 | |
| *** zhurong has joined #openstack-kolla | 05:07 | |
| *** Tony31 has joined #openstack-kolla | 05:18 | |
| Tony31 | good morning | 05:18 |
| *** shyamb has joined #openstack-kolla | 05:31 | |
| *** evrardjp has quit IRC | 05:33 | |
| *** evrardjp has joined #openstack-kolla | 05:34 | |
| *** sorin-mihai has joined #openstack-kolla | 05:48 | |
| *** sean-k-mooney has quit IRC | 05:53 | |
| *** sean-k-mooney has joined #openstack-kolla | 05:55 | |
| *** dking_desktop has joined #openstack-kolla | 06:08 | |
| *** Tony3136 has joined #openstack-kolla | 06:15 | |
| Tony3136 | . | 06:16 |
| *** shyamb has quit IRC | 06:17 | |
| *** shyamb has joined #openstack-kolla | 06:17 | |
| *** Tony31 has quit IRC | 06:18 | |
| *** Tony3136 is now known as Tony31 | 06:18 | |
| *** shyamb has quit IRC | 06:25 | |
| *** shyamb has joined #openstack-kolla | 06:27 | |
| Tony31 | hi guys | 06:38 |
| Tony31 | having some trouble running `kayobe overcloud provision` . Seems to be getting stuck trying to connect to the baremetal VMs. | 06:41 |
| *** zhurong has quit IRC | 06:42 | |
| Tony31 | I was expecting the VMs to pxeboot and come up with the IPs listed in network-allocation.yml but instead it comes up with with an IP in the inspection range | 06:43 |
| Tony31 | any idea what I could be doing wrong here? | 06:43 |
| *** shyamb has quit IRC | 07:00 | |
| *** sean-k-mooney has quit IRC | 07:01 | |
| *** sean-k-mooney has joined #openstack-kolla | 07:03 | |
| osmanlicilegi | morning | 07:04 |
| Tony31 | hi osman | 07:05 |
| *** shyamb has joined #openstack-kolla | 07:18 | |
| *** hjensas has joined #openstack-kolla | 07:24 | |
| Tony31 | could anyone help with an error that I get when running `kayobe overcloud provision` ? | 07:39 |
| Tony31 | pretty please | 07:39 |
| *** shyamb has quit IRC | 07:40 | |
| yoctozepto | mgoddard, mnasiadka: kayobe user-to-be asking for your assistance | 07:42 |
| yoctozepto | Tony31: sorry, neither me nor osmanlicilegi are kayobe folks :< | 07:42 |
| Tony31 | Thanks for the response anyway :) :) | 07:42 |
| Tony31 | I must leave the office today before 5pm to fix my gf's internet but i am planning to come back online today from home so I can try and catch up with the `Kayobe` guys. I think I am really close now :) | 07:43 |
| yoctozepto | Tony31: those gfs always break their internets eh :D | 07:46 |
| Tony31 | :) actually, both hers and mine went down last night but hers never came back. We both use the Australia "NBN" with different ISPs. the NBN is a government network initiative to provide high-speed internet (because previously it was rubbish ADSL >6mb/s everywhere). Mine use fibre to my place but hers uses VDSL. The xDSL all use cheap equipment and | 07:49 |
| Tony31 | it's really bad. OFten these problems happen and they have to "rebuild the port" which I think means delete the config and put it back in again 🙈 | 07:49 |
| Tony31 | sorry I mean <6mb/s | 07:49 |
| Tony31 | her isp told her to configure the modem... it's been working for a month or two and I definitely configured it 🙈 | 07:50 |
| Tony31 | yoctozepto is there a `kolla` equivalent of `kayobe overcloud provision` ? | 07:51 |
| Tony31 | I was looking at this page to find if there was https://docs.openstack.org/kolla-ansible/latest/reference/bare-metal/ironic-guide.html but I didnt piece together anything there, but could be my lack of knowledge of it | 07:52 |
| yoctozepto | Tony31: hah, bad isps in there | 07:52 |
| Tony31 | I mean, do you know if there is a `kolla` equivalent of `kayobe overcloud provision` ? | 07:53 |
| yoctozepto | Tony31: kolla-ansible works on pre-provisioned servers | 07:53 |
| mnasiadka | morning | 07:53 |
| Tony31 | ah | 07:53 |
| yoctozepto | so you have to install the OS yourself | 07:53 |
| Tony31 | good morning | 07:53 |
| yoctozepto | setup ssh access | 07:53 |
| Tony31 | that could be an alternative for me | 07:53 |
| yoctozepto | configure networking per docs | 07:53 |
| yoctozepto | and then can deploy | 07:53 |
| Tony31 | because at the moment I am using VMs to try this out and I am pxebooting them but there is no IPMI or BMC. So I'm running the command and then powering up the VMs manually and crossing fingers. It did work ok for tripleO deployment a few years ago | 07:54 |
| yoctozepto | well, personally I had to abandon kayobe for the time being due to my fujitsu servers not liking ironic access; I still have ironic in plans but well, time is time :-) | 07:54 |
| yoctozepto | oh man, what an overkill | 07:55 |
| Tony31 | i had so many issues with the baremetal part of tripleO, the manual power up was the easy bit :) Then they updated the RPM between the working week and the weekend and so I hit a few bugs which almost caused an abort of the deployment | 07:55 |
| Tony31 | mnasiadka hi - do you have any capacity to try and help me with `kayobe overcloud provision ` please? | 07:56 |
| mnasiadka | Tony31: first tell me what is the problem, then I can tell you if I can help :) | 07:57 |
| openstackgerrit | Piotr Rabiega proposed openstack/kolla master: Add collectd-intel_pmu plugin https://review.opendev.org/704560 | 07:57 |
| Tony31 | OK :) | 07:57 |
| Tony31 | when I run that command, it fails at the `TASK [Ensure the ironic node is available]` | 07:58 |
| Tony31 | I have 2 VMs added as baremetal | 07:58 |
| Tony31 | introspection was performed and I named the nodes as controller0 and compute0 | 07:58 |
| Tony31 | when I power up the VMs they PXE boot and get an IP in the introspection range. I was expecting them to get an ip within the allocation pool | 07:59 |
| Tony31 | this could be pointing to the problem but I havent been able to figure it out yet | 07:59 |
| yoctozepto | and to answer "08:51:51 <Tony31> yoctozepto is there a `kolla` equivalent of `kayobe overcloud provision` ?" -> in kolla-ansible it's just deploy :-) | 08:00 |
| *** dciabrin has joined #openstack-kolla | 08:00 | |
| yoctozepto | and now I go silent and you try to debug kayobe, good luck | 08:00 |
| Tony31 | `# openstack baremetal node list | 08:00 |
| Tony31 | +--------------------------------------+-------------+---------------+-------------+--------------------+-------------+ | 08:00 |
| Tony31 | False | | 08:00 |
| mnasiadka | Tony31: in the sense of having 2 VMs added as baremetal - do you use universe from nothing approach? | 08:03 |
| Tony31 | yes I did do that | 08:03 |
