| *** zhurong has quit IRC | 00:01 | |
| *** egonzalez90 has joined #openstack-kolla | 00:01 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla: Customizations for Designate https://review.openstack.org/350172 | 00:07 |
|---|---|---|
| *** diogogmt has quit IRC | 00:08 | |
| *** britthouser has quit IRC | 00:10 | |
| *** ad_rien_ has joined #openstack-kolla | 00:15 | |
| *** ad_rien_ has quit IRC | 00:18 | |
| *** jtriley has joined #openstack-kolla | 00:22 | |
| *** jtriley has quit IRC | 00:27 | |
| *** egonzalez90 has quit IRC | 00:30 | |
| *** duonghq has joined #openstack-kolla | 00:45 | |
| *** sdake has joined #openstack-kolla | 00:46 | |
| sdake | evening | 00:47 |
| duonghq | evening sdake | 00:47 |
| duonghq | but morning here | 00:48 |
| *** diogogmt has joined #openstack-kolla | 01:05 | |
| *** zhurong has joined #openstack-kolla | 01:05 | |
| *** ravig has joined #openstack-kolla | 01:06 | |
| *** daneyon_ has joined #openstack-kolla | 01:06 | |
| *** daneyon_ has quit IRC | 01:11 | |
| *** jtriley has joined #openstack-kolla | 01:23 | |
| *** zhubingbing_ has joined #openstack-kolla | 01:24 | |
| *** rhallisey has joined #openstack-kolla | 01:24 | |
| *** zhubingbing__ has quit IRC | 01:27 | |
| *** jtriley has quit IRC | 01:27 | |
| *** unicell has quit IRC | 01:29 | |
| *** unicell has joined #openstack-kolla | 01:30 | |
| *** ravig has quit IRC | 01:36 | |
| *** ravig has joined #openstack-kolla | 01:42 | |
| *** unicell has quit IRC | 01:46 | |
| *** ravig has quit IRC | 01:53 | |
| *** ravig has joined #openstack-kolla | 01:58 | |
| *** sdake has quit IRC | 02:01 | |
| *** rhallisey has quit IRC | 02:01 | |
| *** gfidente has quit IRC | 02:11 | |
| *** sdake has joined #openstack-kolla | 02:13 | |
| *** jtriley has joined #openstack-kolla | 02:24 | |
| *** jtriley has quit IRC | 02:29 | |
| *** sdake has quit IRC | 02:34 | |
| duonghq | anybody deploy Kolla in AIO, Cinder with LVM-backed in 2NIC machine? | 02:48 |
| *** ayoung has joined #openstack-kolla | 02:49 | |
| duonghq | mntdevops: has you deployed with above configuration? | 02:50 |
| *** sdake has joined #openstack-kolla | 02:54 | |
| duonghq | anybody deploy Kolla in AIO, Cinder with LVM-backed in 2NIC machine? | 02:54 |
| *** senk_ has joined #openstack-kolla | 03:02 | |
| *** diogogmt has quit IRC | 03:05 | |
| mntdevops | duonghq, yes | 03:20 |
| mntdevops | duonghq, http://mntdevops.com/2016/08/08/iaas-2/ | 03:21 |
| duonghq | mntdevops: I already read that but I cannot figure out why the iscsid keeps try to connect to one IP while the tgtd is listening on another IP (on other NIC) | 03:22 |
| mntdevops | hmm | 03:23 |
| mntdevops | well | 03:23 |
| mntdevops | that's literally all my code in github | 03:24 |
| mntdevops | *shrugs* | 03:24 |
| *** ravig has quit IRC | 03:24 | |
| mntdevops | once I got the cinder-volumes made | 03:24 |
| *** sdake has quit IRC | 03:24 | |
| mntdevops | it worked as advertised | 03:24 |
| mntdevops | https://github.com/mntdevops/netpxemirror-ac/blob/master/ansible/roles/os-aio/files/globals.yml | 03:25 |
| mntdevops | all my interfaces are network_interface with the exception of neutron | 03:25 |
| *** jtriley has joined #openstack-kolla | 03:26 | |
| duonghq | same here | 03:27 |
| duonghq | mntdevops: which is the interface do you use to connect from your laptop to the mp? | 03:29 |
| mntdevops | enp7s0f0 | 03:29 |
| mntdevops | 172.22.10.40 is the vip | 03:30 |
| mntdevops | .50 is the real | 03:30 |
| duonghq | does ovs mesh up with you connection when you deploy? | 03:30 |
| *** senk_ has quit IRC | 03:31 | |
| *** jtriley has quit IRC | 03:31 | |
| *** dave-mccowan has quit IRC | 03:42 | |
| *** ayoung has quit IRC | 03:45 | |
| *** britthouser has joined #openstack-kolla | 03:46 | |
| *** britthou_ has joined #openstack-kolla | 03:47 | |
| *** daneyon_ has joined #openstack-kolla | 03:48 | |
| *** britthouser has quit IRC | 03:50 | |
| *** daneyon_ has quit IRC | 03:53 | |
| *** senk_ has joined #openstack-kolla | 03:54 | |
| *** sdake has joined #openstack-kolla | 03:54 | |
| sdake | duonghq sbezverk is the guy to talk to on that subject | 03:55 |
| sdake | duonghq smells like a bug | 03:55 |
| mntdevops | duonghq: how would i check? | 03:56 |
| duonghq | sdake: I think it is my fault with unusual machine configuration | 03:56 |
| duonghq | this is a physical machine with 2NIC, I only plugged 1 cable | 03:56 |
| mntdevops | ha, same here | 03:56 |
| mntdevops | well | 03:56 |
| mntdevops | the second is plugged in, but not really going anywhere | 03:57 |
| duonghq | mntdevops: is it NO-CARRIER status? | 03:58 |
| duonghq | when I use ssh tunneling trick, it works but I'm not comfortable with tricky solution | 04:01 |
| mntdevops | duonghq: enp7s0f1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500 | 04:15 |
| duonghq | it's ok now, my fault, thank sdake, mntdevops | 04:16 |
| *** fragatina has quit IRC | 04:16 | |
| mntdevops | np | 04:16 |
| *** britthou_ has quit IRC | 04:23 | |
| *** britthouser has joined #openstack-kolla | 04:24 | |
| *** jtriley has joined #openstack-kolla | 04:26 | |
| openstackgerrit | Shaun Smekel proposed openstack/kolla: Customizations for Elasticsearch https://review.openstack.org/354508 | 04:31 |
| *** jtriley has quit IRC | 04:31 | |
| *** senk_ has quit IRC | 04:34 | |
| *** fragatina has joined #openstack-kolla | 04:58 | |
| *** fragatina has quit IRC | 05:03 | |
| *** britthouser has quit IRC | 05:11 | |
| sdake | ok tme to ptfo | 05:26 |
| sdake | night folks | 05:26 |
| *** jtriley has joined #openstack-kolla | 05:27 | |
| duonghq | night sdake | 05:27 |
| *** jtriley has quit IRC | 05:32 | |
| *** sdake has quit IRC | 05:32 | |
| *** coolsvap has joined #openstack-kolla | 05:34 | |
| *** unicell has joined #openstack-kolla | 05:35 | |
| *** senk_ has joined #openstack-kolla | 05:39 | |
| *** unicell has quit IRC | 05:49 | |
| *** unicell has joined #openstack-kolla | 05:49 | |
| *** unicell has quit IRC | 05:50 | |
| *** unicell has joined #openstack-kolla | 05:50 | |
| *** jtriley has joined #openstack-kolla | 06:28 | |
| *** dmueller-evoila has joined #openstack-kolla | 06:29 | |
| *** daneyon_ has joined #openstack-kolla | 06:31 | |
| *** jtriley has quit IRC | 06:33 | |
| *** daneyon_ has quit IRC | 06:35 | |
| *** b_bezak has joined #openstack-kolla | 07:07 | |
| *** rkrum has quit IRC | 07:16 | |
| *** jtriley has joined #openstack-kolla | 07:29 | |
| *** jtriley has quit IRC | 07:34 | |
| *** zhubingbing__ has joined #openstack-kolla | 08:01 | |
| *** zhubingbing_ has quit IRC | 08:04 | |
| *** athomas has joined #openstack-kolla | 08:07 | |
| *** matrohon has joined #openstack-kolla | 08:07 | |
| *** shardy has joined #openstack-kolla | 08:11 | |
| *** Serlex has joined #openstack-kolla | 08:12 | |
| *** daneyon_ has joined #openstack-kolla | 08:19 | |
| *** rmart04 has joined #openstack-kolla | 08:20 | |
| *** daneyon_ has quit IRC | 08:24 | |
| *** jtriley has joined #openstack-kolla | 08:29 | |
| *** joyee has joined #openstack-kolla | 08:30 | |
| *** jtriley has quit IRC | 08:34 | |
| *** gfidente has joined #openstack-kolla | 08:37 | |
| *** gfidente has joined #openstack-kolla | 08:37 | |
| *** b_bezak_ has joined #openstack-kolla | 08:41 | |
| *** b_bezak has quit IRC | 08:41 | |
| *** joyee has quit IRC | 08:43 | |
| *** athomas has quit IRC | 08:58 | |
| *** b_bezak has joined #openstack-kolla | 09:01 | |
| *** athomas has joined #openstack-kolla | 09:03 | |
| *** b_bezak_ has quit IRC | 09:04 | |
| *** janmalax has quit IRC | 09:05 | |
| pbourke | morning can someone please approve https://review.openstack.org/#/c/353408/ | 09:14 |
| patchbot | pbourke: patch 353408 - kolla - Revert "Remove files from /var/lib/apt/lists when ... | 09:14 |
| *** egonzalez90 has joined #openstack-kolla | 09:16 | |
| *** ad_rien_ has joined #openstack-kolla | 09:19 | |
