*** yamamoto has joined #openstack-kuryr | 00:03 | |
*** yamamoto has quit IRC | 00:07 | |
*** yamamoto has joined #openstack-kuryr | 00:24 | |
*** yamamoto has quit IRC | 00:29 | |
*** kiennt2680 has joined #openstack-kuryr | 00:30 | |
*** yamamoto has joined #openstack-kuryr | 00:44 | |
*** yamamoto has quit IRC | 00:51 | |
*** salv-orlando has joined #openstack-kuryr | 01:02 | |
*** salv-orl_ has quit IRC | 01:04 | |
*** yamamoto has joined #openstack-kuryr | 01:07 | |
*** yamamoto has quit IRC | 01:12 | |
*** yamamoto has joined #openstack-kuryr | 01:28 | |
*** yamamoto has quit IRC | 01:32 | |
*** janki has joined #openstack-kuryr | 01:33 | |
*** kiennt2680 has quit IRC | 01:38 | |
*** hongbin_ has joined #openstack-kuryr | 01:40 | |
*** salv-orlando has quit IRC | 01:41 | |
*** salv-orlando has joined #openstack-kuryr | 01:42 | |
*** kiennt2680 has joined #openstack-kuryr | 01:43 | |
*** kiennt2680 has quit IRC | 01:44 | |
*** kiennt2695 has joined #openstack-kuryr | 01:45 | |
*** kiennt2695 has quit IRC | 01:46 | |
*** kiennt2651 has joined #openstack-kuryr | 01:46 | |
*** kiennt2651 has quit IRC | 01:47 | |
*** yamamoto has joined #openstack-kuryr | 01:49 | |
*** kiennt2609 has joined #openstack-kuryr | 01:49 | |
*** yamamoto has quit IRC | 01:55 | |
*** salv-orl_ has joined #openstack-kuryr | 02:00 | |
*** salv-orlando has quit IRC | 02:03 | |
*** yamamoto has joined #openstack-kuryr | 02:12 | |
*** yamamoto has quit IRC | 02:17 | |
*** maysamacedos has quit IRC | 02:27 | |
*** yamamoto has joined #openstack-kuryr | 02:32 | |
*** yamamoto has quit IRC | 02:39 | |
*** jchhatbar has joined #openstack-kuryr | 02:51 | |
*** janki has quit IRC | 02:53 | |
*** yamamoto has joined #openstack-kuryr | 02:55 | |
*** yamamoto has quit IRC | 03:00 | |
*** kiennt2609 has quit IRC | 03:13 | |
*** kiennt2609 has joined #openstack-kuryr | 03:13 | |
*** yamamoto has joined #openstack-kuryr | 03:16 | |
*** yamamoto has quit IRC | 03:21 | |
*** gcheresh_ has joined #openstack-kuryr | 03:31 | |
*** yamamoto has joined #openstack-kuryr | 03:37 | |
*** yamamoto has quit IRC | 03:43 | |
*** yamamoto has joined #openstack-kuryr | 04:00 | |
*** salv-orl_ has quit IRC | 04:04 | |
*** yamamoto has quit IRC | 04:04 | |
*** salv-orlando has joined #openstack-kuryr | 04:04 | |
*** kiennt2609 has quit IRC | 04:05 | |
*** hongbin_ has quit IRC | 04:07 | |
*** janonymous has joined #openstack-kuryr | 04:17 | |
openstackgerrit | wangqi proposed openstack/kuryr-kubernetes master: fix a typo https://review.openstack.org/562108 | 04:20 |
---|---|---|
*** yamamoto has joined #openstack-kuryr | 04:21 | |
*** gcheresh_ has quit IRC | 04:24 | |
*** yamamoto has quit IRC | 04:25 | |
*** yamamoto has joined #openstack-kuryr | 04:41 | |
*** yamamoto has quit IRC | 04:48 | |
*** yamamoto has joined #openstack-kuryr | 04:48 | |
*** gcheresh_ has joined #openstack-kuryr | 04:52 | |
*** jchhatbar has quit IRC | 04:59 | |
*** jchhatbar has joined #openstack-kuryr | 05:00 | |
*** jchhatbar has quit IRC | 05:01 | |
*** jchhatbar has joined #openstack-kuryr | 05:01 | |
*** jchhatbar has quit IRC | 05:02 | |
*** jchhatbar has joined #openstack-kuryr | 05:03 | |
*** pcaruana has joined #openstack-kuryr | 05:05 | |
*** pcaruana has quit IRC | 05:19 | |
*** yboaron_ has joined #openstack-kuryr | 06:11 | |
*** kiennt2609 has joined #openstack-kuryr | 06:24 | |
*** jchhatba_ has joined #openstack-kuryr | 06:30 | |
*** jchhatbar has quit IRC | 06:33 | |
*** kiennt2609 has quit IRC | 06:38 | |
dmellado | dulek: that shouldn't be it | 06:41 |
