| *** hongbin has quit IRC | 00:02 | |
| *** yuanying_ has joined #openstack-kuryr | 00:03 | |
| *** limao has joined #openstack-kuryr | 00:29 | |
| openstackgerrit | dengshaolin proposed openstack/fuxi: Retrieve volumes from cinder with matched metadata https://review.openstack.org/390448 | 00:34 |
|---|---|---|
| openstackgerrit | dengshaolin proposed openstack/fuxi: Fix indexing operation for OrderDict in py3.4 and py3.5 https://review.openstack.org/403958 | 00:40 |
| *** shashank_hegde has quit IRC | 00:46 | |
| *** tonanhngo has quit IRC | 01:01 | |
| *** tonanhngo has joined #openstack-kuryr | 01:10 | |
| *** tonanhngo has quit IRC | 01:14 | |
| openstackgerrit | bailin.zhang proposed openstack/kuryr-libnetwork: Fix the file permissions of README.rst https://review.openstack.org/396519 | 01:20 |
| *** shashank_hegde has joined #openstack-kuryr | 01:21 | |
| *** yamamot__ has joined #openstack-kuryr | 01:41 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/fuxi: Updated from global requirements https://review.openstack.org/373745 | 01:42 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/kuryr-libnetwork: Updated from global requirements https://review.openstack.org/351976 | 01:44 |
| *** yamamot__ has quit IRC | 01:44 | |
| *** shashank_hegde has quit IRC | 01:56 | |
| *** shashank_hegde has joined #openstack-kuryr | 01:57 | |
| *** yamamot__ has joined #openstack-kuryr | 02:04 | |
| *** shashank_hegde has quit IRC | 02:20 | |
| *** hongbin has joined #openstack-kuryr | 02:22 | |
| *** reedip has quit IRC | 02:29 | |
| *** reedip has joined #openstack-kuryr | 02:32 | |
| *** yuanying_ has quit IRC | 03:00 | |
| *** oanson has joined #openstack-kuryr | 03:07 | |
| *** yamamo___ has joined #openstack-kuryr | 03:22 | |
| *** yamamot__ has quit IRC | 03:26 | |
| *** yuanying has joined #openstack-kuryr | 03:59 | |
| *** tonanhngo has joined #openstack-kuryr | 04:08 | |
| *** tonanhngo has quit IRC | 04:09 | |
| *** jgriffith is now known as jgriffith_away | 04:21 | |
| *** tonanhngo has joined #openstack-kuryr | 04:21 | |
| *** tonanhngo has quit IRC | 04:22 | |
| *** hongbin has quit IRC | 04:24 | |
| *** oanson has quit IRC | 04:30 | |
| *** reedip has quit IRC | 04:42 | |
| *** rock has quit IRC | 04:47 | |
| *** tonanhngo has joined #openstack-kuryr | 05:04 | |
| *** shashank_hegde has joined #openstack-kuryr | 05:05 | |
| *** tonanhngo has quit IRC | 05:06 | |
| *** shashank_hegde has quit IRC | 05:12 | |
| *** shashank_hegde has joined #openstack-kuryr | 05:35 | |
| *** shashank_hegde has quit IRC | 06:16 | |
| *** irenab has joined #openstack-kuryr | 06:24 | |
| *** shashank_hegde has joined #openstack-kuryr | 06:27 | |
| irenab | vikasc, hi | 06:38 |
| openstackgerrit | Ilya Chukhnakov proposed openstack/kuryr-kubernetes: [WIP] CNI & VIFBridge binding https://review.openstack.org/404038 | 06:39 |
| *** oanson has joined #openstack-kuryr | 06:41 | |
| ivc_ | irenab, apuimedo, ping | 06:43 |
| irenab | ivc_, pong | 06:43 |
| ivc_ | irenab, apuimedo, just uploaded a working CNI https://review.openstack.org/404038 (still needs some cleanup, routes handling, tests and will also be split in smaller patches) | 06:44 |
| irenab | ivc_, I will take a look and try to run it | 06:45 |
| irenab | ivc_, can you please post your local.conf? | 06:45 |
| ivc_ | irenab, apuimedo, also tried some 'stress-tests' by running "kubectl run nginx --image=nginx --replicas=30" (takes ~2min) and "kubectl delete deployment nginx" (works fine, but seems to exceed k8s timeout with some very unexpected results as it gets called second time by kubelet after pods are deleted and there's no annotations) | 06:47 |