| *** ad_rien_ has quit IRC | 09:20 | |
| openstackgerrit | Eduardo Gonzalez proposed openstack/kolla: Customizations for Designate https://review.openstack.org/350172 | 09:30 |
| *** jtriley has joined #openstack-kolla | 09:30 | |
| *** jtriley has quit IRC | 09:35 | |
| duonghq | pbourke: I'm still running on Ubuntu 14.04, so I'm not pretty sure if Ubuntu 16.04 has completely moved to systemd yet? | 09:38 |
| pbourke | duonghq: it has, either way my suggestion is really to just follow the existing doc patterns. (in other places when we tell people to stop/start a service we give the relevant commands) | 09:39 |
| duonghq | pbourke: I'm updating according to your comment | 09:40 |
| pbourke | thanks! | 09:40 |
| duonghq | but I need to verify the service's name, it'll take a while | 09:40 |
| duonghq | I'm not sure if they change the service name | 09:41 |
| *** matrohon has quit IRC | 09:48 | |
| duonghq | pbourke: do we care about non-lts version? | 09:52 |
| pbourke | duonghq: no just 14.04 and 16.04 | 09:53 |
| duonghq | ok, thanks | 09:53 |
| openstackgerrit | Merged openstack/kolla: Document the use of external Ceph journal drives https://review.openstack.org/324539 | 09:54 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kolla: Updated from global requirements https://review.openstack.org/354303 | 09:56 |
| coolsvap | pbourke: approved | 10:01 |
| pbourke | coolsvap: thanks | 10:01 |
| pbourke | coolsvap: I got the list of rally tests trimmed down | 10:01 |
| pbourke | coolsvap: next we need to do the value tweaks I think? | 10:01 |
| coolsvap | pbourke: yes, sorry I could not help much yesterday | 10:02 |
| *** ad_rien_ has joined #openstack-kolla | 10:02 | |
| openstackgerrit | Merged openstack/kolla: Fix interface address for Neutron-server https://review.openstack.org/343117 | 10:02 |
| pbourke | coolsvap: no prob :) | 10:02 |
| pbourke | hope your hand is doing ok | 10:02 |
| pbourke | when you have a chance can you give me an example of the kind of changes we need to make again | 10:03 |
| *** zhurong has quit IRC | 10:04 | |
| coolsvap | pbourke: not much I have a meeting, will be back in an hour or so | 10:05 |
| pbourke | coolsvap: ok | 10:05 |
| openstackgerrit | Merged openstack/kolla: Revert "Remove files from /var/lib/apt/lists when cleaning up on Ubuntu/Debian" https://review.openstack.org/353408 | 10:08 |
| *** ad_rien_1 has joined #openstack-kolla | 10:09 | |
| *** ad_rien_ has quit IRC | 10:09 | |
| *** ad_rien_1 has quit IRC | 10:09 | |
| *** coolsvap is now known as _coolsvap_ | 10:13 | |
| openstackgerrit | Duong Ha-Quang proposed openstack/kolla: Fix LVM2 in Ubuntu section in Cinder guide https://review.openstack.org/353975 | 10:15 |
| duonghq | pbourke: I edited the patchset, I think it's better then 1st one | 10:17 |
| duonghq | pbourke: thanks | 10:20 |
| *** Serlex has quit IRC | 10:24 | |
| *** Serlex1 has joined #openstack-kolla | 10:24 | |
| *** jtriley has joined #openstack-kolla | 10:31 | |
| *** jtriley has quit IRC | 10:36 | |
| *** duonghq has quit IRC | 10:43 | |
| NachoDuck | For swift multinode do you just need to run the ring setup on each node? (http://docs.openstack.org/developer/kolla/swift-guide.html#rings) | 10:49 |
| Daviey | pbourke: around? | 10:49 |
| pbourke | Daviey: hi | 10:49 |
| Daviey | pbourke: Hey, so last night i tried recreating the routable network with both dhcp explicitly enabled and declared external routing... but this didn't help | 10:50 |
| Daviey | pbourke: Not sure if you saw, but i had this identical issue with vlans and lack of dhcp when using ovirt locally. I moved to linuxbridges purely because of this | 10:50 |
| Daviey | pbourke: I never solved it :/ | 10:51 |
| Daviey | err, not ovirt.. i mean ovs | 10:51 |
| pbourke | NachoDuck: no, you generate the rings on your deploy node. For multinode though you need to add each node to the ring | 10:51 |
| egonzalez90 | duonghq you can unify systemd command into systemctl stop open-iscsi iscsid | 10:52 |
| pbourke | Daviey: can you give me a more general overview of what happened? Last I left it inc0 and others were trying to create an external network for the VMs using one of the VLANs | 10:52 |
| NachoDuck | pbourke: Ah right I see, I'll give that a go | 10:52 |
| pbourke | NachoDuck: word of warning, swift in kolla is not heavily tested | 10:53 |
| pbourke | NachoDuck: it should work, just some potentially some rough edges | 10:53 |
| Daviey | pbourke: exactly that... Sorry, i missrememebred - thought you were part of the conversation. but we aren't getting dhcp responses. | 10:54 |
| Daviey | pbourke: If i'm honest, it is a bit embarrassing if we can't end the week with workable instances. | 10:54 |
| Daviey | pbourke: Do you want to take a look at getting OVS working? Otherwise i think we need to move to linuxbridges | 10:55 |
| pbourke | Daviey: hmm yeah. though it sounds like an openstack thing rather than a kolla thing? | 10:55 |
| Daviey | pbourke: yeah | 10:55 |
| NachoDuck | pbourke: Ok thanks, might see about radosgw instead | 10:55 |
| pbourke | Daviey: happy to help, you would need to drive though. my networking knowledge is improving but this seems tricky | 10:56 |
| Daviey | pbourke: Honestly.. i'm stuck aswell. did some tcpdumps, but that just provded what we expected | 10:57 |
| pbourke | ok, we can request some help via the mailing list | 10:58 |
| Daviey | pbourke: In the short term, shall we just move to LB's? | 10:58 |
| pbourke | this type of networking is bound to have been done in osic before | 10:58 |
| Daviey | we can revist OVS later | 10:58 |
| pbourke | Daviey: yeah if you think it will work definitely | 10:58 |
| *** vbel has quit IRC | 11:00 | |
| *** vbel has joined #openstack-kolla | 11:00 | |
| *** daneyon_ has joined #openstack-kolla | 11:01 | |
| *** ad_rien_ has joined #openstack-kolla | 11:03 | |
| *** JoseMello has joined #openstack-kolla | 11:06 | |
| *** daneyon_ has quit IRC | 11:06 | |
| *** ad_rien_ has quit IRC | 11:08 | |
| sean-k-mooney | Daviey: hi the bifrost patches should be code complete yes. if it will deploy successfully is another matter :) | 11:09 |
| openstackgerrit | Shaun Smekel proposed openstack/kolla: Customizations for Tgtd https://review.openstack.org/354392 | 11:10 |
| *** egonzalez90 has quit IRC | 11:10 | |
| sean-k-mooney | Daviey: it should work but i noticed on my baremetal test last night it tried to deploy but the ironic python agent ramdisk had an internal error when i tried to deploy on physical servers. works fine with vms | 11:10 |
| sean-k-mooney | Daviey: my home servers are old so im not sure if the issue is related to the hardware or if there was a change in the ipa image that was downloaded. | 11:11 |
| sean-k-mooney | Daviey: looking into that today | 11:11 |
| *** zhugx has joined #openstack-kolla | 11:13 | |
| openstackgerrit | Merged openstack/kolla: Updated from global requirements https://review.openstack.org/354303 | 11:14 |
| Daviey | sean-k-mooney: OK, i'm keen to give it a spin | 11:15 |
| Daviey | sean-k-mooney: Does ironic have an inbuilt option for the dhcp server to only respond to pxe dhcp requests? | 11:17 |
| *** b_bezak_ has joined #openstack-kolla | 11:18 | |
| openstackgerrit | Shaun Smekel proposed openstack/kolla: Handle empty package list for install_packages https://review.openstack.org/354429 | 11:18 |
| *** b_bezak has quit IRC | 11:18 | |
| *** dwalsh has joined #openstack-kolla | 11:19 | |
| sean-k-mooney | Daviey: i think so but that may only work when you are using neutron | 11:23 |
| Daviey | right | 11:24 |
| sean-k-mooney | Daviey: i have not had an issue with it conflicting on my home network but i also moved it to a seperate vlan a while ago but initlally it did not seam to conflict when on the same vlan as my main home gateway | 11:24 |
| Daviey | sean-k-mooney: So does that mean it wont work with bifrost? | 11:24 |