dmellado | I've tried it on BM with that tarballed amphorae and works just fine | 06:41 |
*** jchhatba_ is now known as janki | 06:44 | |
*** pcaruana has joined #openstack-kuryr | 06:50 | |
openstackgerrit | Yossi Boaron proposed openstack/kuryr-kubernetes master: Kuryr controller - pluggable handlers documentation https://review.openstack.org/561447 | 07:09 |
*** chenyb4 has joined #openstack-kuryr | 07:49 | |
ltomasbo | dmellado, https://review.openstack.org/#/c/561244/ | 07:49 |
ltomasbo | can you +W that one, it is needed to do other backports | 07:49 |
*** yamamoto has quit IRC | 07:49 | |
*** yamamoto has joined #openstack-kuryr | 07:49 | |
*** yamamoto_ has joined #openstack-kuryr | 07:56 | |
*** yamamoto has quit IRC | 08:00 | |
dmellado | ltomasbo: there you go | 08:01 |
ltomasbo | dmellado, thanks! | 08:01 |
*** garyloug has joined #openstack-kuryr | 08:47 | |
openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-kubernetes master: Add namespace handler https://review.openstack.org/562159 | 09:22 |
openstackgerrit | Michał Dulko proposed openstack/kuryr-kubernetes master: Gate with containerized deployment and OpenShift https://review.openstack.org/557313 | 09:46 |
*** gcheresh has joined #openstack-kuryr | 09:51 | |
*** yboaron_ has quit IRC | 09:51 | |
*** celebdor1 has joined #openstack-kuryr | 09:55 | |
*** gcheresh_ has quit IRC | 09:55 | |
*** celebdor1 is now known as apuimedo | 09:55 | |
*** garyloug has quit IRC | 10:02 | |
*** garyloug has joined #openstack-kuryr | 10:02 | |
*** chenyb4 has quit IRC | 10:17 | |
*** pcaruana has quit IRC | 10:27 | |
*** rh-jelabarre has joined #openstack-kuryr | 11:05 | |
*** gcheresh has quit IRC | 11:08 | |
openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-kubernetes master: Add namespace subnet driver for namespace creation https://review.openstack.org/562247 | 11:19 |
*** yamamoto_ has quit IRC | 11:36 | |
openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-kubernetes master: Namespace deletion functionality for namespace_subnet driver https://review.openstack.org/562249 | 11:38 |
*** gcheresh has joined #openstack-kuryr | 11:58 | |
*** atoth has joined #openstack-kuryr | 11:59 | |
*** yamamoto has joined #openstack-kuryr | 12:03 | |
*** maysamacedos has joined #openstack-kuryr | 12:17 | |
*** pcaruana has joined #openstack-kuryr | 12:19 | |
*** janonymous has quit IRC | 13:06 | |
*** apuimedo has quit IRC | 13:07 | |
openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-kubernetes master: Add namespace subnet driver for namespace creation https://review.openstack.org/562247 | 13:28 |
openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-kubernetes master: Namespace deletion functionality for namespace_subnet driver https://review.openstack.org/562249 | 13:28 |
*** fabian_ has joined #openstack-kuryr | 13:33 | |
*** apuimedo has joined #openstack-kuryr | 13:35 | |
dulek | dmellado, ltomasbo: Probably better place to discuss Kuryr DevStack issues. | 13:45 |
dmellado | ltomasbo: dulek yep | 13:46 |
dmellado | ltomasbo: let's dig in your vm | 13:46 |