| Tony31 | but this time I am designing a production deployment | 08:03 |
| mnasiadka | and Ironic first needs to introspect a node, before it can be set to available state and provisioned | 08:03 |
| Tony31 | on control host VM and seed VM | 08:03 |
| Tony31 | I can try the introspection again | 08:04 |
| mnasiadka | the question is if the introspection fails (and why)? | 08:04 |
| Tony31 | just starting it now :) thanks for your help | 08:05 |
| *** dciabrin has quit IRC | 08:09 | |
| *** cah_link has joined #openstack-kolla | 08:10 | |
| Tony31 | I get this error `dev/null 2>&1 && sleep 0'", "controller0 | FAILED | rc=1 >>", "The requested action \"inspect\" can not be performed on node \"5fb8a19d-60a9-4359-8d8b-a9ecc5cb22bc\" while it is in state \"enroll\". (HTTP 400)non-zero return code"]}` | 08:11 |
| Tony31 | the provisioning state is currently `enroll` | 08:11 |
| Tony31 | do you know what state it needs to be in for `introspection`? | 08:12 |
| openstackgerrit | Eddie Yen proposed openstack/kolla-ansible stable/rocky: Remove /etc/hosts entries pointing hostname to localhost and prevent cloud-init to manage /etc/hosts https://review.opendev.org/706192 | 08:13 |
| *** dciabrin has joined #openstack-kolla | 08:13 | |
| *** k_mouza has joined #openstack-kolla | 08:15 | |
| Tony31 | should the provisioning state be `manageable` ? | 08:16 |
| Tony31 | I could try manually resetting and re-trying | 08:16 |
| *** k_mouza has quit IRC | 08:19 | |
| Tony31 | do you know which process should be making the nodes moved from `enroll` to `manage` ? because I could try running that again | 08:21 |
| *** tonythomas has joined #openstack-kolla | 08:24 | |
| Tony31 | from `openstack baremetal node show` I see `last_error | Failed to validate power driver interface for node 6d45927a-a8e8-46c7-8f31-b0ccb1992400. Error: Missing the following IPMI credentials in node's driver_info: ['ipmi_address']. | | 08:28 |
| Tony31 | could this be a reason? Can I use a fakePXE interface for the VMs ? | 08:29 |
| yoctozepto | Tony31: well, you most likely *have to* :-) | 08:36 |
| yoctozepto | but not sure about fakepxe | 08:36 |
| Tony31 | I'm trying :) | 08:36 |
| yoctozepto | I think it's called differently for power | 08:36 |
| Tony31 | i set a dummy pxe username/password for ipmi and now the error is that there is no ip address | 08:39 |
| Tony31 | `last_error | Failed to validate power driver interface for node 6d45927a-a8e8-46c7-8f31-b0ccb1992400. Error: Missing the following IPMI credentials in node's driver_info: ['ipmi_address']. | | 08:39 |
| yoctozepto | Tony31: yea, cause vms have no bmc | 08:39 |
| yoctozepto | and you did not configure it | 08:39 |
| Tony31 | cant find this variable at all | 08:40 |
| yoctozepto | it won't help to fake it | 08:40 |
| yoctozepto | need to change the driver for power | 08:40 |
| Tony31 | cant find that either, or a list of drivers | 08:40 |
| yoctozepto | ah, ok | 08:40 |
| Tony31 | going to have to go in a sec but i'll be back on later | 08:40 |
| yoctozepto | mgoddard, mnasiadka: so we pinpointed that to Tony31 using vms | 08:41 |
| yoctozepto | now the question is how to handle this in kayobe | 08:41 |
| Tony31 | I'd like to use VMs for controller and at least one compute... but then later on, bring in my 2 x Cisco servers in a chassis which can use ipmi no problem | 08:41 |
| Tony31 | catch you later thanks for your help :) | 08:41 |
| mnasiadka | Tony31: just thinking - do you really follow the guide on universe-from-nothing and use tenks? | 08:45 |
| *** Tony31 has quit IRC | 08:46 | |
| *** carli has joined #openstack-kolla | 08:49 | |
| mnasiadka | yoctozepto: do you have any idea why the upgrade jobs fail here? https://review.opendev.org/#/c/704309/ | 08:54 |
| patchbot | patch 704309 - kolla-ansible - WIP: Remove kolla-ceph - 6 patch sets | 08:54 |
| mnasiadka | seems like it uses inventory from later version, but code from earlier :) | 08:54 |
| *** jaicaa has quit IRC | 08:55 | |
| mnasiadka | and probably it does, since we template out the inventory from tests/templates | 08:55 |
| yoctozepto | mnasiadka: yeah, exactly | 08:56 |
| yoctozepto | inventory has to be compatible with both versions | 08:56 |
| mnasiadka | oh boy, I'll need to add another if in the inventory template | 08:56 |
| yoctozepto | mnasiadka: well, it would be easier if ansible assumed that to be empty and only generated warning | 08:57 |
| *** jaicaa has joined #openstack-kolla | 08:58 | |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: WIP: Remove kolla-ceph https://review.opendev.org/704309 | 09:01 |
| *** k_mouza has joined #openstack-kolla | 09:04 | |
| mnasiadka | yoctozepto: life can't be easy :) | 09:07 |
| mgoddard | morning | 09:10 |
| *** dougsz has joined #openstack-kolla | 09:11 | |
| *** dciabrin has quit IRC | 09:18 | |
| *** dciabrin has joined #openstack-kolla | 09:18 | |
| *** shyamb has joined #openstack-kolla | 09:22 | |
| *** brtknr has quit IRC | 09:40 | |
| openstackgerrit | Mark Goddard proposed openstack/kayobe master: WIP: Use python3 for local kolla-ansible execution https://review.opendev.org/705000 | 09:45 |
| openstackgerrit | Noboru Iwamatsu proposed openstack/kolla-ansible master: Adapt to Octavia Certificate Configuration Guide. https://review.opendev.org/706203 | 09:51 |
| *** ivve has joined #openstack-kolla | 09:58 | |
| *** ktibi has joined #openstack-kolla | 09:59 | |
| *** dciabrin has quit IRC | 10:02 | |
| ktibi | Hi guys, I have a strange behavior on nova. I don't know why, but nova create vms on the same host. When the host is full, nova change. By example, I created 20vms and all VM are on the same host. but I have 4 other hosts with 2 instances. | 10:02 |
| ktibi | I have this behavior only on plateform deploy by kolla :/ | 10:03 |
| ktibi | any suggestions ? | 10:03 |
| *** bjolo has joined #openstack-kolla | 10:05 | |
| *** gfidente|afk is now known as gfidente | 10:06 | |
| openstackgerrit | Merged openstack/kolla-ansible stable/rocky: Use become for Cinder LVM precheck https://review.opendev.org/706172 | 10:08 |
| *** dciabrin has joined #openstack-kolla | 10:11 | |
| yoctozepto | ktibi: odd, which release? | 10:22 |
| yoctozepto | maybe it had some bug in placement/nova which causes this | 10:23 |
| ktibi | queens | 10:23 |
| ktibi | rocky* | 10:23 |
| *** shyamb has quit IRC | 10:32 | |
| *** rpittau|afk is now known as rpittau | 10:45 | |