| sean-k-mooney | Daviey: bifrost wont use neutron correct. | 11:26 |
| sean-k-mooney | Daviey: im not sure if that will case dhcp conflicts though | 11:26 |
| Daviey | sean-k-mooney: So at home, are you using a dedicated provisioning vlan? Is that tagged at switch or on the client? I mean, unless you are using ipxe - how are you tagging? | 11:26 |
| *** egonzalez90 has joined #openstack-kolla | 11:27 | |
| *** egonzalez90 has quit IRC | 11:27 | |
| sean-k-mooney | Daviey: at home the server i deploy from is a tagged member of the vlan and the pxe boot interfaces on the target servers are untagged member of the vlan | 11:27 |
| Daviey | sean-k-mooney: so the tagging is transparent to the server? | 11:28 |
| sean-k-mooney | so the switch handels stripping/pushing the vlan for the target machicnes and i created a vlan subport on the server that is runing ironic | 11:28 |
| Daviey | sean-k-mooney: so you have a dedicated nic for provisioning ? | 11:28 |
| sean-k-mooney | on the target servers yes though if your nics support vlans for pxe you could share one. on the server that is running bifrost i have only one interface | 11:29 |
| sean-k-mooney | Daviey: i could also make this work by using the bifrost host server as a router between the two vlans but i was using one interface for provisioning and the other for internet/management | 11:31 |
| *** jtriley has joined #openstack-kolla | 11:32 | |
| sean-k-mooney | Daviey: if the bifrost server was setup to route between the subnets/vlan only one interface would be need one each system | 11:32 |
| *** kragniz is now known as Guest66666 | 11:36 | |
| *** jtriley has quit IRC | 11:37 | |
| Daviey | sean-k-mooney: I've only got HP servers handy... and AIUI, they don't support vlan PXE. I can force this with ipxe, but this adds some extra complexity | 11:37 |
| Daviey | sean-k-mooney: either i need a seperate DHCP/tftp stack to chainload into ipxe, getting into bifrost.. OR, i could look at the ironic support for uploading an ipxe ISO | 11:38 |
| Daviey | hmm.. | 11:38 |
| Daviey | maybe chainloading is the way to go... this would mean that my static DHCP/tftp i can easily make only respond to pxe requests.. and then safely enter the vlan only in this situation, getting in bifrost that way | 11:39 |
| sean-k-mooney | Daviey: oh well bifrost does replay your pxe rom with an ipxe rom if your servers dont support ipxe | 11:44 |
| sean-k-mooney | Daviey: then only reason i set up the vlan segmentation was that it would reliably get its ip from bifrost to get the ipxe image then relaibly get the ip within the ipxe rom from my main dhcp server and never pull down the ipa ramdisk | 11:46 |
| sean-k-mooney | so it was actully my main dhcp server that was causeing the issues for me not the bifrost one | 11:47 |
| sean-k-mooney | im using hp servers too at home but they are gen6 | 11:47 |
| Daviey | sean-k-mooney: I think i have Gen8 | 11:52 |
| Daviey | sean-k-mooney: when you say replace, you mean via ipmi/iLO/etc ? | 11:52 |
| sean-k-mooney | they way the dnsmas conf is set up if it get a pxe request it respond with a ipxe ramdisk first. if it get an ipxe request it then load the ipa ramdisk | 11:54 |
| sean-k-mooney | Daviey: this is what the dnsmasq.conf looks like | 11:55 |
| sean-k-mooney | http://paste.openstack.org/show/555964/ | 11:55 |
| Daviey | sean-k-mooney: Ah yeah, but the reason i am talking about ipxe is because i can use it to join the vlan at the host level. | 11:56 |
| sean-k-mooney | Daviey: if you have a managed switch that is the simplest way to do it | 12:02 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Neutron bootstrap to create service/project/role https://review.openstack.org/353740 | 12:11 |
| *** rhallisey has joined #openstack-kolla | 12:15 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Script generating kubernetes secrets from passwords.yml https://review.openstack.org/354199 | 12:16 |
| *** srwilkers has joined #openstack-kolla | 12:20 | |
| sbezverk | morning gents, do you know what is wrong with gates? it keeps failing everything.. | 12:25 |
| *** zhubingbing__ has quit IRC | 12:29 | |
| *** zhugx has quit IRC | 12:30 | |
| *** jtriley has joined #openstack-kolla | 12:31 | |
| *** senk_ has quit IRC | 12:35 | |
| *** jtriley has quit IRC | 12:36 | |
| pbourke | sbezverk: it kind of looks like the ubuntu mirrors are down | 12:41 |
| pbourke | i wonder are other projects failing | 12:41 |
| *** rkrum has joined #openstack-kolla | 12:45 | |
| *** daneyon_ has joined #openstack-kolla | 12:50 | |
| *** zhurong has joined #openstack-kolla | 12:50 | |
| *** rkrum has quit IRC | 12:50 | |
| *** daneyon_ has quit IRC | 12:55 | |
| *** dwalsh has quit IRC | 12:56 | |
| sean-k-mooney | _coolsvap_: if you time later you have commited change to ciwatch before. | 12:57 |
| sean-k-mooney | _coolsvap_: think we could add kolla | 12:57 |
| *** dave-mccowan has joined #openstack-kolla | 12:58 | |
| *** duonghq has joined #openstack-kolla | 12:59 | |
| *** senk_ has joined #openstack-kolla | 13:00 | |
| sean-k-mooney | _coolsvap_: kolla dose not currently have thirdparty ci but it would be nice to view the logs the kolla jenkins jobs | 13:00 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Neutron bootstrap to create service/project/role https://review.openstack.org/353740 | 13:06 |
| *** jtriley has joined #openstack-kolla | 13:06 | |
| Daviey | sbezverk / pbourke: Could it be that apt change that someone reverted due to stuff like this? | 13:06 |
| Daviey | someone remove /var/lib/apt/lists or something | 13:07 |
| sbezverk | pbourke: thanks, but it seems it is more than that.. | 13:07 |
| *** mhayden has joined #openstack-kolla | 13:09 | |
| *** fragatina has joined #openstack-kolla | 13:13 | |
| *** sdake has joined #openstack-kolla | 13:14 | |
| sdake | yo | 13:18 |
| *** fragatina has quit IRC | 13:18 | |
| Daviey | sdake: mornin' | 13:18 |
| duonghq | morning | 13:21 |
| *** ppowell has joined #openstack-kolla | 13:23 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Neutron bootstrap to create service/project/role https://review.openstack.org/353740 | 13:28 |
| sdake | whats going on with osic | 13:32 |
| sdake | ws the networking fixed? | 13:32 |
| sdake | pbourke ? | 13:32 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Neutron bootstrap to create service/project/role https://review.openstack.org/353740 | 13:34 |
| pbourke | sdake: afaik we're stuggling with the networking | 13:37 |
| pbourke | sdake: Daviey did a little more with it this morning so will get the latest from him when he's back | 13:38 |
| sdake | did you try liuxbridge? | 13:38 |
| pbourke | that was the next step, I think Daviey was going to try that | 13:40 |
| pbourke | but I dont know how far he got | 13:40 |
| sdake | soundsgood | 13:41 |
| sdake | poer about to go | 13:41 |
| *** b_bezak_ has quit IRC | 13:41 | |
| sdake | i'll be out this morning - dr appointment | 13:41 |
| Daviey | here | 13:42 |
| Daviey | Okay... so.. I deployed with linuxbridge | 13:42 |
| Daviey | but neutron doesn't seem to be responding | 13:42 |
| Daviey | can someone else try running init-runonce and tell me if i am being daft? | 13:42 |
| sdake | init-runonce is not what is needed | 13:43 |
| sdake | a special netowrk i sneeded | 13:43 |
| Daviey | sdake: no.. but that hasn't been run on this new cloud | 13:43 |
| Daviey | sdake: that isn't the issue.. neutron api isn't responding | 13:43 |
| sdake | i see | 13:43 |
| sdake | maybe we ought to try a stable branch... | 13:44 |
| *** b_bezak has joined #openstack-kolla | 13:44 | |
| sdake | using master is a mvoing target of a whole bunch of components of openstack | 13:44 |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: WIP Script generating kubernetes secrets from passwords.yml https://review.openstack.org/354199 | 13:45 |