| ivc_ | irenab, http://paste.openstack.org/show/590746/ | 06:48 |
| ivc_ | irenab, but you also need to 'cd /opt/stack/kuryr-kubernetes && sudo pip install -r requirements.txt' to get rid of 'kuryr.lib' issues | 06:49 |
| irenab | ivc_, thanks | 06:49 |
| irenab | ivc_, regarding the second time delete, worth to check what k8s devs expected in such case. Looks like in our case its just 'success' operation | 06:54 |
| ivc_ | irenab, that behaviour is really weird. everything is cleaned up properly on the first run, but for some reason k8s _sometimes_ calls cni again (and then we fail with our own timeout since '?watch=true' gets stuck) | 06:59 |
| irenab | ivc_, maybe watch should be kicked upon condition? | 07:00 |
| irenab | did you try to ask guys on k8s slack? | 07:00 |
| ivc_ | irenab, what sort of condition do you suggest? like check pod with 'get' first? | 07:01 |
| irenab | actually, I would expect kubelet to take care of it ... | 07:01 |
| *** yamamo___ has quit IRC | 07:02 | |
| irenab | let me take a look on the code | 07:02 |
| ivc_ | you mean kubelet code? | 07:02 |
| irenab | yes, that it will not fire CNI for the false action | 07:03 |
| irenab | checking if pod exists sounds like workaround | 07:04 |
| irenab | also it assumes that pod is removed from the k8s data store after kebelet 'approves' its deleted | 07:05 |
| irenab | kubelet | 07:05 |
| irenab | ivc_, do you know how the k8s Pod delete flow goes? | 07:05 |
| ivc_ | irenab, gotta dig into kubelet source | 07:06 |
| irenab | ivc_, same for me :-) | 07:06 |
| ivc_ | irenab, what i'm afraid of is that probably there's no contract between cni and kube-api that when cni is called pod can be queried from kube-api | 07:07 |
| irenab | ivc_, starting to think the same | 07:08 |
| ivc_ | given the async nature of k8s it might be totally possible and then we are screwed | 07:08 |
| irenab | but still do not understand how you cpuld get pods without annotations | 07:08 |
| irenab | it either with annotations or does not exist | 07:09 |
| ivc_ | nope i'm not getting pods without annotations :) i'm not getting pods at all :) | 07:09 |
| irenab | aha | 07:09 |
| irenab | now it makes sense | 07:09 |
| irenab | so assuming there is no contact between Controller part and net plugin part on when pod is removed from the data model | 07:10 |
| ivc_ | yep it does seems so | 07:10 |
| irenab | CNI should try to keep its sources local | 07:10 |
| ivc_ | we can probably still can get away with our solution with daemon-based cni | 07:11 |
| irenab | yes, also thinking of it | 07:12 |
| irenab | actually why do we need k8s api for the delete case? | 07:12 |
| ivc_ | we need vif details | 07:12 |
| ivc_ | so we can os_vif.unplug | 07:13 |
| irenab | the best way would to get it with other args | 07:13 |
| ivc_ | otherwise we'll have leftover hybrid bridges | 07:13 |
| ivc_ | no way we can get it with other args | 07:13 |
| ivc_ | i mean with non-daemon cni | 07:14 |
| irenab | I was thinking about adding extra power to kubelet | 07:14 |
| irenab | probably longer term | 07:14 |
| irenab | or we must keep local cache | 07:15 |
| irenab | container id -> vif details | 07:15 |
| irenab | probably what you mean by daemon | 07:16 |
| ivc_ | yep. we need daemon anyway. with the current approach there's one hell of a storm of privsep helpers spawned by os_vif | 07:17 |
| ivc_ | and we planned it anyway | 07:18 |
| *** yuanying has quit IRC | 07:20 | |