| *** ricolin_ has joined #openstack-kolla | 10:47 | |
| openstackgerrit | Radosław Piliszek proposed openstack/kolla master: CI: Replace cinder-lvm scenario with zun scenario https://review.opendev.org/706059 | 10:49 |
| yoctozepto | mgoddard, mnasiadka: https://review.opendev.org/706059 let's merge before it gets another conflict... | 10:50 |
| patchbot | patch 706059 - kolla - CI: Replace cinder-lvm scenario with zun scenario - 2 patch sets | 10:50 |
| openstackgerrit | Mark Goddard proposed openstack/kolla master: CentOS 8: Enable hacluster images https://review.opendev.org/702706 | 10:51 |
| yoctozepto | ktibi: hmm, /me running stein; would you mind posting about this to ml? might be it was a known problem | 10:51 |
| *** dciabrin has quit IRC | 10:53 | |
| *** dciabrin has joined #openstack-kolla | 10:54 | |
| yoctozepto | mgoddard: sorry for bother but it would be best if you rebased c8 jobs on top of respective zun's lvm patch | 10:59 |
| *** shyamb has joined #openstack-kolla | 10:59 | |
| mgoddard | yoctozepto: depends if we are backporting zun/lvm to train | 11:00 |
| mgoddard | if not, will be cleaner for me to merge first | 11:00 |
| yoctozepto | mgoddard: I planned to since it's the very same story | 11:00 |
| mgoddard | yoctozepto: k | 11:00 |
| yoctozepto | mgoddard: needless cycles better spared | 11:00 |
| mgoddard | yoctozepto: yours or mine? :p | 11:01 |
| yoctozepto | mgoddard: ci's | 11:01 |
| yoctozepto | mgoddard: I believe ours are very necessary :D | 11:01 |
| mgoddard | ah that's annoying | 11:02 |
| mgoddard | If I build bifrost but it is marked unbuildable, it will build base and openstack-base before telling me it couldn't build bifrost-base | 11:03 |
| yoctozepto | lolz | 11:03 |
| yoctozepto | yeah, kolla awaiting rewrite | 11:03 |
| mnasiadka | it should be pretty easy to just throw an error when the supplied images are unbuildable | 11:04 |
| mgoddard | naturally, it's fiddly | 11:04 |
| mnasiadka | ok, after forging the inventory template - the upgrade jobs work | 11:05 |
| mgoddard | we want to apply filters to parents | 11:05 |
| openstackgerrit | Merged openstack/kolla-ansible stable/rocky: Remove /etc/hosts entries pointing hostname to localhost and prevent cloud-init to manage /etc/hosts https://review.opendev.org/706192 | 11:05 |
| *** Tony31 has joined #openstack-kolla | 11:13 | |
| Tony31 | hi again | 11:13 |
| Tony31 | mnasiadka I did follow the guide but in the end, I didnt need to make any variable changes. Initially I was trying to configure the networking because I understood incorrectly that it would be bridged to the LAN. But instead, if I didnt make any changes everything just worked except for some problems with an ansible module in the venv and missing | 11:15 |
| Tony31 | packages | 11:15 |
| Tony31 | but as I wont use tenks in prod, is it relevant? | 11:15 |
| openstackgerrit | Michal Nasiadka proposed openstack/kolla-ansible master: Remove kolla-ceph https://review.opendev.org/704309 | 11:15 |
| Tony31 | since you mentioned tenks - i'll go and take another look at it | 11:15 |
| yoctozepto | mgoddard: kayobe user-to-be ^ | 11:16 |
| Tony31 | I still have the vm that has the all in one. Openstack runs pretty quick on it tbh | 11:16 |
| mgoddard | hi Tony31 | 11:16 |
| mgoddard | tenks is just used for testing | 11:17 |
| mgoddard | we use it in CI to test ironic | 11:17 |
| mgoddard | and also in a universe from nothing to create baremetal VMs for controllers and computes | 11:17 |
| Tony31 | Hi Mark | 11:18 |
| mgoddard | is there something we can put in the README to make it clearer that there's no bridge required? | 11:18 |
| Tony31 | maybe I can use it as a virtual BMC but I dont want to run VMs inside VMs for the controller / compute | 11:18 |
| yoctozepto | mgoddard: Tony31 has a problem with kayobe bifrosty part | 11:18 |
| mgoddard | +1, it's good for testing but not production | 11:18 |
| yoctozepto | where it requires some power interface to vms | 11:18 |
| Tony31 | yes - I was going to ask if I could make a pull request and suggest some udpates to it? | 11:18 |
| mgoddard | of course | 11:18 |
| yoctozepto | Tony31: absolutely | 11:18 |
| yoctozepto | we love good doc updates :-) | 11:19 |
| Tony31 | my background is networking so I'm probably looking at it from a bad angle | 11:19 |
| mgoddard | that would be much appreciated | 11:19 |
| mgoddard | where were the missing packages btw? | 11:19 |
| Tony31 | ok - will do it after getting this up and running I still want to find out how to do a bunch of things like add a 2nd storage network to the compute/controller, add cinder driver, domain auth., horizon updates etc | 11:20 |
| Tony31 | missing packages I got by installing `yum install qemu-kvm libvirt libvirt-python libguestfs-tools virt-install qemu-img virt-manager libvirt-client virt-viewer bridge-utils` so I am not sure which package(s) fixed the issue. I also needed to reboot afterwards | 11:22 |
| Tony31 | then I found that there's an openstack module that is getting deprecated and I needed to source the kolla venv and run `pip install os-client-config` | 11:23 |
| Tony31 | sorry | 11:23 |
| Tony31 | not kolla | 11:23 |
| Tony31 | tenks venv | 11:23 |
| Tony31 | is there a way I can tell kayobe to ignore the ipmi stuff? I'll power on the VM myself... I know it goes against automation but I'm not sure how else it would work, with control and compute in a VM on oVirt | 11:25 |
| mgoddard | strange - we install os-client-config in the tenks venv | 11:25 |
| Tony31 | cue twilight zone music :) | 11:26 |
| mgoddard | Tony31: so let me see if I understand. You are trying to set up a production environment, and want to create some VMs for your controllers and use those | 11:26 |
| Tony31 | well this is how I have it running with tripleo pike at the moment: single controller/ network VM. 2 x physical compute nodes which are Cisco hardware. Plus an undercloud director thats shut down now. | 11:27 |
| mgoddard | ok | 11:27 |
| mgoddard | do you have enough VM automation to create VMs with a CentOS image and an IP? | 11:27 |
| yoctozepto | (not that one needs too much automation for a single controller) | 11:28 |
| Tony31 | I'm thinking to deploy another controller VM but also a compute VM so that i can manage a DR failover (I'll failover the hypervisor and VMs) because of the absence of knowledge on how to achieve the same with openstack. I have 2 x physical compute nodes in the DR and I could use one for oVirt and the other for openstack. But at the moment both are | 11:28 |