| Daviey | sdake: meh, master is working OK for me locally | 13:45 |
| Daviey | sdake: I did want to update to latest master, but wasn' sure if we were pinning on a version | 13:45 |
| sdake | pinning | 13:45 |
| sdake | its not masster of kolla i am concerned with | 13:46 |
| sdake | its master of openstack services themselves | 13:46 |
| *** sdake has quit IRC | 13:46 | |
| Daviey | pbourke: do you want to have a play? | 13:47 |
| pbourke | Daviey: i'll have a quick look | 13:48 |
| *** senk_ has quit IRC | 13:49 | |
| pbourke | Daviey: mind if I clean up and redeploy? | 13:49 |
| Daviey | pbourke: wait | 13:50 |
| Daviey | pbourke: It is a fresh deployment | 13:50 |
| Daviey | pbourke: no commands, have been run - even init-runonce hasn't | 13:50 |
| Daviey | I've done the deploy 3 times, same result | 13:50 |
| pbourke | NetworkVlanRangeError: Invalid network VLAN range: '732:8000' - '8000 is not a valid VLAN tag' | 13:51 |
| Daviey | hmm | 13:53 |
| Daviey | pbourke: are you not on tmux? | 13:53 |
| pbourke | Daviey: i am | 13:53 |
| *** zhurong has quit IRC | 13:53 | |
| pbourke | Daviey: ah im on the wrong one | 13:53 |
| openstackgerrit | Larry Rensing proposed openstack/kolla: Add support for building vmtp container https://review.openstack.org/354362 | 13:53 |
| sean-k-mooney | pbourke: well that would make sense vlan can only go to 4096 correct | 13:53 |
| *** sdake has joined #openstack-kolla | 13:56 | |
| pbourke | Daviey: its prob left over from whatever the guys were trying yesterday | 13:56 |
| sdake | Daviey sorry power lost | 13:57 |
| sdake | i've got about 1 hr left before i have to depart for a few hours | 13:57 |
| pbourke | Daviey: so that's why neutron-server is failing but unfortunately I simply don't know enough on how achieve what we want | 13:58 |
| sdake | why is neutron-server failing? | 13:58 |
| pbourke | sdake: bad augment file left over from whatever the guys were trying yesterday | 13:59 |
| sdake | ok well lets un-augment | 13:59 |
| sdake | and try just linuxbridge | 13:59 |
| sdake | can someone paste the augment file? | 13:59 |
| pbourke | can we summarise the issue | 13:59 |
| sdake | maybe it has to do with setup of the network | 13:59 |
| sdake | sure | 13:59 |
| *** diogogmt has joined #openstack-kolla | 13:59 | |
| Daviey | pbourke: ok, i changed it to 800 and redeployed | 14:00 |
| Daviey | lets see if that works | 14:00 |
| Daviey | sdake: I am using LB's now | 14:00 |
| sdake | LBs? | 14:01 |
| sdake | load balancers? | 14:01 |
| Daviey | linuxbridge | 14:01 |
| sdake | sorry just woke up | 14:01 |
| sdake | brain booting | 14:01 |
| Daviey | have cawfee | 14:01 |
| sdake | in my cae pepsi | 14:01 |
| sdake | but ya same thing more or less | 14:01 |
| duonghq | Daviey: (just wonder), why you need linuxbridge? | 14:01 |
| sdake | takes a little bit for the sugar and caffeine to hit the system | 14:02 |
| sdake | duonghq because ovs may not be functional | 14:02 |
| Daviey | duonghq: Yesterday was burnt out trying to get access to isntance (and DHCP).. and it wasn't working | 14:02 |
| Daviey | about 4 of us tried, includng looking at packet captures | 14:02 |
| Daviey | to compare, we are trying the same config with linuxbridges | 14:02 |
| duonghq | okay | 14:02 |
| Daviey | (and sort-of-related, i am the identical problem locally) | 14:02 |
| sdake | am/have? | 14:03 |
| duonghq | How many node do you deploy to? | 14:03 |
| sdake | 100+ | 14:03 |
| Daviey | sdake: I am the problem, it seems | 14:03 |
| Daviey | pass me that coffee | 14:03 |
| sdake | my wife was reading some vp's bio yesterday | 14:03 |
| sdake | and at the end she said | 14:03 |
| sdake | "and I buy he highest quality crack cocaine on the planet." | 14:04 |
| sdake | funny woman :) | 14:04 |
| Daviey | Oh, also - i added a stanza to ceph_label to also unmount them if they are still mounted (hit me earlier) | 14:04 |
| Daviey | and a deploy.sh, which i think would be handy for us keeping all commands required to bring it up (including neutron commands) | 14:04 |
| Daviey | sdake: hah | 14:04 |
| sdake | she said it way better then that | 14:05 |
| duonghq | Daviey: You are using master, aren't you? | 14:05 |
| sdake | it was totally natural, like it was part of his bio | 14:05 |
| sdake | i dont know how she comes up with that stuff | 14:05 |
| *** dmueller-evoila has quit IRC | 14:05 | |
| duonghq | I really like coffee's maked from moka pot | 14:06 |
| sdake | not famliar with moka pot | 14:06 |
| sdake | duonghq i've had coffee like that in china and a few other places before | 14:07 |
| duonghq | so, how do you make you cup? | 14:07 |
| sdake | i dont consume coffee | 14:07 |
| sdake | but in the us, we use some machine that has little capsules | 14:07 |
| duonghq | sdake: afaik, moka pot is from Italy | 14:07 |
| sdake | maybe it was europe where i had it | 14:08 |
| duonghq | I try us-style coffee | 14:08 |
| sdake | i've been all over the world | 14:08 |
| sdake | but not itally | 14:08 |
| duonghq | just different flavor | 14:08 |
| sdake | us coffee is dissappointing | 14:08 |
| sdake | mass market junk | 14:08 |
| duonghq | haha | 14:09 |
| *** b_bezak has quit IRC | 14:09 | |
| Daviey | duonghq: master, yes | 14:09 |
| *** b_bezak has joined #openstack-kolla | 14:10 | |
| *** dwalsh has joined #openstack-kolla | 14:10 | |
| sdake | i've got one of these on order http://www.ftknox.com/vaults/the-defender-vault/ with corner vaults | 14:11 |
| sdake | raither corner thingies | 14:11 |
| sdake | if someone managed to oepn that up, they would deserve whats inside | 14:12 |
| duonghq | interestin | 14:13 |
| vbel | best defense is insurance :) | 14:15 |
| *** ad_rien_ has joined #openstack-kolla | 14:15 | |
| pbourke | or a gun | 14:16 |
| pbourke | :p | 14:16 |
| *** _coolsvap_ is now known as coolsvap | 14:16 | |
| Daviey | sdake: You make me feel normal :) | 14:17 |
| Daviey | Chaps, if LB's works (which i am expecting it to)... we need to get some help from neutron people to work out why OVS isn't | 14:18 |
| Daviey | If this is the identical problem i am seeing locally, it is a release issue for us | 14:19 |
| pbourke | Daviey: would be really great if you could summarise the problem on the ML | 14:19 |
| sdake | if ovs is busted in master, then we need to revert back to a stable branch | 14:19 |
| pbourke | Daviey: its hard to keep up with this stuff and that's coming from someone who's keeping an eye on it | 14:19 |
| sdake | if linux bridge works, next steps is to revert to 2.0.2 | 14:19 |
| pbourke | there are probably people here who know the answer but just arent watching IRC 24/7 | 14:20 |
| sdake | it is possible ovs is just broken entirely in master | 14:20 |
| Daviey | sdake: Why not continue with master using LB's until we solve the issue? | 14:20 |
| sdake | well fwiw, I am not sure why we are testing a moving target of openstck | 14:21 |
| sdake | whereas saying "here is the characterization of mitaka" | 14:21 |
| Daviey | sdake: because we can fix master :) | 14:21 |
| sdake | i agree master should be fixed | 14:21 |
| sean-k-mooney | sdake: isnt the cluster ment to be deploying stable/mitaka ? | 14:21 |
| sdake | but the point of osic scale testing is to test scale | 14:21 |
| Daviey | where as, most users of stable/* probably have what they need already | 14:21 |
| Daviey | sdake: FWIW, i've been tracking master for the last 2 months.. it isn't moving that fast | 14:22 |
| sdake | you mean master of kolla, or master of openstack? | 14:22 |
| *** britthouser has joined #openstack-kolla | 14:22 | |
| sdake | i am not concerned with master of kolla | 14:22 |
| Daviey | sdake: both :) | 14:22 |
| sdake | concerned with master of openstack | 14:23 |
| sdake | our gate tests dont actually test sshing int othe vm | 14:23 |
| Daviey | sdake: yeah... rest of the openstack ecosystem looks reasonable | 14:23 |
