| *** gyee has quit IRC | 00:33 | |
| *** cshen has joined #openstack-ansible | 00:49 | |
| *** cshen has quit IRC | 01:23 | |
| *** markvoelker has joined #openstack-ansible | 02:13 | |
| *** markvoelker has quit IRC | 02:18 | |
| *** rh-jelabarre has quit IRC | 02:20 | |
| *** also_stingrayza has joined #openstack-ansible | 03:06 | |
| *** stingrayza has quit IRC | 03:09 | |
| *** markvoelker has joined #openstack-ansible | 03:48 | |
| *** markvoelker has quit IRC | 03:53 | |
| *** KeithMnemonic has quit IRC | 04:55 | |
| *** udesale has joined #openstack-ansible | 04:57 | |
| *** cshen has joined #openstack-ansible | 05:05 | |
| *** yasemind has joined #openstack-ansible | 05:21 | |
| *** shyamb has joined #openstack-ansible | 05:29 | |
| *** shyamb has quit IRC | 05:45 | |
| *** markvoelker has joined #openstack-ansible | 05:49 | |
| *** shyamb has joined #openstack-ansible | 05:53 | |
| *** markvoelker has quit IRC | 05:54 | |
| *** shyam89 has joined #openstack-ansible | 06:03 | |
| *** cshen has quit IRC | 06:03 | |
| *** shyamb has quit IRC | 06:06 | |
| CeeMac | morning | 06:17 |
|---|---|---|
| *** arkan has joined #openstack-ansible | 06:38 | |
| *** arkan has quit IRC | 06:50 | |
| *** arkan has joined #openstack-ansible | 06:50 | |
| *** markPilon has joined #openstack-ansible | 06:52 | |
| *** this10nly has joined #openstack-ansible | 06:53 | |
| *** cshen has joined #openstack-ansible | 07:05 | |
| *** cshen has quit IRC | 07:09 | |
| *** shyam89 has quit IRC | 07:17 | |
| *** udesale has quit IRC | 07:17 | |
| *** tosky has joined #openstack-ansible | 07:33 | |
| *** arkan has quit IRC | 07:50 | |
| *** andrewbonney has joined #openstack-ansible | 07:51 | |
| *** udesale has joined #openstack-ansible | 07:57 | |
| openstackgerrit | Hervé Beraud proposed openstack/openstack-ansible master: Use unittest.mock instead of mock https://review.opendev.org/734452 | 07:57 |
| *** also_stingrayza is now known as stingrayza | 08:08 | |
| *** jbadiapa has joined #openstack-ansible | 08:18 | |
| *** shyamb has joined #openstack-ansible | 08:24 | |
| *** shyam89 has joined #openstack-ansible | 08:36 | |
| *** shyamb has quit IRC | 08:39 | |
| *** spatel has joined #openstack-ansible | 08:49 | |
| *** arkan has joined #openstack-ansible | 08:51 | |
| *** spatel has quit IRC | 08:53 | |
| *** aedc has joined #openstack-ansible | 08:53 | |
| arkan | hi | 08:58 |
| arkan | jrosser: I'm making progress with magnum | 08:58 |
| arkan | just I want to ask, right now it passed 11 min. | 08:59 |
| arkan | kube-master status - Create In Progress | 08:59 |
| arkan | is this normal ? | 08:59 |
| arkan | in magnum logs, I don't see errors | 08:59 |
| *** cshen has joined #openstack-ansible | 09:02 | |
| *** cshen has quit IRC | 09:06 | |
| *** shyamb has joined #openstack-ansible | 09:24 | |
| *** shyam89 has quit IRC | 09:25 | |
| *** shyam89 has joined #openstack-ansible | 09:41 | |
| *** shyamb has quit IRC | 09:44 | |
| *** aedc has quit IRC | 09:54 | |
| *** aedc has joined #openstack-ansible | 09:54 | |
| *** shyamb has joined #openstack-ansible | 09:56 | |
| *** shyam89 has quit IRC | 10:00 | |
| *** udesale_ has joined #openstack-ansible | 10:01 | |
| *** udesale has quit IRC | 10:04 | |
| *** shyamb has quit IRC | 10:12 | |
| *** sshnaidm|afk is now known as sshnaidm|off | 10:16 | |
| *** markPilon has quit IRC | 10:25 | |
| *** shyamb has joined #openstack-ansible | 10:32 | |
| *** shyamb has quit IRC | 10:47 | |
| *** arkan has quit IRC | 10:49 | |
| *** mgariepy has quit IRC | 10:58 | |
| *** cshen has joined #openstack-ansible | 11:03 | |
| *** cshen has quit IRC | 11:07 | |
| *** shyamb has joined #openstack-ansible | 11:20 | |
| *** pcaruana has quit IRC | 11:33 | |
| *** pcaruana has joined #openstack-ansible | 11:44 | |
| *** mgariepy has joined #openstack-ansible | 12:02 | |