| Tony31 | in ovirt | 11:28 |
| mgoddard | if so I would suggest skipping the bifrost part for the controller VM - no need to pretend it's baremetal | 11:28 |
| yoctozepto | ^ +1 | 11:29 |
| Tony31 | ... I'd like to make the controller HA but I havent gone down that road yet - I Wanted to ask you guys could I achieve it later on? I think I can, just have not searched for the docs | 11:29 |
| mgoddard | just create a CentOS VM and add it to your inventory | 11:29 |
| mgoddard | yes, kolla supports HA | 11:29 |
| Tony31 | ok thanks Mark | 11:29 |
| yoctozepto | Tony31: it's easy then, kolla-ansible does HA for you | 11:29 |
| Tony31 | That was on my mind today | 11:29 |
| yoctozepto | just slap more controllers :D | 11:29 |
| mgoddard | etc/kayobe/inventory/hosts: | 11:29 |
| Tony31 | :) | 11:29 |
| yoctozepto | into the inventory | 11:29 |
| yoctozepto | mgoddard: you have commands in kayobe that proxy to k-a deploy, no? | 11:30 |
| Tony31 | How do I get rid of the 2 fake baremetal things in the seed / bifrost container? | 11:30 |
| mnasiadka | yoctozepto: your zun change just failed on ubuntu/debian :D | 11:30 |
| mgoddard | similar to this: https://github.com/openstack/kayobe-config-dev/blob/master/etc/kayobe/inventory/hosts | 11:30 |
| mgoddard | add a host to the controllers group | 11:30 |
| mgoddard | define it's IP in etc/kayobe/network-allocation.yml | 11:30 |
| yoctozepto | mnasiadka: on debian, on grafana | 11:30 |
| Tony31 | yes got it :) thanks - will treat it like the seed | 11:30 |
| yoctozepto | mnasiadka: kolla ci goes red | 11:30 |
| yoctozepto | again | 11:31 |
| mnasiadka | uhh | 11:31 |
| Tony31 | my seed is a VM I installed myself on the ovirt host using cloud-init | 11:31 |
| mgoddard | openstack baremetal node delete <node> | 11:31 |
| mnasiadka | yoctozepto: what's wrong with grafana on debian? :) | 11:31 |
| mgoddard | to remove from bifrost | 11:31 |
| Tony31 | ah great - so just typical openstack commands then :) | 11:31 |
| yoctozepto | mnasiadka: everything | 11:31 |
| Tony31 | thanks a lot :) | 11:31 |
| mgoddard | yep | 11:31 |
| mgoddard | you need to run them from inside the bifrost container though | 11:31 |
| yoctozepto | mnasiadka: INFO:kolla.common.utils.grafana:E: Failed to fetch https://packages.grafana.com/oss/deb/dists/stable/main/binary-amd64/Packages.bz2 File has unexpected size (14821 != 14971). Mirror sync in progress? [IP: 151.101.198.217 443] | 11:31 |
| Tony31 | I guess I Went down the wrong path because I was trying to apply lessons learned from tripleo | 11:31 |
| mgoddard | but it sounds like you've done that already | 11:31 |
| Tony31 | yep :) was looking around there today | 11:32 |
| mgoddard | we're a lot more flexible than tripleo :) | 11:32 |
| mgoddard | bring your own VMs - not a problem | 11:32 |
| Tony31 | when I deployed tripleo it kept placing the controller on the compute node. it was sheer luck I got it deployed how I Wanted it. | 11:32 |
| mgoddard | lol | 11:32 |
| yoctozepto | mgoddard: and we have much less code to break :-) | 11:32 |
| mgoddard | it's a bit of a jelly | 11:32 |
| Tony31 | i tagged and named and everything, but it just picked a random host for the placement | 11:32 |
| Tony31 | :) kayobe is much better | 11:33 |
| yoctozepto | :D | 11:33 |
| mgoddard | woop | 11:33 |
| yoctozepto | wait till you see bikolla | 11:33 |
| mgoddard | I would suggest going bare metal for computes in production though | 11:33 |
| yoctozepto | ^ +1 | 11:33 |
| mgoddard | depending on what you need for DR | 11:33 |
| Tony31 | another question i had - but it's probably in the docs so i can find it another time. if i make changes to the deployment and want to execute those, is there a process guide to follow? | 11:34 |
| Tony31 | i have 2 x bare metal at the moment.. the issue I am trying to avoid is the different networks in DR compared to prod | 11:34 |
| mgoddard | kayobe overcloud service reconfigure | 11:34 |
| mgoddard | that will regenerate config and containers if necessary | 11:34 |
| yoctozepto | mnasiadka: thankfully, debian is non-voting | 11:34 |
| Tony31 | the physical machine in DR will have VLANs in different subnets. Additionally the storage is replicated there but the IP's and IQNs are different | 11:34 |
| mgoddard | https://docs.openstack.org/kayobe/latest/administration/overcloud.html#reconfiguring-containerised-services | 11:35 |
| Tony31 | thanks mark - saving :) | 11:35 |
| Tony31 | also when i was configuring the networks, it asks for gateways and vlans - well in most cases I dont need to tag vlans because the baremetal chassis takes care of that, and so does the hypervisor (ovirt). I did have an issue with the seed because it had 2 x default gateways and initially it broke networking | 11:36 |
| Tony31 | how do you take care of that? | 11:36 |
| Tony31 | I say "asks" what I mean is, in the yml file it has commented out fields... I just filled them all out | 11:37 |
| Tony31 | :) | 11:37 |
| *** rlljorge has joined #openstack-kolla | 11:38 | |
| yoctozepto | Tony31: these are to properly configure baremetals | 11:40 |
| yoctozepto | when it *would* configure interfaces witg tags | 11:40 |
| Tony31 | ah ok :) | 11:40 |
| Tony31 | I planned to try deploying and see if it worked and go from there | 11:40 |
| openstackgerrit | Mark Goddard proposed openstack/kolla master: CentOS 8: Add deploy jobs in CI https://review.opendev.org/705952 | 11:40 |
| Tony31 | thanks for the info | 11:40 |
| yoctozepto | yw | 11:41 |
| mgoddard | Tony31: you don't need all the fields for every network. In general there should only be one network with a gateway | 11:41 |
| Tony31 | yoctozepto btw I managed to fix that internet in under 60 seconds. no joke. All modem lights were green. So I traced to `1.1.1.1` and it worked. nslookup was going to an `fe80:` ipv6 address.. so i turned off ipv6 and everything worked. . I think the outage last night was them deploying ipv6. I've yet to check though | 11:42 |
| Tony31 | Mark, again I think I am using past tripleo experience and causing a problem for myself... I need a default gateway for the nodes.. for all of them (except the controller) it's the management network. | 11:42 |
| yoctozepto | Tony31: oh, you are one of those baddies who disable ipv6 | 11:43 |