| sdake | so we have no validation anywhere that ovs works or doesn't work | 14:23 |
| sean-k-mooney | sdake: well ovs should work on stable/mitaka | 14:24 |
| sean-k-mooney | sdake: im using it now | 14:24 |
| sdake | precisely my point - we need to move o nwith testing now, record an ovs bug, and finish the job on scale testing | 14:24 |
| Daviey | neutron api is workig now | 14:24 |
| sdake | sean-k-mooney yes i know for certain stable works - not sure about master | 14:25 |
| *** b_bezak has quit IRC | 14:25 | |
| pbourke | Daviey: do you know how to create the networks? | 14:26 |
| *** williamc_ has joined #openstack-kolla | 14:26 | |
| Daviey | pbourke: yeah, lets create the default ones first | 14:26 |
| *** williamcaban has quit IRC | 14:27 | |
| sdake | what was broken with neutron api not working? | 14:28 |
| sdake | misconfiguration? | 14:28 |
| pbourke | sdake: yes | 14:28 |
| sdake | out of curiosity what was misconfigured | 14:29 |
| *** zhubingbing has joined #openstack-kolla | 14:29 | |
| sdake | hey zhubingbing | 14:29 |
| zhubingbing | hey | 14:29 |
| zhubingbing | hi | 14:29 |
| sdake | hey britthouser - back from vacaton? | 14:29 |
| britthouser | lol....not yet. In SJC office, for just a couple hours this morning then heading to airport | 14:30 |
| Daviey | grrrr | 14:31 |
| Daviey | still busted | 14:31 |
| sdake | Daviey neutron net-show -> paste | 14:31 |
| pbourke | Daviey: do we have a route to 172.21.116.0/22 | 14:32 |
| Daviey | i need to take a quick call | 14:32 |
| Daviey | brb | 14:32 |
| *** ad_rien_ has quit IRC | 14:33 | |
| *** duonghq has quit IRC | 14:36 | |
| *** daneyon_ has joined #openstack-kolla | 14:38 | |
| *** senk_ has joined #openstack-kolla | 14:39 | |
| *** dwalsh has quit IRC | 14:41 | |
| *** daneyon_ has quit IRC | 14:43 | |
| sdake | enjoy: https://www.youtube.com/watch?v=myyWXKeBsNk | 14:44 |
| sdake | al pacino was fantastic in that movie | 14:47 |
| sdake | sdake->shower | 14:47 |
| *** michauds has joined #openstack-kolla | 14:50 | |
| zhubingbing | I owe the AWS machine, the machine will start paying money - -） | 14:51 |
| sbezverk | zhubingbing how can you own aws machine? just curious.. | 14:52 |
| *** huikang has joined #openstack-kolla | 14:52 | |
| *** dwalsh has joined #openstack-kolla | 14:54 | |
| zhubingbing | okay，AWS cloud computing platform, you do not have to do？ | 14:54 |
| *** srwilkers has quit IRC | 14:55 | |
| *** srwilkers has joined #openstack-kolla | 14:59 | |
| openstackgerrit | zhubingbing proposed openstack/kolla: repair gnocchi dockerfile https://review.openstack.org/348388 | 15:00 |
| *** michauds has quit IRC | 15:03 | |
| *** britthouser has quit IRC | 15:06 | |
| openstackgerrit | zhubingbing proposed openstack/kolla: repair aodh image dockerfile https://review.openstack.org/350715 | 15:06 |
| *** ayoung has joined #openstack-kolla | 15:10 | |
| openstackgerrit | zhubingbing proposed openstack/kolla: Add aodh role https://review.openstack.org/351027 | 15:11 |
| *** britthouser has joined #openstack-kolla | 15:12 | |
| *** rmart04 has quit IRC | 15:13 | |
| *** senk_ has quit IRC | 15:15 | |
| *** huikang_ has joined #openstack-kolla | 15:15 | |
| *** michauds has joined #openstack-kolla | 15:18 | |
| *** mliima has joined #openstack-kolla | 15:19 | |
| *** haplo37_ has joined #openstack-kolla | 15:30 | |
| openstackgerrit | Merged openstack/kolla: Customizations for OpenStack Base https://review.openstack.org/354416 | 15:32 |
| openstackgerrit | Merged openstack/kolla: Fix LVM2 in Ubuntu section in Cinder guide https://review.openstack.org/353975 | 15:32 |
| *** ayoung has quit IRC | 15:33 | |
| *** wirehead_ has quit IRC | 15:36 | |
| openstackgerrit | Merged openstack/kolla: Customizations for Designate https://review.openstack.org/350172 | 15:38 |
| openstackgerrit | Merged openstack/kolla: Customizations for Kuryr https://review.openstack.org/354424 | 15:38 |
| openstackgerrit | Merged openstack/kolla: Fix prechecks of a vip address https://review.openstack.org/353214 | 15:38 |
| openstackgerrit | Merged openstack/kolla: Fix prechecks of a vip address https://review.openstack.org/353221 | 15:38 |
| *** wirehead_ has joined #openstack-kolla | 15:38 | |
| *** sdake_ has joined #openstack-kolla | 15:39 | |
| *** mark-casey has joined #openstack-kolla | 15:41 | |
| sdake_ | shower->sdake | 15:41 |
| *** sdake has quit IRC | 15:41 | |
| sdake_ | hey folks | 15:41 |
| sdake_ | any luck with osic running? | 15:42 |
| openstackgerrit | Merged openstack/kolla: Keystone interface address and memcached override https://review.openstack.org/328569 | 15:45 |
| openstackgerrit | Merged openstack/kolla: Customizations for Glance https://review.openstack.org/352977 | 15:46 |
| *** michauds has quit IRC | 15:46 | |
| openstackgerrit | Mauricio Lima proposed openstack/kolla: Customizations for Manila https://review.openstack.org/350601 | 15:47 |
| openstackgerrit | zhubingbing proposed openstack/kolla: repair aodh image dockerfile https://review.openstack.org/350715 | 15:50 |
| *** duonghq has joined #openstack-kolla | 15:50 | |
| *** senk_ has joined #openstack-kolla | 15:51 | |
| *** senk_ has quit IRC | 15:51 | |
| *** jtriley has quit IRC | 15:54 | |
| openstackgerrit | zhubingbing proposed openstack/kolla: Add aodh role https://review.openstack.org/351027 | 15:54 |
| *** senk_ has joined #openstack-kolla | 15:54 | |
| *** lrensing_ has joined #openstack-kolla | 15:55 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Glance bootstrap to create service/project/role https://review.openstack.org/346215 | 15:56 |
| *** sdake_ has quit IRC | 15:59 | |
| *** haplo37_ has quit IRC | 16:00 | |
| *** michauds has joined #openstack-kolla | 16:02 | |
| *** Serlex1 has quit IRC | 16:03 | |
| *** huikang_ has quit IRC | 16:03 | |
| *** duonghq has left #openstack-kolla | 16:17 | |
| *** mliima has quit IRC | 16:22 | |
| *** daneyon_ has joined #openstack-kolla | 16:27 | |
| *** ad_rien_ has joined #openstack-kolla | 16:29 | |
| *** fragatina has joined #openstack-kolla | 16:30 | |
| *** daneyon_ has quit IRC | 16:31 | |
| *** fragatin_ has joined #openstack-kolla | 16:31 | |
| *** dwalsh has quit IRC | 16:32 | |
| *** athomas has quit IRC | 16:32 | |
| *** fragatina has quit IRC | 16:35 | |
| *** ad_rien_ has quit IRC | 16:42 | |
| *** lrensing_ has quit IRC | 16:44 | |
| *** senk_ has quit IRC | 16:49 | |
| *** mark-casey has quit IRC | 16:50 | |
| *** senk_ has joined #openstack-kolla | 16:52 | |
| *** unicell has quit IRC | 16:55 | |
| *** unicell has joined #openstack-kolla | 16:56 | |
| *** britthouser has quit IRC | 16:57 | |
| *** srwilkers has quit IRC | 16:59 | |
| *** huikang_ has joined #openstack-kolla | 17:04 | |
| *** huikang_ has quit IRC | 17:09 | |
| *** mliima has joined #openstack-kolla | 17:11 | |
| *** mark-casey has joined #openstack-kolla | 17:12 | |
| *** ayoung has joined #openstack-kolla | 17:13 | |
| *** coolsvap has quit IRC | 17:21 | |
| *** ntpttr__ has joined #openstack-kolla | 17:23 | |
| *** ppowell has quit IRC | 17:24 | |
| *** ad_rien_ has joined #openstack-kolla | 17:24 | |
| *** huikang_ has joined #openstack-kolla | 17:29 | |
| *** ad_rien_ has quit IRC | 17:32 | |
| ntpttr__ | hey kolla folks, does kolla provide a way to install tempest and configure it to run tests, or does that have to be done manually after you have a working deployment? | 17:32 |
| kfox1111 | so, how do we handle configmaps in kolla-kubernetes that probably shouldn't be generated by kolla? | 17:33 |
| *** ad_rien_ has joined #openstack-kolla | 17:33 | |
| rhallisey | kfox1111, do you have an example? | 17:34 |
| kfox1111 | ok. so, I'm playing around with deployments and safe shutdown of glance-api (should work the same way for all openstack api servers though) | 17:35 |