| irenab | so the CNI Add/Del will just send request to daemon and block for reply? | 07:21 |
| ivc_ | yeah | 07:21 |
| *** pcaruana has joined #openstack-kuryr | 07:21 | |
| *** saneax has joined #openstack-kuryr | 07:21 | |
| irenab | its just annoying that for Del we need it only for the extra vif info | 07:23 |
| ivc_ | its not extra | 07:24 |
| irenab | I mean which is not pushed to CNI | 07:24 |
| ivc_ | its part of vif. think of ovs-hybrid case at least - there's that linux bridge | 07:24 |
| ivc_ | ah, right | 07:24 |
| ivc_ | well, k8s cni-plugin is still experimental, so maybe it could be customized the way we need | 07:26 |
| *** yuanying has joined #openstack-kuryr | 07:26 | |
| ivc_ | the best place for our data is the netcfg json that is streamed to cni.stdin | 07:27 |
| irenab | I need some time to go over the code and try to dig into k8s code. Maybe we can consult with k8s guys | 07:27 |
| irenab | ivc_, I agree | 07:27 |
| irenab | and get it in 'push' from kubelet | 07:27 |
| ivc_ | irenab, i wonder why did we not find out that k8s behaviour with midokura's PoC | 07:29 |
| *** dimak has joined #openstack-kuryr | 07:37 | |
| *** yamamot__ has joined #openstack-kuryr | 07:39 | |
| irenab | ivc_, I do not think it was properly tested on scale + maybe previous k8s version behaved differently | 07:42 |
| *** lezbar has joined #openstack-kuryr | 08:08 | |
| *** lezbar has quit IRC | 08:17 | |
| *** yuanying has quit IRC | 08:17 | |
| janonymous | alraddarla_: cool, you could try https://review.openstack.org/#/c/394788/, https://review.openstack.org/#/c/394251/ , https://review.openstack.org/#/c/393996/ if that's fine | 08:19 |
| *** lezbar has joined #openstack-kuryr | 08:19 | |
| *** irenab has quit IRC | 08:37 | |
| *** irenab has joined #openstack-kuryr | 08:39 | |
| openstackgerrit | Merged openstack/kuryr: Nested-Containers: vlan driver https://review.openstack.org/361993 | 08:49 |
| apuimedo | ivc_: "the best place for our data is the netcfg json that is streamed to cni.stdin" what makes you say that? | 08:50 |
| *** shashank_hegde has quit IRC | 08:50 | |
| *** shashank_hegde has joined #openstack-kuryr | 08:59 | |
| *** shashank_hegde has quit IRC | 08:59 | |
| *** pmannidi has quit IRC | 09:00 | |
| *** garyloug has joined #openstack-kuryr | 09:01 | |
| *** lmdaly has joined #openstack-kuryr | 09:02 | |
| irenab | apuimedo, did you read the discussion we had with ivc_ this morning? | 09:11 |
| *** gsagie has joined #openstack-kuryr | 09:40 | |
| *** ram____ has joined #openstack-kuryr | 09:44 | |
| *** lezbar has quit IRC | 09:44 | |
| *** openstackgerrit has quit IRC | 09:48 | |
| *** openstackgerrit has joined #openstack-kuryr | 09:48 | |
| apuimedo | irenab: I did | 09:50 |
| apuimedo | I'm a bit ill so maybe I didn't get all to perfection | 09:50 |
| apuimedo | but I thought the netcfg json that is passed to stdin is more for static configuration | 09:50 |
| apuimedo | so it would not serve for passing ovs-hybrid bridge names nor uuids | 09:51 |
| apuimedo | If we want to avoid needing watch for delete we could just put data on tempfiles like /run/kuryr/uuid_pod | 09:51 |
| apuimedo | irenab: ivc_: ^^ | 09:51 |
| irenab | apuimedo, when do you suggest to place the tempfiles? On Add? | 09:57 |
| apuimedo | just before returning success on add | 09:57 |
| irenab | apuimedo, workaround, but possible. so daemon is not needed | 10:01 |
| irenab | apuimedo, rest and get well soon! | 10:01 |
| apuimedo | irenab: nah... | 10:03 |
| apuimedo | no rest | 10:03 |
| apuimedo | irenab: https://www.youtube.com/watch?v=jB2nWiEO9T0 | 10:04 |