| Tony31 | the controller, however, has a default gateway out via the external API network so the horizone dashboard can be loaded. I'll need to check again | 11:43 |
| Tony31 | yoctozepto only when it's not set up or not set up properly :) | 11:43 |
| Tony31 | im an ipv6 forum gold member | 11:44 |
| openstackgerrit | Mark Goddard proposed openstack/kolla-ansible master: CentOS 8: Add deploy jobs in CI https://review.opendev.org/693544 | 11:44 |
| Tony31 | lol | 11:44 |
| rlljorge | Morning ! I have a bug with train release and I try deploy stein. Stein is compatible with ansible 2.9.2 and kolla-ansible 9.1.0.dev160 ? | 11:44 |
| yoctozepto | Tony31: nice! | 11:45 |
| mgoddard | Tony31: there are two approaches. You could use ansible group_vars to set different gateways for different groups (needs to be removed from networks.yml which has precedence) | 11:45 |
| yoctozepto | rlljorge: that kolla ansible is for master/ussuri | 11:45 |
| Tony31 | ok thank you - makes perfect sense. | 11:45 |
| mgoddard | or you can use policy-based routing on the external network to avoid asymmetric routing | 11:45 |
| yoctozepto | rlljorge: stein is 8.* | 11:46 |
| Tony31 | i like your first option - i had not thought of it, but now you said it, it does make sense | 11:46 |
| mgoddard | https://docs.openstack.org/kayobe/latest/configuration/network.html#advanced-policy-based-routing | 11:46 |
| rlljorge | yoctozepto thanks | 11:46 |
| Tony31 | so just hypothetically thinking... I could deploy openstack using kayobe, then manually create a VM on openstack. Then deploy/reconfigure openstack with kayobe and make that VM a controller node :D | 11:48 |
| yoctozepto | Tony31: yeah, though that would be trying to go worse than tripleo :-) | 11:49 |
| yoctozepto | (in the sense of complicating life) | 11:49 |
| Tony31 | :) | 11:50 |
| *** brtknr has joined #openstack-kolla | 11:54 | |
| Tony31 | thanks again guys - im off to get some dinner. have a good one :) | 11:59 |
| *** shyamb has quit IRC | 12:11 | |
| *** pcaruana has quit IRC | 12:13 | |
| *** Tony31 has quit IRC | 12:15 | |
| yoctozepto | mgoddard: it really does not seem like I need that dance for mariadb | 12:19 |
| yoctozepto | mgoddard: mind you it still runs recovery when necessary | 12:20 |
| yoctozepto | (xtra->maria) | 12:20 |
| *** kplant has joined #openstack-kolla | 12:23 | |
| rlljorge | fatal: [compute02]: FAILED! => {"changed": false, "msg": "Unable to start service docker: Job for docker.service failed because the control process exited with error code. See \"systemctl status docker.service\" and \"journalctl -xe\" for details.\n"} | 12:34 |
| hrw | https://marcin.juszkiewicz.com.pl/fosdem/ - covers all years with videos (2015-2020) | 12:34 |
| rlljorge | when I try execute bootstrap-servers | 12:34 |
| *** shyamb has joined #openstack-kolla | 12:34 | |
| yoctozepto | rlljorge: need to see that journalctl | 12:34 |
| rlljorge | yoctozepto http://paste.openstack.org/show/789220/ | 12:36 |
| yoctozepto | rlljorge: would be nice to see above Feb 06 09:35:36 control01.sianet.com.br systemd[1]: docker.service failed. | 12:37 |
| rlljorge | yoctozepto: http://paste.openstack.org/show/789221/ | 12:37 |
| rlljorge | I Think the problem is here | 12:37 |
| yoctozepto | rlljorge: indeed - did you previously bootstrap it with the latest kolla-ansible? | 12:40 |
| rlljorge | yoctozepto is | 12:40 |
| yoctozepto | looks like bootstrapping it again with an older one will kill it (which is not that unexpected) | 12:40 |
| yoctozepto | for your case just delete the json file | 12:41 |
| yoctozepto | on all bootstrapped host | 12:41 |
| yoctozepto | then rerun the playbook | 12:41 |
| rlljorge | yoctozepto perfect ! Works | 12:41 |
| yoctozepto | :-) | 12:41 |
| *** cah_link has quit IRC | 12:42 | |
| rlljorge | yoctozepto: tkank you | 12:42 |
| yoctozepto | rlljorge: you're welcome | 12:43 |
| *** shyamb has quit IRC | 12:45 | |
| *** rpittau is now known as rpittau|bbl | 12:52 | |
| *** jopdorp has joined #openstack-kolla | 12:57 | |
| jopdorp | hi! | 12:57 |
| yoctozepto | hi jopdorp | 12:57 |
| jopdorp | any hints about how to add my custom theme to my kolla ansible deployment? | 12:57 |
| jopdorp | I'm currently working with both an all-in-one deployment and a multinode | 12:57 |
| yoctozepto | jopdorp: customize the horizon image, build it, push it to registry and configure k-a to use it | 12:58 |
| jopdorp | which registry? | 12:58 |
| yoctozepto | jopdorp: docker registry, probably one you want to deploy internally | 12:58 |
| jopdorp | right | 12:59 |
| jopdorp | okay, thanks! | 12:59 |
| yoctozepto | minimal: https://docs.docker.com/registry/ | 12:59 |
| yoctozepto | but I use harbor | 12:59 |
| yoctozepto | https://goharbor.io/ | 12:59 |
| jopdorp | thanks! | 13:00 |
| kplant | ^ as do i | 13:00 |
| *** dciabrin has quit IRC | 13:00 | |
| *** dciabrin has joined #openstack-kolla | 13:01 | |
| kplant | yoctozepto: have you found a better way to delete a tag (and all images associated with that tag) than having to write code to do so? | 13:01 |
| kplant | it seems docker specifically left that ability out of non-docker enterprise >_< | 13:01 |
| *** ricolin_ has quit IRC | 13:02 | |
| jopdorp | will this fix be merged to train? | 13:03 |
| jopdorp | https://review.opendev.org/#/c/705949/2 | 13:03 |
| patchbot | patch 705949 - kolla-ansible (stable/train) - Fix keystone fernet bootstrap - 2 patch sets | 13:03 |
| yoctozepto | kplant: you mean like prune? | 13:03 |
| yoctozepto | jopdorp: yeah, it needs a little more fixing on master still | 13:03 |
| kplant | yoctozepto: in the registry, not on the local docker hosts | 13:03 |
| jopdorp | okay, I think I encountered a bug that will be fixed by it | 13:04 |
| yoctozepto | kplant: ah, you mean the basic docker registry? I thought it was painful to manage so moved to harbor | 13:04 |
| yoctozepto | kplant: it's doable from harbor ui afair | 13:04 |
| kplant | yeah, but i have the same probably with harbor | 13:05 |
| kplant | i see no way to mass delete images :-| | 13:05 |
| yoctozepto | jopdorp: care to share? | 13:05 |
| yoctozepto | kplant: ah, mass... | 13:05 |
| kplant | yeah, like delete all images with tag "stein-dev" | 13:05 |
| hrw | oraclelinux fails in stein... | 13:06 |
| hrw | do we care? | 13:06 |
| yoctozepto | hrw: badly? | 13:07 |
| yoctozepto | hrw: ah, tripleo | 13:08 |