| kfox1111 | the problem is, there is no safe way to drain one out without something in front. | 17:35 |
| kfox1111 | also, there is no relyable way to do ssl on all of them. | 17:35 |
| kfox1111 | so to kill two birds with one stone, I want change the api_listen they do to 127.0.0.1, and add an haproxy sidecar that listens on 0.0.0.0 and passes traffic to 127.0.0.1. | 17:36 |
| kfox1111 | haproxy can do a safe shutdown with a lifecycle hook, keeping thinkgs alive until all connections terminate. | 17:36 |
| kfox1111 | it also can ssl terminate if we want. | 17:36 |
| kfox1111 | but that means the haproxy config is per api server. | 17:37 |
| kfox1111 | I think I can add it as another templated pod in the pod section? | 17:37 |
| rhallisey | yes | 17:37 |
| rhallisey | interersting idea | 17:38 |
| rhallisey | haproxy dameon set | 17:38 |
| rhallisey | or maybe just side car | 17:38 |
| rhallisey | idk | 17:38 |
| kfox1111 | yeah. I'm thinking sidecar, as they then share 127.0.0.1. | 17:38 |
| kfox1111 | otherwise, someone could access the api directly and safe shutdown wouldn't see it. | 17:39 |
| rhallisey | that's true | 17:39 |
| kfox1111 | I'll poste a WPI POC soonish. | 17:39 |
| *** michauds has quit IRC | 17:39 | |
| rhallisey | kfox1111, https://blueprints.launchpad.net/kolla-kubernetes/+spec/api-termination | 17:40 |
| kfox1111 | hmm.... its wanting a 'container' section instead of a template one for glance. | 17:41 |
| kfox1111 | ah. cool. | 17:41 |
| kfox1111 | can you have both templates and a containers: section in service_resources.yml? or are they exclusive? | 17:41 |
| kfox1111 | oh. nm. I put it in the wrong place. | 17:42 |
| *** huikang_ has quit IRC | 17:43 | |
| *** vhosakot has joined #openstack-kolla | 17:45 | |
| kfox1111 | ok. that did generate a correctish looking configmap. | 17:45 |
| kfox1111 | I think this will all create correctly, but I think cleanup might fail? Can you tag on an object type to things in service_resources.yml? | 17:46 |
| *** michauds has joined #openstack-kolla | 17:51 | |
| *** gfidente has quit IRC | 17:52 | |
| *** mliima has quit IRC | 17:56 | |
| *** ad_rien_ has quit IRC | 17:57 | |
| *** fragatin_ has quit IRC | 18:02 | |
| kfox1111 | rhallisey: this raises the same problem again though.... in a glance template, there's no haproxy image var defined. | 18:04 |
| kfox1111 | can we just load all the image names always? | 18:05 |
| kfox1111 | or at least consider haproxy to be a rather unique case and always load that one too? | 18:05 |
| sbezverk | kfox1111: this approach breaks the concept of having a common image for kolla ansible and kolla kubernetes.. if I understand correctly wha you asked.. | 18:08 |
| kfox1111 | its a common image. | 18:09 |
| kfox1111 | it breaks kolla-ansible's idea of there only being one haproxy though. | 18:09 |
| kfox1111 | in kolla-ansible a service only has one image. | 18:09 |
| sbezverk | kfox1111 I do not think it is problem | 18:09 |
| kfox1111 | this would have one service with two images. | 18:09 |
| sbezverk | you can probably get around it in kube templates | 18:10 |
| kfox1111 | yeah, I could hard code it easily enough. but as its very flexible with <distro>-<src/binary>:<tag> | 18:11 |
| kfox1111 | it would be nice if we could reuse it somehow. | 18:11 |
| kfox1111 | the variable definition that is. | 18:11 |
| kfox1111 | its defined here: grep haproxy_image_full ../kolla/ansible/roles/haproxy/defaults/main.yml | 18:12 |
| kfox1111 | haproxy_image_full: "{{ haproxy_image }}:{{ haproxy_tag }}" | 18:12 |
| kfox1111 | maybe just always loading that additional role file would work? | 18:12 |
| kfox1111 | yup. that works. | 18:15 |
| kfox1111 | a bit of a hack. probably better to get haproxy's role bits to move to the common role. | 18:15 |
| kfox1111 | but for now, it works. | 18:15 |
| *** dwalsh has joined #openstack-kolla | 18:17 | |
| *** britthouser has joined #openstack-kolla | 18:18 | |
| *** britthou_ has joined #openstack-kolla | 18:19 | |
| openstackgerrit | Tin Lam proposed openstack/kolla: Add support for building vmtp container https://review.openstack.org/354362 | 18:20 |
| *** sdake has joined #openstack-kolla | 18:21 | |
| *** britthouser has quit IRC | 18:22 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP - Deployment, Safe Shutdown, & Scaling for glance-api https://review.openstack.org/354895 | 18:31 |
| *** senk_ has quit IRC | 18:34 | |
| kfox1111 | rhallisey: sbezverk: here's another question. we templated out common disk/pv/pvc's. which is great. | 18:39 |
| kfox1111 | I see two cases where they arn't needed though based on a condition. | 18:39 |
| kfox1111 | glance_backend_ceph == yes and cinder_backend_ceph == yes? | 18:40 |
| sdake | any uck on a running osic cluster? | 18:40 |
| sdake | luck | 18:40 |
| *** ad_rien_ has joined #openstack-kolla | 18:40 | |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP - Deployment, Safe Shutdown, & Scaling for glance-api https://review.openstack.org/354895 | 18:41 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP - Deployment, Safe Shutdown, & Scaling for glance-api https://review.openstack.org/354895 | 18:42 |
| openstackgerrit | Kevin Fox proposed openstack/kolla-kubernetes: WIP - Deployment, Safe Shutdown, & Scaling for glance-api https://review.openstack.org/354895 | 18:43 |
| rhallisey | kfox1111, not sure though because the volumes won't be exposed to kubernetes | 18:44 |
| *** ad_rien_ has quit IRC | 18:44 | |
| *** huikang_ has joined #openstack-kolla | 18:44 | |
| rhallisey | one thing though, that config file could be per micro service | 18:44 |
| rhallisey | if only 1/4 services needed persistent storage | 18:45 |
| rhallisey | ex) cinder-volume | 18:45 |
| rhallisey | or glance-registry | 18:45 |
| *** sdake has quit IRC | 18:45 | |
| rhallisey | kfox1111, but you could make kube aware | 18:46 |
| rhallisey | our current method doesnt' support that | 18:46 |
| sbezverk | kfox1111 I think if the servicedoes not require pv/pvc it does not go then insto service_resource.yml. | 18:46 |
| openstackgerrit | zhubingbing proposed openstack/kolla: Add trove role https://review.openstack.org/354901 | 18:47 |
| kfox1111 | sbezverk: the issue is, glance-api needs a volume if not storing to ceph. | 18:47 |
| kfox1111 | if its using rbd directly, it can scale out better without a volume. | 18:47 |
| kfox1111 | so the template for loading disks in for that service is dependent on how the service is configured. | 18:48 |
| kfox1111 | :/ | 18:48 |
| *** huikang_ has quit IRC | 18:48 | |
| kfox1111 | this may be another case where workflow probably should be driving, since it knows the config of glance... | 18:49 |
| rhallisey | ya | 18:49 |
| kfox1111 | it can skip running the disk create commands for glance_backend_ceph=yes | 18:49 |
| rhallisey | we need to split out config | 18:49 |
| kfox1111 | k. then I'll not worry about it. :) | 18:49 |
| rhallisey | we will :0 | 18:49 |
| rhallisey | :) | 18:49 |
| rhallisey | kfox1111, workflow is #1 priority after we nail down nova and neutron | 18:50 |
| kfox1111 | k. | 18:50 |
| sbezverk | rhallisey: speaking about neutron. please review bootstrap it is finished.. | 18:51 |
| sbezverk | kfox1111: any modification for neutron bootstar should be done after this PS gets merged.. | 18:51 |
| rhallisey | sure | 18:51 |
| kfox1111 | sbezverk: this one? https://review.openstack.org/#/c/353740 | 18:51 |
| patchbot | kfox1111: patch 353740 - kolla-kubernetes - Neutron bootstrap to create service/project/role | 18:51 |
| sbezverk | kfox1111: yep | 18:52 |
| rhallisey | I think the gate was suddenly made coting | 18:52 |
| rhallisey | voting.. | 18:52 |
| * rhallisey fixes that | 18:52 | |
| *** britthou_ has quit IRC | 18:54 | |
| kfox1111 | sbezverk: I'm a bit worried that method of passing openstack_neutron_auth through will be pretty fragile. | 18:54 |