| *** rh-jelabarre has joined #openstack-ansible | 12:07 | |
| *** sakharkar has joined #openstack-ansible | 12:08 | |
| sakharkar | Openstack ansible train support ubuntu 20? | 12:09 |
| *** shyamb has quit IRC | 12:13 | |
| admin0 | quick question .. is my br-vxlan is on diffrent subnets, will openstack/neutron work ? | 12:37 |
| admin0 | say i have servers in 3 racks, and each rack has its own subnet .. ( not a layer2 but a layer3) .. would openstack/osa still work fine ? | 12:38 |
| admin0 | how does vxlan/vtep discovery work ? say instance1 (in rack1) wants to talk to a newly created instance on server10 on rack10 .. | 12:38 |
| admin0 | i do know how this works .. question is if the router forwards multicast, can those vxlan ips be on l3 and not l3 | 12:43 |
| admin0 | not l2 | 12:44 |
| watersj | admin0, not expert, your switches will need to support bxvlan. You'll need to configure and allow similar vlans. Nicest thing about bxvlan its allows network segemnt to extend across layer 3 which is just your case. | 12:44 |
| watersj | admin0, mistyped, vxlan (not bvxlan) | 12:45 |
| *** spatel has joined #openstack-ansible | 12:50 | |
| jrosser | admin0: if you set up subnets in each rack, make sure they all route, use PIM for multicast and add routes on the computer/network nodes to the whole tunnel cidr then it works fine. I have a deployment like that. | 12:51 |
| jrosser | there is a “pod” example config in openstack-ansible/etc which will give you some ideas | 12:53 |
| jrosser | you do not need vxlan support in your switches to make that work because the vtep are all on the hosts | 12:55 |
| *** spatel has quit IRC | 12:55 | |
| watersj | jrosser, they can cross different subnets? | 12:55 |
| jrosser | the vtep ip can be in deifferent subnets | 12:56 |
| jrosser | so long as those subnets can route between themselves, and the hosts have a static route added to the entire L3 cidr via the local TOR gateway ip | 12:57 |
| jrosser | you need multicast routing as well to make the Mac learning work | 12:57 |
| jrosser | watersj: what you say about needing vxlan on the switches is right though if you ever need a L2 service/provider network like might be needed for octavia or trove | 12:59 |
| jrosser | but I guess you could also do that as extra Linux vtep, lots of options really | 13:00 |
| *** cshen has joined #openstack-ansible | 13:04 | |
| *** cshen has quit IRC | 13:10 | |
| *** spatel has joined #openstack-ansible | 14:01 | |
| watersj | when you create you masakari segment hosts, should you use mgmt ip assigned to compute node (br-mgmt) | 14:06 |
| watersj | or hostname? | 14:07 |
| watersj | Host with name 172.29.236.33 could not be found | 14:08 |
| watersj | which exists | 14:08 |
| watersj | guess not, Hypervisor with name 172.29.236.33 could not be found | 14:10 |
| *** this10nly has quit IRC | 14:12 | |
| *** aedc_ has joined #openstack-ansible | 14:28 | |
| *** npalladium has joined #openstack-ansible | 14:29 | |
| *** npalladium has quit IRC | 14:29 | |
| *** npalladium has joined #openstack-ansible | 14:29 | |
| *** aedc has quit IRC | 14:31 | |
| * npalladium Having an issue with Openstack-Ansible. Not quite sure if I can ask here; please let me know if there is a better place. | 14:46 | |
| * npalladium the TASK [lxc_hosts : Ensure that the LXC cache has been prepared] is failing on Ubuntu 18.04. I'm not the first person with this error and someone else seems to have gotten it on 16.04. | 14:47 | |
| * npalladium it fails at the cmd": "chroot /var/lib/machines/ubuntu-bionic-amd64 /opt/cache-prep-commands.sh > /var/log/lxc-cache-prep-commands.log 2>&1" | 14:48 | |
| * npalladium That probably corresponds to "chroot {{ lxc_image_cache_path }} /opt/cache-prep-commands.sh > /var/log/lxc-cache-prep-commands.log 2>&1" in lxc_cache_preparation.yml. | 14:49 | |