*** fabian_ is now known as chenyb4 | 13:46 | |
dulek | ltomasbo: https://gist.githubusercontent.com/dulek/518e7ef3685a1a08c39a684bd5ade28d/raw/5e483b039b1847a237942e57a139e673f0ac8e73/gistfile1.txt | 13:46 |
dmellado | ltomasbo: https://github.com/danielmellado.keys | 13:47 |
dulek | dmellado: Oh, that's cool, I can share my key like that? | 13:48 |
dmellado | dulek: totally, it'll share whichever keys you got there on github | 13:48 |
dulek | AWESOME. | 13:48 |
dmellado | xD | 13:48 |
dmellado | we might fail in discovering what's up but at least you got to learn something new xD | 13:49 |
ltomasbo | dulek, dmellado stack@38.145.33.129 | 13:49 |
dulek | ltomasbo: Hm, almost no iptables rules? | 13:50 |
ltomasbo | dulek, dmellado I have a tmux session | 13:50 |
ltomasbo | if you want to join! | 13:50 |
dmellado | ltomasbo: yep, I'm in | 13:51 |
ltomasbo | see, the cni crash | 13:51 |
ltomasbo | dmellado, and if I start it manually | 13:51 |
ltomasbo | I doubt it is sg, since it should not connect if the traffic is blocked | 13:52 |
dulek | WTF is LB VIP port down? | 13:52 |
dmellado | dulek: ltomasbo it looks to me like some issue on the cert, but I'm not familiar with that | 13:52 |
ltomasbo | it could be that something drops the connection? | 13:52 |
dulek | s/VIP/VIF | 13:52 |
ltomasbo | dulek, that is normal, right? | 13:52 |
ltomasbo | it is a port that gets added to the amphora as allows-address-pair, right? | 13:53 |
dulek | ltomasbo, dmellado: It's not certificate issue. | 13:53 |
dulek | I mean it is - curl don't get the server cert. | 13:53 |
dulek | But that's due to connection issue, not certificate. | 13:53 |
*** janonymous has joined #openstack-kuryr | 13:53 | |
dulek | Just don't think about certificates, trust me. | 13:53 |
ltomasbo | dulek, see the tmux | 13:53 |
dulek | Sec. | 13:54 |
ltomasbo | 10.0.0.129 is attached to octavia-lb-vrrp as allowed_address_pairs | 13:54 |
dulek | This one's down? :P | 13:54 |
ltomasbo | dulek, curl gets the ca.crt, right? | 13:54 |
ltomasbo | dulek, yep, but that is how octavia works | 13:55 |
dulek | Lemme show you the real issue. | 13:55 |
ltomasbo | it is normal that the lbaas VIP is down | 13:55 |
ltomasbo | apuimedo, ^^ can you confirm? | 13:55 |
dulek | ltomasbo, dmellado: See? | 13:55 |
dulek | Container has host networking. | 13:55 |
dmellado | a-ha | 13:55 |
dmellado | so it doesn't fetch a ***** | 13:55 |
dulek | So if LB is not accessible from the host we're screwed. | 13:55 |
dulek | And it is not for some reason. | 13:56 |
dulek | Now in the gate that worked even for Octavia. | 13:56 |
dulek | I'll check some more recent runs. | 13:56 |
ltomasbo | dulek, perhaps is a bug on ovs-firewall with allow-address-pairs... | 13:57 |
ltomasbo | it won't be the first time... | 13:57 |
dulek | Oh, waaaait… There's no gate with both Octavia and containers. | 13:57 |
dulek | Fuck. | 13:57 |
ltomasbo | so, that is why it is broken... xD | 13:57 |
ltomasbo | now it will be difficult to know when... | 13:58 |
dulek | Anything that's not tested is broken. :D | 13:58 |
ltomasbo | yep, that is the golden rule! | 13:58 |
dulek | Eh, okay, so what now. | 13:58 |
ltomasbo | I can try another thing | 13:58 |
dmellado | lol | 13:58 |
ltomasbo | maybe I can re-stack, with ovs-hybrid | 13:58 |