| apuimedo | like this | 10:04 |
| apuimedo | ivc_: what do you think of the files? | 10:04 |
| *** reedip has joined #openstack-kuryr | 10:11 | |
| *** limao has quit IRC | 10:11 | |
| apuimedo | irenab: ivc_: vikasc: https://review.openstack.org/#/c/404124/ | 10:18 |
| openstackgerrit | Merged openstack/kuryr-libnetwork: Updated from global requirements https://review.openstack.org/351976 | 10:23 |
| *** ram____ has quit IRC | 10:44 | |
| openstackgerrit | Merged openstack/kuryr-libnetwork: Build docker busybox image locally https://review.openstack.org/399958 | 11:42 |
| *** yamamot__ has quit IRC | 11:46 | |
| apuimedo | vikasc: https://review.openstack.org/395453 | 11:51 |
| *** vikasc has quit IRC | 11:56 | |
| *** gsagie has quit IRC | 12:14 | |
| *** lmdaly has quit IRC | 12:30 | |
| apuimedo | ivc_: irenab: I just posted a gate that tries to install devstack https://review.openstack.org/#/c/404191/ | 12:31 |
| apuimedo | of course, it will fail until we solve the kuryr-lib release | 12:31 |
| apuimedo | which I also submitted | 12:31 |
| *** limao has joined #openstack-kuryr | 12:45 | |
| *** garyloug has quit IRC | 12:51 | |
| ivc_ | apuimedo, thats fantastic! | 12:59 |
| *** limao_ has joined #openstack-kuryr | 13:04 | |
| ivc_ | apuimedo, regarding the 'files' approach, i'm not sure. one part of me thinks its terribly UGLY, but on the other hand its quite simple. | 13:05 |
| ivc_ | apuimedo, but i'd prefer to avoid it if we can. imo the problem would solve itself as soon as we get to 'recycling' ports/vifs | 13:07 |
| *** limao has quit IRC | 13:07 | |
| ivc_ | apuimedo, irenab, and our current cni approach is ineffective at best anyway (our cni is fat&slow, multiple IPDB instances flood netlink with unneded spam, os-vif spawns a separate sudo/rootwrap process on each cni exec) so we'll need to do something about it anyway even if it wasn't for the issue with de-sync between cni and k8s-api | 13:14 |
| *** yamamoto has joined #openstack-kuryr | 13:20 | |
| *** yamamoto_ has joined #openstack-kuryr | 13:22 | |
| *** yamamoto has quit IRC | 13:25 | |
| irenab | ivc_, 'one part of me' ? | 13:25 |
| irenab | ivc_, I think you are in favor of the daemon from the day one. | 13:27 |
| ivc_ | irenab, well, not exactly 'from the day one', but i am :) | 13:31 |
| irenab | ivc_, for recycling ports/vifs, what do you have in mind? per Clusrter or per node? | 13:32 |
| ivc_ | irenab, performance-wise per-node would be best imo | 13:35 |
| ivc_ | irenab, best solution would be if we could keep container-side interface and relocate it into some 'recycling' namespace(s) | 13:39 |
| irenab | ivc_, neutron related, you distribute ports across nodes? | 13:41 |
| ivc_ | irenab, but i need to deal with services/lbaas first, so wdyt if we merge the current 'fat' cni (it still requires some tweaks and maybe has to be split in multiple patches) | 13:41 |
| irenab | ivc_, with files as apuimedo suggested? | 13:42 |
| ivc_ | irenab, no, without files | 13:42 |
| irenab | ivc_, we can. I think we keep kuryr-k8s experimental anyway | 13:42 |
| irenab | ivc_, so how delete is resolved? | 13:42 |
| ivc_ | irenab, it works now, you just see some unwanted messages in k8s logs | 13:43 |
| irenab | how vif is removed ? | 13:43 |
| alraddarla_ | janonymous, i'll take care of the three you mentioned. thanks for letting me help :) | 13:43 |
| ivc_ | irenab, https://review.openstack.org/#/c/404038/1/kuryr_kubernetes/cni/binding/base.py@57 | 13:44 |
| irenab | ivc_, in case pod is not in k8s data model, will CNI be able to clean the host properly? | 13:44 |