| *** aedc_ has quit IRC | 14:49 | |
| * npalladium the log file ends with E: The repository 'https://download.ceph.com/debian-octopus bionic Release' does not have a Release file . | 14:49 | |
| * npalladium Any help on this would be appreciated. | 14:50 | |
| *** pcaruana has quit IRC | 14:58 | |
| *** cshen has joined #openstack-ansible | 15:06 | |
| *** pcaruana has joined #openstack-ansible | 15:11 | |
| *** cshen has quit IRC | 15:11 | |
| admin0 | jrosser, thanks .. i understood the pods now :) | 15:12 |
| *** gyee has joined #openstack-ansible | 15:26 | |
| *** hwoarang has quit IRC | 15:27 | |
| npalladium | Heya. Just realised all the messages were prepended with me. Do I send them again? (Apologies for the spam, but I do require help :D ) | 15:30 |
| *** udesale_ has quit IRC | 15:37 | |
| watersj | wow, that was an easier fix. change /etc/masakarimonitors/masakarimonitors.conf host = ip to fqdn | 15:37 |
| jrosser | npalladium: I guess it is failing to “apt update” in the chroot, you can always paste the log file at paste.openstack.org if necessary | 15:49 |
| jrosser | however this really is just a case of old fashioned debugging, I think that repo config has been copied from your host | 15:49 |
| jrosser | so first check that the host can apt update | 15:50 |
| jrosser | maybe then enter the choot manually and try the same thing until you find what is broken | 15:51 |
| npalladium | The host can indeed apt update. Manually running the command fails too with the same error. | 15:53 |
| jrosser | did you chroot and poke around in a shell or just run the script? | 15:55 |
| npalladium | Just ran script. Oops. Legwork it is. Thanks! | 15:55 |
| jrosser | perhaps the apt key is missing, something like that, will need some detective work | 15:56 |
| npalladium | I guess so too. Either a missing apt key or some utility like gnupg missing; but something like the latter should've been caught by someone else too I guess. | 15:57 |
| jrosser | is this a straightforward AIO build? | 15:58 |
| npalladium | Not AIO but close enough; all infra on one host and compute on another. I'm using the example openstack_user_config.yml without changing much. | 16:03 |
| jrosser | I would recommend having an AIO handy because it is an excellent reference for issues like this | 16:10 |
| jrosser | 8core/8G ram VM and a couple of hours to set up are time well invested | 16:11 |
| admin0 | jrosser, +1 | 16:11 |
| admin0 | an AIO is always handy to test a lot of stuff | 16:11 |
| jrosser | export SCENARIO=aio_lxc_ceph and you’ll get something identical to what we run CI tests on multiple times a day | 16:12 |
| *** mgariepy has quit IRC | 16:13 | |
| *** jbadiapa has quit IRC | 16:21 | |
| jrosser | admin0: another thing with pods is you need to think very carefully about where infra and network nodes go | 16:33 |
| jrosser | becasue they are running vrrp (keepalived) for haproxy and neutron so need a L2 adjacency somehow | 16:33 |
| *** yolanda has quit IRC | 16:39 | |
| npalladium | jrosser: Noted about the AIO with thanks. Currently a bit constrained (I'm an undergrad deploying it for my uni), but will try to have one setup in the near future. | 16:47 |
| *** markvoelker has joined #openstack-ansible | 16:59 | |
| *** cshen has joined #openstack-ansible | 17:07 | |
| *** mgariepy has joined #openstack-ansible | 17:09 | |
| *** cshen has quit IRC | 17:12 | |
| *** owalsh has quit IRC | 17:21 | |
| *** andrewbonney has quit IRC | 17:36 | |
| * prometheanfire has an odd thing where ovs wants to create a bridge and attach the networkd created bridge of the same name to it, the bridge is defined in provider_networks as the only place so not sure why ovs is trying to use the same name | 17:37 | |