dulek | ltomasbo: Might be a good idea. | 13:58 |
ltomasbo | to see if that is the culprit or not | 13:58 |
dmellado | ltomasbo if you're using ovs-firewall that might be it | 13:58 |
ltomasbo | ok, it will take 10 min or so, let's try it out | 13:59 |
ltomasbo | dulek, dmellado: I'll ping you when it finishes... | 13:59 |
ltomasbo | not sure if I prefer it to be an ovs-firewall issue or not... | 13:59 |
dmellado | ltomasbo: ack, thanks! | 13:59 |
dmellado | dulek: in any case, TODO: add octavia + containers gate | 14:00 |
dmellado | xD | 14:00 |
dmellado | oh, FYI ian's patch got merged | 14:00 |
dulek | dmellado: DIB? | 14:00 |
dulek | dmellado: Is Octavia cloning DIB or we need to wait for a release? | 14:01 |
dmellado | that's what I'm trying to figure out now | 14:01 |
dmellado | probably they'll just pip it, but let's check | 14:01 |
dmellado | https://review.openstack.org/#/c/561479/10 | 14:01 |
dmellado | to be clear | 14:01 |
*** kiennt26_ has joined #openstack-kuryr | 14:04 | |
*** apuimedo has quit IRC | 14:06 | |
dmellado | dulek: it's cloned | 14:11 |
dmellado | so we'd be able to check it out now | 14:11 |
dulek | dmellado: Rechecing. | 14:11 |
dmellado | awesome | 14:12 |
dulek | Fingers crossed. ;) | 14:12 |
*** chenyb4 has quit IRC | 14:15 | |
dmellado | ltomasbo: dulek all right, devstack completed! xD | 14:18 |
dulek | ltomasbo: Finished. | 14:18 |
dmellado | heh, we both saw that | 14:18 |
dulek | ltomasbo: HA! | 14:18 |
dmellado | so f*** ovs-firewall | 14:19 |
dmellado | \o/ | 14:19 |
dulek | dmellado: List the pods please. | 14:19 |
dulek | ltomasbo: ^ :P | 14:19 |
dulek | Looks good | 14:19 |
ltomasbo | yep | 14:19 |
ltomasbo | again... | 14:19 |
dulek | ltomasbo: Will you file a bug that K8s API LB is broken with ovs-firewall? | 14:20 |
ltomasbo | dulek, sure! | 14:20 |
*** celebdor1 has joined #openstack-kuryr | 14:20 | |
*** celebdor1 is now known as apuimedo | 14:20 | |
apuimedo | ltomasbo: did you read my response on VIP active status? | 14:21 |
dmellado | apuimedo: I guess we didn't | 14:21 |
dmellado | apuimedo: go for it xD | 14:21 |
apuimedo | I was saying that it is normal | 14:21 |
apuimedo | since it is never bound | 14:21 |
apuimedo | it only gets added to allows ip addresses of the amphorae ports | 14:21 |
ltomasbo | apuimedo, I did not... | 14:21 |
apuimedo | so nothing sets it active | 14:21 |
ltomasbo | apuimedo, that is what I thought | 14:21 |
ltomasbo | thanks! | 14:22 |
dmellado | yeah, that was our idea | 14:22 |
ltomasbo | dulek, dmellado apuimedo I'm pinging kuba about the ovs-firewall issue | 14:22 |
dmellado | ltomasbo: this brings back memories | 14:23 |
ltomasbo | xD | 14:24 |
apuimedo | which issue?! | 14:24 |
apuimedo | we have a new issue? | 14:24 |
dmellado | apuimedo: lazy emerit PTL, read the backlog xD xD | 14:25 |
* dmellado can't but think about mamastova when hearing emerit | 14:25 | |
apuimedo | dmellado: I got disconnected | 14:26 |
dulek | apuimedo: eavesdrop.openstack.org/irclogs ;) | 14:27 |
dulek | apuimedo: ltomasbo tried containerzied with Octavia and ovs-firewall. | 14:28 |
dulek | apuimedo: We've hit K8s API LB connection issue. | 14:28 |
dulek | apuimedo: Same as the one you've debugged on gate vm. | 14:28 |
dmellado | related to ovs-firewall this time | 14:29 |