| yoctozepto | hrw: easy to fix, disable building | 13:08 |
| yoctozepto | hrw: only tripleo wants it | 13:08 |
| jopdorp | i can't find the exact error i was facing right now, sorry! | 13:09 |
| yoctozepto | jopdorp: no problem, the fix is needed anyways | 13:10 |
| yoctozepto | the good thing is code/logic quality grows with every cycle :-) | 13:10 |
| rlljorge | me again LOL ! Another erro | 13:12 |
| rlljorge | fatal: [control03]: FAILED! => {"changed": true, "msg": "'Traceback (most recent call last):\\n File \"/tmp/ansible_ymVEod/ansible_module_kolla_docker.py\", line 983, in main\\n result = bool(getattr(dw, module.params.get(\\'action\\'))())\\n File \"/tmp/ansible_ymVEod/ansible_module_kolla_docker.py\", line 731, in | 13:12 |
| rlljorge | recreate_or_restart_container\\n self.start_container()\\n File \"/tmp/ansible_ymVEod/ansible_module_kolla_docker.py\", line 744, in start_container\\n self.pull_image()\\n File \"/tmp/ansible_ymVEod/ansible_module_kolla_docker.py\", line 556, in pull_image\\n repository=image, tag=tag, stream=True\\n File | 13:12 |
| rlljorge | \"/usr/lib/python2.7/site-packages/docker/api/image.py\", line 415, in pull\\n self._raise_for_status(response)\\n File \"/usr/lib/python2.7/site-packages/docker/api/client.py\", line 263, in _raise_for_status\\n raise create_api_error_from_http_exception(e)\\n File \"/usr/lib/python2.7/site-packages/docker/errors.py\", line 31, in | 13:12 |
| rlljorge | create_api_error_from_http_exception\\n raise cls(e, response=response, explanation=explanation)\\nNotFound: 404 Client Error: Not Found (\"manifest for kolla/centos-binary-fluentd:8.0.0 not found: manifest unknown: manifest unknown\")\\n'"} | 13:12 |
| jopdorp | yoctozepto: do you have any docs about how to use a docker image from my own registry? | 13:14 |
| rlljorge | jopdorp from docker | 13:14 |
| rlljorge | ops | 13:14 |
| rlljorge | my fault sorry | 13:14 |
| jopdorp | setting up my own registry won't be a problem | 13:16 |
| jopdorp | but how to I tell kolla to use it for the horizon image? | 13:16 |
| kplant | what you should do is build them all into your local registry | 13:17 |
| kplant | https://docs.openstack.org/kolla/latest/admin/image-building.html | 13:17 |
| kplant | you could technically only put horizon in there, change the registry and pull/deploy with "--tags horizon" | 13:17 |
| kplant | but i don't think that's a great idea | 13:17 |
| kplant | unless you're only testnig | 13:17 |
| kplant | yoctozepto: i think harbor has a custom api call, you can throw a DELETE at /repositories/{repo_name}/tags/{tag} | 13:18 |
| *** kklimonda has joined #openstack-kolla | 13:18 | |
| kplant | why didn't they put that in the UI >:( | 13:18 |
| *** skramaja has quit IRC | 13:18 | |
| *** gmann has quit IRC | 13:23 | |
| *** vkmc has quit IRC | 13:24 | |
| jopdorp | kplant: so how would I tell kolla to use my own registry? | 13:26 |
| *** irclogbot_0 has quit IRC | 13:26 | |
| jopdorp | never mind, found it at: https://docs.openstack.org/kolla/latest/admin/image-building.html | 13:26 |
| openstackgerrit | Mark Goddard proposed openstack/kolla-ansible master: CI: Use auto-detected python interpreter except on CentOS 7 https://review.opendev.org/706265 | 13:27 |
| *** cz3 has quit IRC | 13:28 | |
| *** irclogbot_0 has joined #openstack-kolla | 13:29 | |
| *** generalfuzz has quit IRC | 13:30 | |
| *** rpittau|bbl is now known as rpittau | 13:30 | |
| hrw | yoctozepto: I look at https://review.opendev.org/#/c/634019 and have that feeling that fluentd Dockerfile needs to be rewritten from scratch | 13:32 |
| patchbot | patch 634019 - kolla - Do not gem install for binary type - 6 patch sets | 13:32 |
| hrw | gave -1 | 13:34 |
| *** dciabrin has quit IRC | 13:55 | |
| *** jopdorp has quit IRC | 13:55 | |
| *** dciabrin has joined #openstack-kolla | 13:57 | |
| *** dciabrin has quit IRC | 13:57 | |
| *** cz3 has joined #openstack-kolla | 14:00 | |
| rlljorge | Hi there ! I have 2 noob questions ;-) 1) Can I rollback/downgrade from train to stein for example ? 2) Can I change Horizon logo using kolla ? | 14:01 |
| *** generalfuzz has joined #openstack-kolla | 14:02 | |
| hrw | rlljorge: 2. create own theme and add it into image via template overrides | 14:03 |
| *** kplant has quit IRC | 14:04 | |
| rlljorge | hrw Do you know if exists any documentation or example by this ? | 14:04 |
| *** vkmc has joined #openstack-kolla | 14:06 | |
| *** kplant has joined #openstack-kolla | 14:06 | |
| hrw | rlljorge: moment | 14:07 |
| hrw | rlljorge: https://git.linaro.org/leg/sdi/kolla/ldc-overlay.git/commit/?id=2c4abf12eb37a9636e315cfc0dc3b7221d17f342 | 14:08 |
| hrw | rlljorge: this override adds Linaro's theme to horizon | 14:08 |
| rlljorge | hrw Great ! I will try .... thank you ! | 14:09 |
| yoctozepto | hrw: +1 to your -1 | 14:10 |
| yoctozepto | rlljorge: downgrades are not possible | 14:11 |
| hrw | rlljorge: if you want your theme to be default then the easiest way is to have just one theme installed ;D | 14:11 |
| yoctozepto | this is irrespective of deployment method | 14:11 |
| openstackgerrit | Merged openstack/kolla-ansible stable/rocky: Fix qemu loading of ceph.conf (permission error) https://review.opendev.org/706073 | 14:37 |
| openstackgerrit | Merged openstack/kolla-ansible stable/stein: Fix qemu loading of ceph.conf (permission error) https://review.opendev.org/706072 | 14:37 |
| openstackgerrit | Merged openstack/kolla stable/stein: nova-libvirt: add UEFI packages to support UEFI instances https://review.opendev.org/706061 | 14:41 |
| rlljorge | The bug https://bugs.launchpad.net/cinder/+bug/1860739 not happens in stein release .... In a new deployment what the best choise in your opinion , go head with stein release and upgrade in future or wait for fix in train release ? backup/restore is important for my customers. | 14:43 |
| openstack | Launchpad bug 1860739 in OpenStack Compute (nova) "Cinder backup failed for restore volume snapshot (for instance snapshot)" [Undecided,Confirmed] | 14:43 |
| rlljorge | yoctozepto: your opinion is very importante .... | 14:46 |
| rlljorge | kplant your opinion is very important | 14:46 |
| mnasiadka | yoctozepto, hrw, mgoddard: https://review.opendev.org/#/c/704309/ | 14:48 |
| patchbot | patch 704309 - kolla-ansible - Remove kolla-ceph - 8 patch sets | 14:48 |