| sbezverk | kfox1111 it is done that way with kolla | 18:55 |
| kfox1111 | it relies on passing json and shell vars through a template engine and a shell. | 18:55 |
| kfox1111 | sure. that doesn't mean its not fragile. :) | 18:56 |
| sbezverk | kfox1111: using alpha features or software is even more fragile | 18:56 |
| kfox1111 | what did you think of patch set #6? | 18:56 |
| kfox1111 | could be. could be. | 18:56 |
| sbezverk | kfox1111: I do not mind at all if you after it is done, you take over and add it | 18:57 |
| kfox1111 | I don't think init containers are going to be singificantly changed though. petsets require them too. | 18:57 |
| sbezverk | I just want to move forward | 18:57 |
| kfox1111 | k. but I'd kind of like it resolved before more things are added, like the glance bootstrap. | 18:57 |
| sbezverk | kfox1111: adding any bits in image, will take a lot of time as it will impact kolla as well | 18:58 |
| sbezverk | we need to show glance and neutron | 18:58 |
| kfox1111 | it would take a while to add a container with just one package in it into kolla? | 18:59 |
| sbezverk | I personally do not think at all its worth it as kolla uses the same approach and it works well | 18:59 |
| kfox1111 | kolla-ansible does things that I don't think fit well as an Op, and one reason I'm not using it. :/ | 19:00 |
| kfox1111 | I'm trying to fix them now, in kolla-kubernetes, so I can use it in a big deployment. | 19:00 |
| kfox1111 | so, don't consider what kolla-ansible did as cannon law. | 19:00 |
| *** fragatina has joined #openstack-kolla | 19:01 | |
| kfox1111 | for our production cloud, we have 3 mariadb's, 2 rabbit clusters, 3 different controller sub clusters, etc. I don't think it can be done with kolla-ansible. | 19:01 |
| kfox1111 | to get to our scale, it was nessisary to do it that way. :/ | 19:02 |
| *** dwalsh has quit IRC | 19:03 | |
| rhallisey | kfox1111 sbezverk what do you think can be an alternative? | 19:03 |
| kfox1111 | for which part? | 19:03 |
| rhallisey | how openstack_neutron_auth is moved | 19:03 |
| *** vhosakot has quit IRC | 19:04 | |
| kfox1111 | I guess I can live with it for now, but as an op, its really hard to understand. I'd recommend first thing after glance/nova/neutron/basic workflow, we fix it. | 19:04 |
| sbezverk | rhallisey: there is no any evidence that the way it is done now is really fragile, I do not know about any bugs or something like that.. | 19:05 |
| kfox1111 | but if we start pushing on kolla to get a jq container, it should be ready by then. | 19:05 |
| *** ntpttr__ has quit IRC | 19:05 | |
| *** ad_rien_ has joined #openstack-kolla | 19:05 | |
| rhallisey | sbezverk, I"m not saying it necessary is, but if there is a possibility to improve we should | 19:05 |
| sbezverk | kfox1111: it will take a lot of time, remember how long it took just to get simple patches on kolla side | 19:05 |
| kfox1111 | sbezverk: yeah, but that was touching a lot of plumbing. this is just a new container. | 19:06 |
| rhallisey | kfox1111, sbezverk that can happen after a kolla kolla-ansible split | 19:06 |
| sbezverk | and who can guarantee that jq is stable or when it gets stable?? | 19:06 |
| rhallisey | much easier at that point | 19:06 |
| kfox1111 | sbezverk: yeah, I agree it will work as is. it just makes me cringe as an op. other ops will feel the same. | 19:06 |
| kfox1111 | sbezverk: thats the power of containers. use the version that's stable. :) | 19:06 |
| rhallisey | kfox1111, ok hold on to that concern then. Let's address it when we can | 19:07 |
| kfox1111 | k. | 19:07 |
| rhallisey | sounds like it is going to be post repo split | 19:07 |
| kfox1111 | I can submit the jq container request. it won't be big. | 19:08 |
| kfox1111 | actually, a new container isn't really needed. jq could just be added to the existing kolla_toolbox. | 19:08 |
| *** dwalsh has joined #openstack-kolla | 19:09 | |
| *** daneyon_ has joined #openstack-kolla | 19:09 | |
| *** ad_rien_ has quit IRC | 19:10 | |
| openstackgerrit | zhubingbing proposed openstack/kolla: Add gnocchi ansible role https://review.openstack.org/349351 | 19:11 |
| openstackgerrit | Kevin Fox proposed openstack/kolla: Add jq https://review.openstack.org/355003 | 19:12 |
| kfox1111 | there. now we can see how long that takes. :) | 19:12 |
| *** daneyon_ has quit IRC | 19:13 | |
| kfox1111 | sbezverk: does this neutron bootstrap actually work? | 19:15 |
| kfox1111 | '$DATABASE_PASSWORD' is single quoted. | 19:16 |
| sbezverk | kfox1111: yes | 19:16 |
| kfox1111 | oh, but its passed as a single arg, to sh -c... so it will deal with the whole string as a shell chunk... | 19:17 |
| kfox1111 | and there is a " before it, so the ' will be ignored... | 19:17 |
| kfox1111 | but the " could be eaten by the yaml parser.... | 19:18 |
| *** senk_ has joined #openstack-kolla | 19:18 | |
| kfox1111 | I'm assuming its not, as its working... | 19:18 |
| kfox1111 | ok. I think this is good enough for now. we can revisit the rest later. | 19:23 |
| kfox1111 | I -workflow'ed it as I think the other patch needs to merge first though. | 19:23 |
| *** wirehead_ has quit IRC | 19:25 | |
| sbezverk | kfox1111: well, it is not clear where secret generator needs to go, it is easier to be a separate tool, but logically it sould be a part of kolla-kubernetes.. so I do not know.. | 19:25 |
| sbezverk | I will address all comments, but I am not sure it ever gets merged.. | 19:26 |
| *** senk_ has quit IRC | 19:28 | |
| *** fragatina has quit IRC | 19:28 | |
| *** wirehead_ has joined #openstack-kolla | 19:28 | |
| kfox1111 | yeah. it could be merged as another cli subcommand in kolla-kubernetes... hmm.. | 19:29 |
| kfox1111 | rhallisey: do you have any thoughts there? | 19:29 |
| *** fragatina has joined #openstack-kolla | 19:29 | |
| *** fragatina has quit IRC | 19:29 | |
| rhallisey | kfox1111, I was thinking it could sit in the q to make the neutron patch demoable | 19:29 |
| rhallisey | we need to rework our CLI | 19:29 |
| rhallisey | we need to make rest calls | 19:29 |
| kfox1111 | heh. yeah... | 19:30 |
| rhallisey | this could be added during that time | 19:30 |
| kfox1111 | well, if we don't want to merge the neutron patch right now, then it doesn't matter to wait. | 19:30 |
| kfox1111 | just looking at what it would take to get it merged. | 19:30 |
| kfox1111 | if we're going to have to redo the cli anyway, then I'd say merging it as standalone wouldn't be terible. | 19:31 |
| kfox1111 | we can then merge it into a good place when the cli's reworked. | 19:32 |
| kfox1111 | I do think the neutron patch should merge soonish. | 19:33 |
| *** daneyon has quit IRC | 19:34 | |
| rhallisey | kfox1111, true | 19:34 |
| rhallisey | wfm | 19:34 |
| sbezverk | rhallisey: just addressed your comments about the script. If you can point me to the doc how to make rest calls to kube, please let me know. | 19:34 |
| kfox1111 | k. | 19:34 |
| rhallisey | sbezverk, the rest calls are a super blocker. But if you want to do it | 19:35 |
| * rhallisey looks for a doc | 19:35 | |
| kfox1111 | wait. | 19:35 |
| sbezverk | rhallisey: it is easier to try it on something small like this | 19:35 |
| kfox1111 | rhallisey: I'm thinking the opposite. | 19:35 |
| rhallisey | kfox1111, sure | 19:35 |
| kfox1111 | I'd like to see '--dry-ryn' show me exactly the kubectl calls that should be made to reproduce what its going to do. | 19:35 |
| kfox1111 | run | 19:36 |
| kfox1111 | so having it exec has some advantages there. you can just stub out the exec with a print. | 19:36 |
| rhallisey | the problem though with exec is were limited to kubectl's info | 19:36 |
| rhallisey | so when it comes to querying the resources | 19:37 |
| kfox1111 | I really like the ability to ./secret-generator.py --dry-run > foo.sh, inspect foo.sh. then bash foo.sh when I'm satisfied it won't break anything. | 19:37 |