*** hongbin_ has joined #openstack-kuryr | 14:40 | |
dmellado | btw apuimedo irenab ltomasbo dulek related to *next* PTG it'll be on NA, September 10-14, 2018. | 14:43 |
dmellado | there's a *lot* of time still but we might need to be aware of these dates | 14:43 |
dmellado | (and need to vote whether we do that virtual or attend) | 14:43 |
apuimedo | dmellado: where? | 14:43 |
dmellado | personally I think it was worth going there f2f, besides the blizzard and so | 14:44 |
apuimedo | NA is like saying "Asia" | 14:44 |
dmellado | that's still not known | 14:44 |
dmellado | probably Canada, IIUC | 14:44 |
apuimedo | dmellado: not even if coast or middle? | 14:44 |
dmellado | but still not disclosed | 14:44 |
dmellado | not for now, I guess I'll be able to get more info soon | 14:45 |
dmellado | will discuss there | 14:45 |
dmellado | I'd vote for Montreal, if it were my call xD | 14:45 |
dulek | dmellado: Montreal was on the list as alternative for Denver. | 14:46 |
dulek | dmellado: So it might be it. | 14:46 |
apuimedo | reusing visa | 14:46 |
dulek | dmellado: Kuryr team is mostly in EMEA. It might be hard to get all of us to NA. | 14:47 |
apuimedo | xD | 14:47 |
apuimedo | if dmellado doesn't wash his passport again | 14:47 |
apuimedo | dulek: I'd rather do a KTG in Brno, TLV or Barcelona | 14:47 |
dmellado | dulek: that's partially why I was thinking about VTG | 14:47 |
dmellado | and maybe some internal mid-cycle | 14:47 |
dmellado | but we'll get to that | 14:48 |
dmellado | apuimedo: LOL | 14:48 |
dulek | dmellado: You've washed your passport? | 14:48 |
dmellado | I kept that safe this time | 14:48 |
dmellado | dulek: YEAH | 14:48 |
apuimedo | dulek: twice | 14:48 |
dulek | Nice. BTW I needed no visa to get to Canada. | 14:48 |
dmellado | it was on my pants after coming from Sydney and as I was really tired I dropped everything on the washing machine without realizing xD | 14:48 |
apuimedo | xD | 14:48 |
dmellado | and after that it happened again because I left that with a lot of clothings and my GF didn't realize | 14:48 |
dulek | dmellado: Well, 10 hour flight is an acceptable excuse. | 14:48 |
dmellado | after that | 14:49 |
dmellado | I keep the latest one inside a safe place | 14:49 |
dmellado | xD | 14:49 |
dmellado | dulek: if only 10, the longest flight back to Sydney was 15 hours xD | 14:49 |
dulek | I'm glad it's not Sydney then. | 14:49 |
dmellado | dulek: for me it was first 7 hours to Abu-Dhabi, 3 hours layover | 14:50 |
dmellado | and then 15 hours | 14:50 |
dmellado | yeepee | 14:50 |
dmellado | xD | 14:50 |
dmellado | after reading, watching films, finishing newest mario and sleeping I still had like 8 hours left | 14:51 |
dmellado | xD | 14:51 |
dulek | apuimedo, irenab: Fix for pip 10 in stable/queens for Kuryr: https://review.openstack.org/#/c/562295/ | 14:52 |
dulek | dmellado: Probably you have rights as well. ^ | 14:52 |
*** yamamoto has quit IRC | 14:53 | |
dmellado | dulek: I should but looks like I don't | 14:54 |
dmellado | dulek: we've got the new cli command | 14:55 |
dmellado | apudo | 14:55 |
dmellado | which is the equivalent of sudo for that repo | 14:55 |
dmellado | apudo merge it xD | 14:55 |
dmellado | apuimedo: ^^xD | 14:55 |
dulek | dmellado: I've added the comment. | 14:57 |
apuimedo | I agree with dmellado's comment | 14:58 |