| ivc_ | irenab, during the first call to cni pod is available, but k8s calls cni second time in some cases | 13:45 |
| apuimedo | ivc_: we should still find out why that second call happens | 13:46 |
| apuimedo | irenab: did you check on k8s? | 13:46 |
| irenab | ivc_, not sure we can assume if during the first time Pod is in k8s data model. | 13:47 |
| irenab | apuimedo, not yet | 13:48 |
| apuimedo | I still prefer if we saved the trip to k8s :P | 13:48 |
| apuimedo | let's say that the split small exec / daemon solves it | 13:49 |
| irenab | apuimedo, agree on saving the trip for delete. Not sure we can do it for Add | 13:49 |
| apuimedo | in that solution, I assume the daemon keeps a dict of all the vifs for the host | 13:49 |
| apuimedo | then when it receives the request from the executable, it doesn't need to go to k8s anymore, since it has the data from watch | 13:50 |
| apuimedo | if it doesn't have data, it should either 'get' or timeout | 13:50 |
| ivc_ | irenab, apuimedo, my point is that if you consider the 'fat' cni approach only as a temporary one (as i do) that will allow us to continue with testing (and we don't run 100s of pods for tests, right?) and services. and once we get to functionally-complete impl, we rework cni into daemon/exec | 13:50 |
| apuimedo | which is what happens now | 13:50 |
| apuimedo | I'm talking about the lean cni now | 13:51 |
| irenab | ivc_, I am ok with you proposal for taking steps | 13:51 |
| irenab | apuimedo, what do you think? | 13:51 |
| apuimedo | the only satisfactory solution for lean cni (split) that I can see is that the dict has a big size with LRU eviction. On first delete we just None the entry, and then when adding more, we collect the oldest Noned entries | 13:52 |
| apuimedo | irenab: I'm okay with taking steps | 13:52 |
| apuimedo | I'm just thinking out loud whether files still make sense with a lean split cni | 13:52 |
| irenab | apuimedo, define lean split | 13:53 |
| ivc_ | irenab, daemon+exec | 13:53 |
| apuimedo | right | 13:53 |
| ivc_ | apuimedo, i'd prefer if we did not persist state on filesystem if that could be avoided | 13:54 |
| irenab | exec called by kubelet and transfers call to daemon, blocks till gets answer? | 13:54 |
| ivc_ | irenab, yes | 13:54 |
| irenab | apuimedo, so you considered Daemon to maintain files? | 13:54 |
| ivc_ | irenab, apuimedo, and no os-vif/oslo.<anything> in exec for sure | 13:54 |
| apuimedo | irenab: no. With the split | 13:55 |
| apuimedo | I was considering keeping a big dictionary on memory | 13:55 |
| irenab | files versus store in mem? | 13:55 |
| apuimedo | that keeps data for all the pods that get scheduled to the host | 13:55 |
| irenab | this whagt daemon will maintain | 13:56 |
| apuimedo | yup | 13:59 |
| apuimedo | and I was proposing a dict with LRU eviction that allows us to track double delete invokation | 13:59 |
| apuimedo | and saying that it is still more complex than a files based approach | 14:00 |
| apuimedo | but I guess we can try both | 14:00 |
| *** lezbar has joined #openstack-kuryr | 14:01 | |
| ivc_ | apuimedo, i don't think we need LRU eviction. its just a matter of how much memory-per pod our data structures will require and how many pods you can expect per node | 14:02 |
| apuimedo | ivc_: I understand the concern about keeping files | 14:02 |
| apuimedo | I've dealt with its pains in the past | 14:02 |
| apuimedo | ivc_: the LRU is to keep track for older non-existing things | 14:03 |