| prometheanfire | https://gist.github.com/prometheanfire/6c63397edca59d071a8735385e6cccbb for the ouc.yml stanza | 17:38 |
| *** owalsh has joined #openstack-ansible | 17:42 | |
| npalladium | Quick update on the chache prep in the chroot issue: Couldn't reproduce the issue upon trying again now; I don't know what parameters have changed though as I haven't changed anything. The play now reached almost the end before erroring out. I'll try to reproduce it later after getting the current setup working. jrosser, thanks a lot! | 17:47 |
| jrosser | good to hear it’s got further :) | 17:49 |
| *** arkan has joined #openstack-ansible | 17:50 | |
| *** chandankumar is now known as raukadah | 18:02 | |
| *** NobodyCam has quit IRC | 18:12 | |
| *** NobodyCam has joined #openstack-ansible | 18:12 | |
| jamesdenton | prometheanfire the question is.. why a networkd-managed bridge? | 18:15 |
| npalladium | jrosser: Haha thanks again! On a completely different note, any place a newbie to Openstack but with some Python and Ansible experience can contribute? :P | 18:17 |
| prometheanfire | jamesdenton: because I need the host to be on that network, it's my understanding that provider networks should be created outside of OSA's control | 18:17 |
| prometheanfire | jamesdenton: I see that os_neutron creates the ovs networks using that same name | 18:18 |
| jamesdenton | prometheanfire at some point a decision was made to assist in creating the actual OVS provider bridges because it was often a missed step. | 18:18 |
| jamesdenton | and in most deployments you wouldn't also need a bridge with the same name at the host level | 18:18 |
| jamesdenton | we solved for the 99% not the 1% :D | 18:18 |
| prometheanfire | wooo, I'm the 1% | 18:19 |
| jamesdenton | i'm curious, what does the host need with an IP on br-4009? Using the host as a gateway for that network w/ masquerade? | 18:19 |
| prometheanfire | jamesdenton: I guess the question I have is, should I just name the network diferent in (the container_bridge value)? | 18:19 |
| prometheanfire | ya, that's the outside access | 18:20 |
| jamesdenton | the value of container_bridge is ultimately what will be used to create the OVS bridge, and thus is used in neutron config files | 18:20 |
| jrosser | npalladium: if you like what you see with openstack-ansible we are always looking out for new contributors to help maintain | 18:20 |
| prometheanfire | jrosser: right, I see that, what's confusing is that it's not using network_interface for the neutron config file | 18:21 |
| jamesdenton | it will connect the interface specified as network_interface to the bridge specified by container_bridge | 18:22 |
| prometheanfire | bridge_mappings = vlan-4009:br-4009 (same value as net_name) and then container_bridge and not network_interface | 18:22 |
| prometheanfire | but it also seems to want to connect the container_bridge value | 18:22 |
| jamesdenton | if networkd is creating the bridge as an 'ovs' bridge, then the task should be OK. if a linuxbridge, then you may have issue | 18:22 |
| jrosser | npalladium: for something with more python there is a set of ansible modules for interacting with the openstack api, to create vm/networks and so on. those modules were recently spun out out of ansible itself | 18:23 |
| jamesdenton | right, provider label:provider bridge. in 'ovs-vsctl show' you'll see network_interface connected to container_bridge | 18:23 |
| prometheanfire | jamesdenton: https://gist.github.com/prometheanfire/6c63397edca59d071a8735385e6cccbb | 18:23 |
| jamesdenton | what does 'ip link show br-4009' provide? | 18:23 |
| prometheanfire | 13: br-4009: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP mode DEFAULT group default qlen 1000 | 18:24 |
| prometheanfire | I didn't specify it to be a ovs bridge | 18:24 |
| *** johnsom has quit IRC | 18:24 | |