apuimedo | on needing comment in the tox install | 14:58 |
dulek | apuimedo: Fixed. ;) | 14:58 |
dmellado | apuimedo: would you do the honors now? | 14:58 |
apuimedo | dmellado has merge rights | 14:58 |
apuimedo | dulek: dmellado: I'll merge it when zuul posts the test results and they are +1 | 15:01 |
apuimedo | gotta go afk for 10 minutes | 15:01 |
dmellado | fair enough | 15:01 |
dulek | Sure. | 15:01 |
dmellado | I'll be afk from today one, gotta go to my favorite place again | 15:02 |
dmellado | so tired of hospitals, meh | 15:02 |
dmellado | today on | 15:02 |
dmellado | heh, somthing's odd with the gerrit groups https://review.openstack.org/#/admin/groups/996,members | 15:03 |
dmellado | anyways, folks, I'm off, good afternoon! | 15:05 |
apuimedo | bak | 15:09 |
*** gcheresh has quit IRC | 15:17 | |
dulek | apuimedo: Success again: https://review.openstack.org/#/c/562295/ | 15:18 |
apuimedo | gotcha | 15:19 |
*** kiennt26_ has quit IRC | 15:22 | |
openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-kubernetes master: Add namespace subnet driver for namespace creation https://review.openstack.org/562247 | 15:27 |
openstackgerrit | Luis Tomas Bolivar proposed openstack/kuryr-kubernetes master: Namespace deletion functionality for namespace_subnet driver https://review.openstack.org/562249 | 15:27 |
*** gcheresh has joined #openstack-kuryr | 15:40 | |
*** premsankar has joined #openstack-kuryr | 15:46 | |
*** gcheresh has quit IRC | 15:46 | |
openstackgerrit | Antoni Segura Puimedon proposed openstack/kuryr-kubernetes master: devstack: Fix octavia api lb kubelet iface access https://review.openstack.org/562322 | 15:47 |
dulek | ltomasbo: Are you able to confirm if this solves the issue. ^ | 15:53 |
dulek | ? | 15:53 |
*** yamamoto has joined #openstack-kuryr | 15:53 | |
apuimedo | dulek: is https://review.openstack.org/#/c/561974/1..2/playbooks/tempest-kuryr-kubernetes-base/run.yaml cheating? | 15:54 |
apuimedo | xD | 15:54 |
dulek | OH MY, SOMEONE NOTICED!? | 15:54 |
apuimedo | dulek: xD | 15:55 |
dulek | apuimedo: Idea is simple - Octavia installs diskimage-builder from git if it's in LIBS_FROM_GIT. | 15:55 |
dulek | apuimedo: https://github.com/openstack/octavia/blob/master/devstack/plugin.sh#L33-L38 | 15:55 |
dulek | apuimedo: If it's not… well I don't know how it installs it. :D | 15:55 |
apuimedo | dulek: ok. I can live with it | 15:55 |
ltomasbo | dulek, testing it right now | 15:55 |
dulek | apuimedo: But lets see if this passes tests. | 15:56 |
apuimedo | yes, yes. I'm holding on the w | 15:56 |
dulek | apuimedo: If it will that would mean DIB workaround is solving the issue and we can think what's the proper way of including it. | 15:56 |
*** yamamoto has quit IRC | 16:00 | |
dulek | apuimedo: Is it normal that my env is missing service_pod_access SG? | 16:05 |
apuimedo | dulek: nope. It is not | 16:06 |
apuimedo | dulek: paste me `openstack security group list` | 16:06 |
dulek | It's created only for Octavia. | 16:06 |
apuimedo | really? | 16:07 |
apuimedo | I forgot that | 16:07 |
apuimedo | you're right | 16:07 |
*** maysamacedos has quit IRC | 16:08 | |
dulek | Yup. Anyway LBaaS v2 was working fine for ltomasbo. | 16:08 |
ltomasbo | dulek, yep | 16:08 |
apuimedo | so I'll put the test | 16:08 |
apuimedo | for it being octavia and l3 mode | 16:09 |