| hrw | mnasiadka: 'Kolla-Ansible Ceph deployment mechanism has been deprecated in Train [1].' - had to read it three times to notice kolla-ansible not ceph-ansible ;D | 14:49 |
| mnasiadka | haha | 14:49 |
| *** rphillips has quit IRC | 14:49 | |
| *** rphillips has joined #openstack-kolla | 14:50 | |
| hrw | mnasiadka: and looks like I need to read external ceph docs first | 14:51 |
| mnasiadka | hrw: happy reading | 14:51 |
| hrw | mnasiadka: as all those cinder_backend_ceph like vars gets now set to 'no' | 14:51 |
| hrw | while it feels like '{{ external_ceph }}' would be better fit there | 14:52 |
| hrw | or sth | 14:52 |
| hrw | mnasiadka: the more I look at it the less I understand ;D | 14:57 |
| *** vmixor has joined #openstack-kolla | 15:04 | |
| hrw | to setup ceph cluster I need 3 machines for osd, right? mon runs on each, mgr on any? | 15:04 |
| mnasiadka | think of mon and mgr as the control plane - usually 3 nodes | 15:05 |
| mnasiadka | osd is where the data resides | 15:05 |
| hrw | so 3 trash pc + 1TB hdd to each | 15:06 |
| *** hrw has quit IRC | 15:15 | |
| *** hrw has joined #openstack-kolla | 15:15 | |
| *** TrevorV has joined #openstack-kolla | 15:30 | |
| *** spiette has quit IRC | 15:32 | |
| openstackgerrit | Adeberg proposed openstack/kolla-ansible master: Swift: remove meta field from rsync command https://review.opendev.org/706303 | 15:32 |
| *** spiette has joined #openstack-kolla | 15:34 | |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: WIP: Create 'infra' type of images https://review.opendev.org/698753 | 16:00 |
| hrw | just rebased to see CI failures | 16:00 |
| mgoddard | yoctozepto: you wanted to ping be about a role for CA certs? https://review.opendev.org/#/c/699888/23/ansible/roles/zun/tasks/config.yml@42 | 16:09 |
| patchbot | patch 699888 - kolla-ansible - Copy CA into containers. - 23 patch sets | 16:09 |
| *** jopdorp has joined #openstack-kolla | 16:20 | |
| jopdorp | hi! | 16:20 |
| jopdorp | how would we go about configuring ldap as the backend for keystone? | 16:20 |
| jopdorp | i found this doc | 16:21 |
| jopdorp | https://docs.openstack.org/kolla-ansible/rocky/reference/keystone-guide.html | 16:21 |
| jopdorp | it hints at using keystone_token_provider somehow | 16:21 |
| noxoid_ | /etc/kolla/config/keystone/domains/keystone.internal.conf | 16:22 |
| noxoid_ | assuming you wanted the domain "internal" to be backed by ldap | 16:22 |
| noxoid_ | you can then put your ldap config in that file as defined by keystone documentation | 16:22 |
| jopdorp | could send a link to that documentation? | 16:23 |
| jopdorp | i found this | 16:23 |
| jopdorp | https://docs.openstack.org/kolla-ansible/latest/admin/advanced-configuration.html | 16:23 |
| noxoid_ | https://docs.openstack.org/keystone/stein/admin/configuration.html#integrate-identity-with-ldap | 16:23 |
| jopdorp | right | 16:24 |
| jopdorp | it's a keystone override config right? | 16:24 |
| jopdorp | how could i derive this file /etc/kolla/config/keystone/domains/keystone.internal.conf | 16:24 |
| jopdorp | ? | 16:24 |
| jopdorp | I feel like that is the key to making many configuration changes | 16:24 |
| noxoid_ | http://paste.openstack.org/show/789232/ | 16:26 |
| noxoid_ | thats what we use to integrate with our freeipa service. not all settings may be correct but it minimally works | 16:26 |
| jopdorp | that helps, thank you! | 16:26 |
| jopdorp | still, how did you find that it had to be in this file: /etc/kolla/config/keystone/domains/keystone.internal.conf | 16:27 |
| jopdorp | ? | 16:27 |
| noxoid_ | by reading the kolla-ansible source code | 16:27 |
| jopdorp | right | 16:27 |
| jopdorp | okay | 16:27 |
| hrw | ok. failed as expected | 16:40 |
| *** rpittau is now known as rpittau|afk | 16:40 | |
| *** jopdorp has quit IRC | 16:44 | |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: WIP: Create 'infra' type of images https://review.opendev.org/698753 | 16:45 |
| hrw | some debugs and dropped many CI jobs | 16:46 |
| *** k_mouza has quit IRC | 16:49 | |
| *** jovial[m] has joined #openstack-kolla | 16:51 | |
| *** weshay|ruck is now known as weshay | 16:52 | |
| jovial[m] | How can we stop old config files persisting in the fluentd container? | 16:52 |
| jovial[m] | I was thinking we could just copy the folders and not *.conf in config_files - thoughts? | 16:54 |
| jovial[m] | that should nuke the old files on restart, right? | 16:54 |
| jovial[m] | any downsides to that? | 16:55 |
| *** carli has quit IRC | 17:00 | |
| mgoddard | yoctozepto: so much for those saved CI cycles :p | 17:03 |
| mnasiadka | mgoddard: you talk about it like it's a princess worth saving :) | 17:04 |
| mgoddard | and I am the frog | 17:04 |
| mgoddard | jovial[m]: I'm not sure about the behaviour of directory copies in that script | 17:05 |
| mgoddard | but nuking implies some extra work | 17:05 |
| jovial[m] | merge: merges the source directory into the target directory instead of replacing it. Boolean, defaults to false. | 17:06 |
| mgoddard | maybe we need an option like rsync's --delete | 17:06 |
| *** ktibi has quit IRC | 17:06 | |
| mgoddard | anyway, heading of | 17:06 |
| openstackgerrit | Will Szumski proposed openstack/kolla-ansible master: Delete stale fluent config on restart https://review.opendev.org/706327 | 17:15 |
| *** ivve has quit IRC | 17:22 | |
| openstackgerrit | Radosław Piliszek proposed openstack/kolla master: CI: Replace cinder-lvm scenario with zun scenario https://review.opendev.org/706059 | 17:25 |
| openstackgerrit | Merged openstack/kolla-ansible stable/train: Fix multiple issues with MariaDB handling https://review.opendev.org/705414 | 17:26 |
| *** k_mouza has joined #openstack-kolla | 17:26 | |
| *** k_mouza has quit IRC | 17:30 | |
| openstackgerrit | Radosław Piliszek proposed openstack/kolla master: CI: Replace cinder-lvm scenario with zun scenario https://review.opendev.org/706059 | 17:32 |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: WIP: Create 'infra' type of images https://review.opendev.org/698753 | 17:33 |
| *** evrardjp has quit IRC | 17:33 | |
| hrw | let's see will it pass | 17:34 |
| *** evrardjp has joined #openstack-kolla | 17:34 | |
| *** k_mouza has joined #openstack-kolla | 17:37 | |
| openstackgerrit | Radosław Piliszek proposed openstack/kolla master: CI: Replace cinder-lvm scenario with zun scenario https://review.opendev.org/706059 | 17:40 |