| ivc_ | apuimedo, not necessary. if we do not delete container-side interface but move it into some kuryr-managed netns, we could use it as source of truth | 14:05 |
| apuimedo | why would we not delete the ifaces when DELETE is called? | 14:06 |
| ivc_ | to get lightning-fast add on recycled ports | 14:07 |
| ivc_ | *ADD | 14:07 |
| apuimedo | oh, that | 14:07 |
| apuimedo | right | 14:07 |
| apuimedo | it flew out of my mind | 14:07 |
| apuimedo | :P | 14:07 |
| ivc_ | :P | 14:07 |
| apuimedo | gotta think about it a bit more | 14:07 |
| ivc_ | apuimedo, it would be similar to pre-creating ports in advance (something that was discussed already iirc) | 14:12 |
| apuimedo | yes, yes | 14:12 |
| ivc_ | apuimedo, irenab, also consider that it could be an option to get rid of annotations and instead use some sort of controller --> cni-daemon rpc | 14:16 |
| irenab | ivc_, apuimedo : did you run devstack lately? | 14:16 |
| ivc_ | irenab, with or without the cni patch? | 14:17 |
| irenab | ivc_, I am starting to realise wheat 'part of me' means :-) | 14:17 |
| irenab | ivc_, actually patch is not relevant, it fails for k8s-api | 14:17 |
| ivc_ | irenab, i think i've forgot to update kuryr.conf cni config (you need to add "kuryr_conf": "/etc/kuryr/kuryr.conf") | 14:17 |
| irenab | devstack does not get there far... k8s-api fails with Invalid Auth Config | 14:18 |
| apuimedo | irenab: I didn't run it with the cni patch | 14:18 |
| apuimedo | :O | 14:18 |
| irenab | apuimedo, even before this | 14:18 |
| ivc_ | irenab, i think you had the same error before | 14:19 |
| irenab | No RSA key provided | 14:19 |
| irenab | ivc_, I think it was nsenter, so added it a priori | 14:20 |
| irenab | apuimedo, now with test you add , we will have an indication if its supposed to work | 14:23 |
| ivc_ | irenab, well the test clearly states that it does not work :) | 14:24 |
| *** lmdaly has joined #openstack-kuryr | 14:26 | |
| ivc_ | irenab, apuimedo, btw maybe what we are seeing with cni could actually be a bug in k8s and its cni plugin | 14:31 |
| irenab | ivc_, I want to reproduce it in my devstack and post question on k8s slack | 14:33 |
| ivc_ | irenab, cool | 14:34 |
| irenab | once devstack is happy to serve me ... | 14:34 |
| ivc_ | :) | 14:34 |
| ivc_ | irenab, you'll also need to update cni/net.d/kuryr.conf with { "name": "kuryr", "type": "kuryr", "debug": true, "kuryr_conf": "/etc/kuryr/kuryr.conf" } | 14:35 |
| ivc_ | irenab, and probably something else :) i need to restack on fresh env to cleanup the deployment part of the patch | 14:37 |
| irenab | ivc_, share the details once you get some more gotchas :-() | 14:37 |
| ivc_ | irenab, i'm gonna do a clean restack today and will try to update the patch so it is usable | 14:38 |
| irenab | ivc_, thanks. I will continue tomorrow | 14:41 |
| *** irenab has quit IRC | 14:45 | |
| *** garyloug has joined #openstack-kuryr | 14:47 | |
| kristianjf | Hey guys, new to channel. I was going through some of your slides/doc but having trouble finding which openstack release is scheduled for k8s-cni integration. | 15:04 |
| kristianjf | Can anyone point me in the right direction? | 15:05 |
| ivc_ | kristianjf, you can probably get an official statement from apuimedo | 15:16 |
| kristianjf | ivc_: thanks! | 15:17 |
| *** dougbtv has joined #openstack-kuryr | 15:17 | |
| *** hongbin has joined #openstack-kuryr | 15:22 | |
| *** saneax is now known as saneax-_-|AFK | 15:23 | |
| *** tonanhngo has joined #openstack-kuryr | 15:33 | |
| openstackgerrit | Hongbin Lu proposed openstack/fuxi: Add devstack plugin for Fuxi https://review.openstack.org/402244 | 15:39 |