| rhallisey | let me compare. Last I looked kubectl was a not verbose enough | 19:37 |
| rhallisey | kfox1111, we can still accomplish it even with rest calls | 19:37 |
| sbezverk | rhallisey: we could do kubectl -o yaml | 19:38 |
| rhallisey | just translate the rest call | 19:38 |
| kfox1111 | true. it just makes me as an op more likely to believe it when I deside not to --dry-run as the code looks the same. | 19:38 |
| kfox1111 | more of a trust building thing. | 19:38 |
| rhallisey | kfox1111, :) | 19:38 |
| rhallisey | need to have some trust | 19:38 |
| kfox1111 | I'd say, use rest if there's a performance benifit or the cli just can't cut it. | 19:38 |
| kfox1111 | an ops job is to figure out all the ways something can bite you. :) | 19:39 |
| kfox1111 | the easier you make it on an op to figure that out, the more ops will like your stuff. :) | 19:39 |
| rhallisey | kfox1111, what it would come down to is if the cli can provide enough info for the workflow to get a proper pod status | 19:39 |
| kfox1111 | I'm much more ok with doing a rest call for getting status back. read only is rarely dangerious. :) | 19:40 |
| kfox1111 | its the changy things that I pay close attention to. | 19:40 |
| rhallisey | ok maybe the query is a rest call | 19:40 |
| rhallisey | and the others are cli | 19:40 |
| kfox1111 | that sounds good to me. :) | 19:41 |
| rhallisey | if we do rest everywhere though, it means an op gives an ip of kube and is good to go | 19:41 |
| rhallisey | well I guess it would be the same with kubectl | 19:41 |
| rhallisey | you just need to point it at a kube somewhere | 19:42 |
| kfox1111 | right. | 19:42 |
| kfox1111 | ok... meeting next. bbl. | 19:42 |
| rhallisey | sbezverk, ok hold off on the rest calls then | 19:42 |
| sbezverk | rhallisey: :-) | 19:42 |
| *** akwasnie has quit IRC | 19:43 | |
| *** akwasnie has joined #openstack-kolla | 19:45 | |
| openstackgerrit | Kabir Sial proposed openstack/kolla: Customizations for dind https://review.openstack.org/353203 | 19:49 |
| *** matrohon has joined #openstack-kolla | 19:49 | |
| *** huikang_ has joined #openstack-kolla | 19:51 | |
| *** huikang_ has quit IRC | 19:56 | |
| *** daneyon has joined #openstack-kolla | 19:56 | |
| *** huikang_ has joined #openstack-kolla | 20:10 | |
| *** inc0 has joined #openstack-kolla | 20:13 | |
| inc0 | hey | 20:13 |
| inc0 | wassup folks? | 20:13 |
| *** daneyon has quit IRC | 20:19 | |
| mark-casey | inc0: hey | 20:20 |
| *** zhubingbing has quit IRC | 20:20 | |
| *** huikang_ has quit IRC | 20:22 | |
| *** huikang_ has joined #openstack-kolla | 20:23 | |
| *** ntpttr__ has joined #openstack-kolla | 20:26 | |
| *** huikang_ has quit IRC | 20:28 | |
| *** huikang_ has joined #openstack-kolla | 20:28 | |
| *** huikang__ has joined #openstack-kolla | 20:29 | |
| *** sdake has joined #openstack-kolla | 20:32 | |
| *** shardy has quit IRC | 20:32 | |
| *** huikang_ has quit IRC | 20:33 | |
| *** ntpttr__ has quit IRC | 20:34 | |
| sdake | sup cats | 20:34 |
| *** huikang__ has quit IRC | 20:37 | |
| *** huikang has quit IRC | 20:38 | |
| *** huikang has joined #openstack-kolla | 20:39 | |
| *** haplo37_ has joined #openstack-kolla | 20:39 | |
| *** rhallisey has quit IRC | 20:43 | |
| *** huikang has quit IRC | 20:43 | |
| sbezverk | sdake could you fix kubernetes gates, they became voting all of a sudden | 20:45 |
| *** inc0 has quit IRC | 20:47 | |
| -openstackstatus- NOTICE: The Gerrit service on review.openstack.org is restarting for a scheduled upgrade, but should return to service momentarily: http://lists.openstack.org/pipermail/openstack-dev/2016-August/101394.html | 20:49 | |
| sdake | wtf | 20:50 |
| sdake | sbezverk link me a voted gate review please | 20:50 |
| sbezverk | sdake: https://review.openstack.org/#/c/346215/ | 20:54 |
| patchbot | sbezverk: patch 346215 - kolla-kubernetes - Glance bootstrap to create service/project/role | 20:54 |
| *** daneyon has joined #openstack-kolla | 20:55 | |
| *** daneyon_ has joined #openstack-kolla | 20:57 | |
| -openstackstatus- NOTICE: The Mediawiki service at wiki.openstack.org will be offline from 21:00 UTC until approximately 23:00 UTC for a planned upgrade http://lists.openstack.org/pipermail/openstack-dev/2016-August/101395.html | 20:58 | |
| *** ChanServ changes topic to "The Mediawiki service at wiki.openstack.org will be offline from 21:00 UTC until approximately 23:00 UTC for a planned upgrade http://lists.openstack.org/pipermail/openstack-dev/2016-August/101395.html" | 20:58 | |
| *** jtriley has joined #openstack-kolla | 21:00 | |
| *** daneyon_ has quit IRC | 21:03 | |
| *** jtriley has quit IRC | 21:08 | |
| *** fragatina has joined #openstack-kolla | 21:11 | |
| *** mark-casey has quit IRC | 21:12 | |
| *** fragatina has quit IRC | 21:19 | |
| sdake | sbezverk https://review.openstack.org/355087 - best i can do for the moment | 21:23 |
| sdake | Jeffrey4l when you get around - if you want to fix this properly that wfm | 21:23 |
| sdake | i am going to take a nap atm | 21:23 |
| *** sdake has quit IRC | 21:23 | |
| *** huikang has joined #openstack-kolla | 21:38 | |
| *** ntpttr__ has joined #openstack-kolla | 21:39 | |
| *** dwalsh has quit IRC | 21:43 | |
| *** fragatina has joined #openstack-kolla | 21:48 | |
| *** fragatina has quit IRC | 21:50 | |
| *** haplo37_ has quit IRC | 21:50 | |
| *** fragatina has joined #openstack-kolla | 21:50 | |
| jrich523 | hey guys, anyone have a good link to building a compute node? something with beginner details would be nice.. also, im use VMWare Workstation to host all of it | 21:58 |
| *** dave-mccowan has quit IRC | 22:02 | |
| *** britthouser has joined #openstack-kolla | 22:04 | |
| *** britthou_ has joined #openstack-kolla | 22:06 | |
| *** huikang has quit IRC | 22:07 | |
| *** huikang has joined #openstack-kolla | 22:08 | |
| *** britthouser has quit IRC | 22:09 | |
| *** michauds has quit IRC | 22:11 | |
| *** huikang_ has joined #openstack-kolla | 22:12 | |
| *** huikang has quit IRC | 22:13 | |
| openstackgerrit | Serguei Bezverkhi proposed openstack/kolla-kubernetes: Script generating kubernetes secrets from passwords.yml https://review.openstack.org/354199 | 22:14 |
| *** britthou_ has quit IRC | 22:27 | |
| *** britthouser has joined #openstack-kolla | 22:27 | |
| *** ntpttr__ has quit IRC | 22:29 | |
| *** britthouser has quit IRC | 22:32 | |
| *** ntpttr__ has joined #openstack-kolla | 22:33 | |
| *** britthouser has joined #openstack-kolla | 22:34 | |
| *** huikang_ has quit IRC | 22:47 | |
| *** huikang has joined #openstack-kolla | 22:48 | |
| *** huikang has quit IRC | 22:52 | |
| -openstackstatus- NOTICE: ok https://wiki.openstack.org/ is now running Mediawiki 1.27.0; please let us know in #openstack-infra if anything seems wrong | 23:00 | |
| *** ChanServ changes topic to "ok https://wiki.openstack.org/ is now running Mediawiki 1.27.0; please let us know in #openstack-infra if anything seems wrong" | 23:00 | |
| *** ChanServ changes topic to "Kolla IRC meetngs on Wednesdays @ 16:00 UTC - see agenda @ https://goo.gl/OXB0DL - Community Submitted Talks for Occata https://goo.gl/s8Wldv - IRC channel is *LOGGED* @ http://goo.gl/3mzZ7b (old logs from #kolla http://goo.gl/VKpPzA)" | 23:06 | |
| -openstackstatus- NOTICE: https://wiki.openstack.org/ is now running Mediawiki 1.27.0; please let us know in #openstack-infra if anything seems wrong | 23:06 | |
| *** harmw has quit IRC | 23:20 | |
| *** britthouser has quit IRC | 23:32 | |
| *** sdake has joined #openstack-kolla | 23:33 | |
| *** daneyon_ has joined #openstack-kolla | 23:40 | |
| *** huikang has joined #openstack-kolla | 23:41 | |
| *** daneyon_ has quit IRC | 23:45 | |
| sdake | sbezverk did the gates get fixed? | 23:46 |
| *** ntpttr__ has quit IRC | 23:48 | |
| *** ntpttr__ has joined #openstack-kolla | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!