| openstackgerrit | Radosław Piliszek proposed openstack/kolla master: CI: Replace cinder-lvm scenario with zun scenario https://review.opendev.org/706059 | 17:42 |
| hrw | infra build progresses ;D | 17:43 |
| *** dougsz has quit IRC | 17:54 | |
| *** jonaspaulo has quit IRC | 17:59 | |
| hrw | INFO:kolla.common.utils.ceph-base:Successfully tagged kolla/debian-infra-ceph-base:9.1.0 | 17:59 |
| hrw | https://3358148dd61163c40ee5-fed9d885f48c89b65645af8cb1093bc8.ssl.cf5.rackcdn.com/698753/7/check/kolla-build-debian-source/a2823f5/kolla/build/ceph-base.log | 17:59 |
| hrw | YAY! | 18:00 |
| hrw | https://3358148dd61163c40ee5-fed9d885f48c89b65645af8cb1093bc8.ssl.cf5.rackcdn.com/698753/7/check/kolla-build-debian-source/a2823f5/kolla/docker-info.txt shows source AND infra images built | 18:01 |
| hrw | now starts the problem... | 18:01 |
| hrw | CI needs to build binary/source/infra at one machine | 18:02 |
| *** dcapone2004 has joined #openstack-kolla | 18:32 | |
| dcapone2004 | Does kolla-ansible support deployment using the neutron "trunk" plugin? | 18:32 |
| dcapone2004 | I have attempted to modify the neutron.conf file in /etc/kolla/config to append trunk to the list of service-plugins and ran a reconfigure, but the neutron.conf generated and placed in the containers does not include the trunk service plugin | 18:33 |
| *** spiette has quit IRC | 18:40 | |
| *** spiette has joined #openstack-kolla | 18:45 | |
| yoctozepto | dcapone2004: hmm, it looks missing from default service_plugins | 18:47 |
| yoctozepto | dcapone2004: the ugly solution is to write down all service_plugins that are in use atm | 18:47 |
| yoctozepto | and add trunk in an override | 18:47 |
| yoctozepto | then reconfigure | 18:47 |
| *** ivve has joined #openstack-kolla | 18:47 | |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: WIP: Create 'infra' type of images https://review.opendev.org/698753 | 18:50 |
| hrw | this build will be tricky | 18:51 |
| dcapone2004 | when you say in an override, you mean in the /etc/kolla/config/neutron/neutron.conf file? | 18:56 |
| dcapone2004 | because that is basically what I did, but it doesn't seem to pick up the added service_plugin | 18:57 |
| dcapone2004 | [DEFAULT] | 18:58 |
| dcapone2004 | service_plugins=firewall_v2,vpnaas,trunk | 18:58 |
| dcapone2004 | is what is in the /etc/kolla/config/neutron/neutron.conf file | 18:58 |
| dcapone2004 | and inside the neutron-server container after reconfigure, I get... | 19:08 |
| dcapone2004 | docker exec -it 5adc50321697 cat /etc/neutron/neutron.conf | grep plugin | 19:08 |
| dcapone2004 | core_plugin = ml2 | 19:08 |
| dcapone2004 | service_plugins = firewall_v2,router,vpnaas | 19:08 |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: WIP: Create 'infra' type of images https://review.opendev.org/698753 | 19:10 |
| yoctozepto | dcapone2004: it's /etc/kolla/config/neutron.conf | 19:13 |
| yoctozepto | use directory for per-service overrides | 19:13 |
| dcapone2004 | ok...I thought neutron would be considered a "service"....what would be considered a service for purposes of overrides so I can update my docs | 19:14 |
| *** k_mouza has quit IRC | 19:19 | |
| *** k_mouza has joined #openstack-kolla | 19:25 | |
| *** k_mouza has quit IRC | 19:30 | |
| dcapone2004 | nice...that worked, thanks...just wish things were a bit more clear on what is a service and what isn't to know the correct place for overrides | 19:41 |
| openstackgerrit | Radosław Piliszek proposed openstack/kolla master: CI: Replace cinder-lvm scenario with zun scenario https://review.opendev.org/706059 | 19:42 |
| *** rlljorge has quit IRC | 19:51 | |
| *** jonaspaulo has joined #openstack-kolla | 19:51 | |
| openstackgerrit | Radosław Piliszek proposed openstack/kolla master: CentOS 8: Add deploy jobs in CI https://review.opendev.org/705952 | 19:53 |
| yoctozepto | dcapone2004: well, the docs mention both locations afair | 19:54 |
| *** openstackstatus has joined #openstack-kolla | 19:56 | |
| *** ChanServ sets mode: +v openstackstatus | 19:56 | |
| dcapone2004 | yes, I thought it meant that it should work in either location, not necessarily that only one of the locations would work and to try all | 19:57 |
| openstackgerrit | Will Szumski proposed openstack/kolla-ansible master: Delete stale fluent config on restart https://review.opendev.org/706327 | 20:00 |
| *** tonythomas has quit IRC | 20:07 | |
| yoctozepto | dcapone2004: nah, dirs are for specific services like neutron-server for the api server | 20:10 |
| *** kplant has quit IRC | 20:11 | |
| *** gfidente is now known as gfidente|afk | 20:17 | |
| *** jonaspaulo has quit IRC | 20:24 | |
| openstackgerrit | Rafael Folco proposed openstack/kolla master: [WIP] centos8: remove rabbitmq pinning https://review.opendev.org/703640 | 20:32 |
| openstackgerrit | Rafael Folco proposed openstack/kolla master: [WIP] centos8: remove rabbitmq pinning https://review.opendev.org/703640 | 20:33 |
| dking_desktop | Could somebody point me in the right direction as to what is going on when I see the following error in the ironic-inspector log: NotFound: Can not transition from state 'starting' on event 'start' (no defined transition) | 20:46 |
| openstackgerrit | Marcin Juszkiewicz proposed openstack/kolla master: WIP: Create 'infra' type of images https://review.opendev.org/698753 | 20:47 |
| *** irclogbot_0 has quit IRC | 20:53 | |
| *** irclogbot_3 has joined #openstack-kolla | 20:54 | |
| openstackgerrit | Antony Messerli proposed openstack/kolla master: Add ssh clients to Ironic Conductor container https://review.opendev.org/706149 | 21:02 |
| *** k_mouza has joined #openstack-kolla | 21:11 | |
| *** k_mouza has quit IRC | 21:16 | |
| *** TrevorV has quit IRC | 21:16 | |
| *** dcapone2004 has quit IRC | 21:17 | |
| *** gfidente|afk has quit IRC | 21:22 | |
| yoctozepto | dking_desktop: best try on #openstack-ironic | 21:40 |
| dking_desktop | yoctozepto: Thank you. | 21:45 |
| *** vmixor has quit IRC | 22:08 | |
| *** dking_desktop has quit IRC | 22:09 | |
| *** negronjl has quit IRC | 22:12 | |
| *** negronjl has joined #openstack-kolla | 22:17 | |
| *** ivve has quit IRC | 22:31 | |
| *** k_mouza has joined #openstack-kolla | 22:49 | |
| openstackgerrit | Adeberg proposed openstack/kolla-ansible master: Swift: remove meta field from rsync command https://review.opendev.org/706303 | 22:58 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!