| *** johnsom has joined #openstack-ansible | 18:25 | |
| jamesdenton | right. that's why you have an error. So you could just rename either bridge, since i'm guessing the idea is to cross connect w/ a veth? | 18:25 |
| prometheanfire | yep | 18:25 |
| prometheanfire | ya, I'll rename the ovs one as either ovs or osa | 18:25 |
| prometheanfire | see what happens :D | 18:25 |
| jamesdenton | there ya go | 18:25 |
| npalladium | jrosser: That sounds interesting and I like the project in general too. That said Openstack itself seems daunting. (Also I'm not quite sure of netiquette on a large IRC channel; I seem to be interrupting other threads. Please let me know if I should change something or shift the discussion elsewhere.) | 18:26 |
| prometheanfire | jamesdenton: it seems like we are overloading the variable | 18:27 |
| jamesdenton | meh | 18:27 |
| prometheanfire | also, https://github.com/systemd/systemd/issues/2613 heh | 18:28 |
| jrosser | npalladium: don’t worry, it’s fine. etiquette is “just ask” | 18:28 |
| jamesdenton | reinventing the wheel. always fun. | 18:28 |
| prometheanfire | jamesdenton: you see that systemd added sr-iov support? | 18:28 |
| jamesdenton | yes, you pointed that out. haven't had time to play with it though | 18:28 |
| prometheanfire | ah, cool | 18:28 |
| prometheanfire | jamesdenton: thanks for puting a pad between my head and the wall :D | 18:42 |
| jamesdenton | i'm here for you | 18:43 |
| *** cshen has joined #openstack-ansible | 19:08 | |
| *** cshen has quit IRC | 19:12 | |
| tow | jrosser: after overcoming the repo_hosts issue, the deployment is pretty smooth (20.1.3) but for the central logging, rsyslog containers are created, but nothing else seems to have been configured, as far as we understand defining log_hosts is all it takes, correct? | 19:16 |
| jrosser | tow: rsyslog is pretty much all replaced with systemd journals now | 19:17 |
| jrosser | there is some config to do journal forwarding over http to one of the log hosts perhaps? systemd-journal-remote iirc | 19:20 |
| *** owalsh has quit IRC | 19:20 | |
| *** cshen has joined #openstack-ansible | 19:20 | |
| *** owalsh has joined #openstack-ansible | 19:29 | |
| *** mloza has quit IRC | 19:35 | |
| *** mloza has joined #openstack-ansible | 19:36 | |
| tow | jrosser: oh it looks like it is forced disabled even if log_hosts are defined: journald_remote_enabled | 19:39 |
| jrosser | you should check the git history for that, I think there is a related systemd bug | 19:40 |
| tow | yeah, there is a comment in the playbook, journalctl vacuum issue, long standing bug... okay thanks | 19:44 |
| *** mloza has quit IRC | 20:27 | |
| *** spatel has quit IRC | 20:29 | |
| *** markvoelker has quit IRC | 20:42 | |
| *** arkan has quit IRC | 20:49 | |
| admin0 | new openstack/train install .. there is no openstack in the utility container | 20:57 |
| *** rpittau has quit IRC | 21:12 | |
| *** rpittau has joined #openstack-ansible | 21:13 | |
| *** jonmasters has quit IRC | 21:36 | |
| *** jonmasters has joined #openstack-ansible | 21:36 | |
| *** gyee has quit IRC | 21:40 | |
| *** gyee has joined #openstack-ansible | 21:45 | |
| *** also_stingrayza has joined #openstack-ansible | 21:49 | |
| *** melwitt is now known as jgwentworth | 21:49 | |
| *** stingrayza has quit IRC | 21:52 | |
| *** rh-jelabarre has quit IRC | 22:06 | |
| *** markvoelker has joined #openstack-ansible | 22:22 | |
| *** markvoelker has quit IRC | 22:27 | |
| *** markvoelker has joined #openstack-ansible | 22:47 | |
| *** markvoelker has quit IRC | 22:52 | |
| *** tosky_ has joined #openstack-ansible | 22:52 | |
| *** tosky has quit IRC | 22:53 | |
| *** gyee has quit IRC | 23:01 | |
| *** tosky_ is now known as tosky | 23:01 | |
| *** tosky has quit IRC | 23:38 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!