ltomasbo | apuimedo, I can give it a try to l2 mode after this one... | 16:09 |
ltomasbo | (it is taking a bit more than the 10 minutes I said) | 16:09 |
apuimedo | ltomasbo: in l2 mode the origin of the communication is from the same subnet as kubelet | 16:10 |
apuimedo | so it'll work | 16:10 |
openstackgerrit | Antoni Segura Puimedon proposed openstack/kuryr-kubernetes master: devstack: Fix octavia api lb kubelet iface access https://review.openstack.org/562322 | 16:10 |
ltomasbo | apuimedo, probably, I remember a different sg rule being added for l2 | 16:12 |
ltomasbo | but maybe was to the pod itself | 16:12 |
apuimedo | ltomasbo: none should be necessary | 16:12 |
apuimedo | that's for pods | 16:12 |
apuimedo | and in fact, for l2 mode that one should not be necessary either | 16:12 |
ltomasbo | \o/ it worked | 16:13 |
ltomasbo | apuimedo, actually, based on devstack/plugin | 16:16 |
ltomasbo | it maybe be needed to add the octavia_pod_access_sg_id | 16:16 |
ltomasbo | apuimedo, rule --> "k8s pod subnet allowed from k8s-pod-subnet" | 16:17 |
ltomasbo | ahh, no, that is needed for the pods to be reached from the amphora | 16:18 |
ltomasbo | probably not for the kubelet | 16:18 |
*** maysamacedos has joined #openstack-kuryr | 16:21 | |
dulek | ltomasbo: How about restacking your LBaaS v2 VM with OpenShift instead of Kubernetes? | 16:26 |
dulek | ltomasbo: I have similar issue on my local env. | 16:26 |
dulek | Works fine with K8s but not with OpenShift. | 16:26 |
dulek | And I'm using LBaaS + iptables_hybrid, so it's not SG. | 16:27 |
ltomasbo | LBaasV2 + Kubernetes (without apuimedo's patch?) | 16:27 |
ltomasbo | dulek, ^^ | 16:27 |
ltomasbo | dulek, just the local.conf.openshift config? | 16:28 |
dulek | ltomasbo: What I do is comment out all enable_service kubernetes-* and enable_service kubelet. | 16:29 |
ltomasbo | umm | 16:29 |
dulek | ltomasbo: And then: enable_service openshift-master openshift-node | 16:29 |
dulek | DOCKER_CGROUP_DRIVER="systemd" | 16:29 |
ltomasbo | I think you need to leave the kuryr-kubernetes, right? | 16:29 |
dulek | ltomasbo: kuryr-kubernetes doesn't match kubernetes-*. :D | 16:30 |
ltomasbo | sorry, you/re right | 16:30 |
ltomasbo | ok, that I have it | 16:31 |
ltomasbo | I need to go, but I'll restack it and take a look later/tomorrow | 16:31 |
*** yamamoto has joined #openstack-kuryr | 16:56 | |
*** yamamoto has quit IRC | 17:00 | |
dulek | Whooo, I've got a really nice bug! | 17:03 |
dulek | apuimedo: Interested? | 17:03 |
apuimedo | dulek: if it's nice I am | 17:03 |
dulek | apuimedo: I've had hard time reproducing this on my env: http://logs.openstack.org/13/557313/5/experimental/kuryr-kubernetes-tempest-daemon-containerized-openshift-lbaasv2/0158ebb/controller/logs/kubernetes/pod_logs/kube-system-kuryr-controller-3371820712-z69zv.txt.gz#_2018-04-18_10_32_19_946 | 17:04 |
dulek | apuimedo: But I finally got it! | 17:04 |
dulek | apuimedo: You know when it happens? When pod is created on its own and not through deployment. | 17:04 |
dulek | apuimedo: (this is OpenShift specific) | 17:04 |
apuimedo | :O | 17:04 |
apuimedo | did you file a bug on both lp and bz? | 17:05 |
dulek | apuimedo: I assume this means that it doesn't ring a bell. :D | 17:05 |