| *** oanson has quit IRC | 15:46 | |
| openstackgerrit | Darla Ahlert proposed openstack/kuryr-libnetwork: Unittest with mock https://review.openstack.org/394788 | 15:52 |
| openstackgerrit | Darla Ahlert proposed openstack/kuryr-libnetwork: Unittests with mock https://review.openstack.org/394251 | 16:00 |
| *** dimak has quit IRC | 16:28 | |
| *** limao_ has quit IRC | 16:36 | |
| *** pcaruana has quit IRC | 16:36 | |
| *** akanksha__ has joined #openstack-kuryr | 16:38 | |
| *** akanksha__ is now known as ApheleiaS | 16:41 | |
| openstackgerrit | Darla Ahlert proposed openstack/kuryr-libnetwork: Unittests with Mock and add a TestCase https://review.openstack.org/393996 | 16:54 |
| dougbtv | any chance anyone has seen kuryr being used with openshift? e.g. openshift on openstack with a shared neutron network between the two? | 16:58 |
| dougbtv | curious if there's any documentation on trying to get that up, if possible. | 16:59 |
| *** shashank_hegde has joined #openstack-kuryr | 17:02 | |
| *** shashank_hegde has quit IRC | 17:12 | |
| *** shashank_hegde has joined #openstack-kuryr | 17:16 | |
| *** neil_ has joined #openstack-kuryr | 17:16 | |
| *** neiljerram has quit IRC | 17:17 | |
| *** diogogmt has joined #openstack-kuryr | 17:31 | |
| *** neil_ has quit IRC | 17:44 | |
| *** irenab has joined #openstack-kuryr | 17:48 | |
| *** shashank_hegde has quit IRC | 17:51 | |
| *** kristianjf has quit IRC | 17:52 | |
| *** dougbtv has quit IRC | 17:58 | |
| *** neil_ has joined #openstack-kuryr | 18:00 | |
| *** jgriffith_away is now known as jgriffith | 18:05 | |
| *** irenab has quit IRC | 18:11 | |
| *** tonanhngo has quit IRC | 18:14 | |
| *** lmdaly has quit IRC | 18:20 | |
| *** shashank_hegde has joined #openstack-kuryr | 18:21 | |
| *** shashank_hegde has quit IRC | 18:25 | |
| openstackgerrit | Darla Ahlert proposed openstack/kuryr-libnetwork: Unittests with Mock and add a TestCase https://review.openstack.org/393996 | 18:27 |
| openstackgerrit | Darla Ahlert proposed openstack/kuryr-libnetwork: Unittests with Mock and add a TestCase https://review.openstack.org/393996 | 18:34 |
| *** jgriffith is now known as jgriffith_away | 18:40 | |
| *** shashank_hegde has joined #openstack-kuryr | 18:50 | |
| *** garyloug has quit IRC | 19:00 | |
| *** tonanhngo has joined #openstack-kuryr | 19:14 | |
| *** oanson has joined #openstack-kuryr | 19:27 | |
| *** lezbar has quit IRC | 19:47 | |
| *** jgriffith_away is now known as jgriffith | 19:56 | |
| *** dougbtv has joined #openstack-kuryr | 19:57 | |
| alraddarla_ | janonymous, ping ? | 20:36 |
| *** diogogmt has quit IRC | 20:40 | |
| *** diogogmt has joined #openstack-kuryr | 20:42 | |
| *** oanson has quit IRC | 20:43 | |
| *** ajo has quit IRC | 21:16 | |
| *** ajo has joined #openstack-kuryr | 21:17 | |
| *** ajo has quit IRC | 21:43 | |
| *** diogogmt has quit IRC | 21:46 | |
| *** ajo has joined #openstack-kuryr | 21:46 | |
| *** diogogmt has joined #openstack-kuryr | 22:12 | |
| *** ajo has quit IRC | 22:40 | |
| *** ajo has joined #openstack-kuryr | 22:41 | |
| *** diogogmt has quit IRC | 23:01 | |
| *** yuanying has joined #openstack-kuryr | 23:16 | |
| openstackgerrit | Hongbin Lu proposed openstack/fuxi: Add basic fullstack volume tests https://review.openstack.org/403941 | 23:18 |
| openstackgerrit | Hongbin Lu proposed openstack/fuxi: Add basic fullstack volume tests https://review.openstack.org/403941 | 23:22 |
| *** pmannidi has joined #openstack-kuryr | 23:35 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!