dulek | apuimedo: Not yet, I was still thinking it's my patch fault. | 17:05 |
dulek | apuimedo: This happens in containerized deployment only, so I assume this is related to SA missing something. | 17:06 |
apuimedo | dulek: it does not | 17:07 |
apuimedo | ring a bell | 17:07 |
apuimedo | dulek: such a big wall of text | 17:08 |
apuimedo | it's hard to see the actual problem | 17:08 |
apuimedo | oh | 17:08 |
dulek | apuimedo: List of fields that can be edited, metadata is not one of them. | 17:08 |
apuimedo | spec: Forbidden: pod updates may not change fields other than `spec.containers[*].image`, `spec.initContainers[*].image`, `spec.activeDeadlineSeconds` or `spec.tolerations` | 17:08 |
apuimedo | wtf | 17:08 |
apuimedo | why would metadata not be editable | 17:09 |
apuimedo | openshift version? | 17:09 |
dulek | apuimedo: 3.7.1, this is what we run in the gate. | 17:09 |
dulek | I've just annotated that pod manually through `kubectl annotate`. | 17:10 |
dulek | This is definitely some ServiceAccount permissions issue. | 17:10 |
apuimedo | probably | 17:13 |
apuimedo | dulek: and we should move to 3.8 | 17:13 |
apuimedo | sorry, 3.9 | 17:13 |
apuimedo | the binaries are finally there https://github.com/openshift/origin/releases/tag/v3.9.0 | 17:13 |
dulek | apuimedo: Okay, retest that with 3.9 and if it's still failing I'll start filing bugs. | 17:14 |
dulek | I'll retest | 17:14 |
apuimedo | thanks dulek | 17:15 |
*** garyloug has quit IRC | 17:28 | |
*** janki has quit IRC | 17:37 | |
*** yamamoto has joined #openstack-kuryr | 17:56 | |
*** yamamoto has quit IRC | 18:02 | |
dulek | Looks like stuff's working for 3.9, I'll follow up in an hour or two. | 18:06 |
*** pcaruana has quit IRC | 18:12 | |
celebdor | Cool | 18:43 |
*** yamamoto has joined #openstack-kuryr | 18:57 | |
*** yamamoto has quit IRC | 19:03 | |
*** apuimedo has quit IRC | 19:03 | |
*** apuimedo has joined #openstack-kuryr | 19:09 | |
*** pcaruana has joined #openstack-kuryr | 19:13 | |
*** pcaruana has quit IRC | 19:20 | |
openstackgerrit | Michał Dulko proposed openstack/kuryr-kubernetes master: Gate with containerized deployment and OpenShift https://review.openstack.org/557313 | 19:28 |
openstackgerrit | Michał Dulko proposed openstack/kuryr-kubernetes master: Raise OpenShift version to 3.9.0 https://review.openstack.org/562366 | 19:28 |
*** atoth has quit IRC | 19:30 | |
*** premsankar has quit IRC | 19:55 | |
*** yamamoto has joined #openstack-kuryr | 19:59 | |
*** yamamoto has quit IRC | 20:04 | |
openstackgerrit | Antoni Segura Puimedon proposed openstack/kuryr-kubernetes master: devstack: fix octavia allocation ranges https://review.openstack.org/562067 | 20:20 |
*** yamamoto has joined #openstack-kuryr | 20:44 | |
*** dmellado has quit IRC | 20:45 | |
*** rh-jelabarre has quit IRC | 21:42 | |
*** premsankar has joined #openstack-kuryr | 21:59 | |
*** yamamoto_ has joined #openstack-kuryr | 22:04 | |
*** yamamoto has quit IRC | 22:08 | |
*** pmannidi has joined #openstack-kuryr | 22:19 | |
*** apuimedo has quit IRC | 23:04 | |
*** maysamacedos has quit IRC | 23:13 | |
*** maysamacedos has joined #openstack-kuryr | 23:13 | |
*** maysamacedos has quit IRC | 23:15 | |
*** hongbin_ has quit IRC | 23:18 | |
*** maysamacedos has joined #openstack-kuryr | 23:42 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!