*** liusheng has quit IRC | 00:51 | |
openstackgerrit | liusheng proposed openstack/nimble: Add tempest tests for instance creation https://review.openstack.org/404552 | 01:14 |
---|---|---|
*** liusheng has joined #openstack-nimble | 01:15 | |
*** kevinz has joined #openstack-nimble | 01:22 | |
zhenguo | morning nimble! | 01:39 |
*** yuntongjin has joined #openstack-nimble | 01:40 | |
*** yuntongjin has quit IRC | 01:49 | |
*** yuntongjin has joined #openstack-nimble | 02:11 | |
shaohe_feng | zhenguo: morning. | 02:40 |
shaohe_feng | zhenguo: send you a email. | 02:40 |
shaohe_feng | zhenguo: about: | 02:41 |
zhenguo | shaohe_feng: yes, just received | 02:41 |
shaohe_feng | 1. can we support sanity retry? | 02:41 |
shaohe_feng | 2. progess notification | 02:41 |
zhenguo | shaohe_feng: Sorry, I have landed the taskflow codes :P | 02:42 |
shaohe_feng | zhenguo: These ideas just comes out from my creating failure on my evn. | 02:43 |
shaohe_feng | zhenguo: It does not matter. We can still improve it, until it is perfect. | 02:44 |
zhenguo | shaohe_feng: sure, hah | 02:45 |
zhenguo | shaohe_feng: I have replied the mail, | 02:45 |
zhenguo | shaohe_feng: please keep you env for some tests, you can try to add NetworkError to no_reschedule_exc_types in OnFailureRescheduleTask. | 02:52 |
shaohe_feng | zhenguo: Yes, let me try. Guess, your new updated, is more sanity and give better User Experience | 02:54 |
zhenguo | shaohe_feng: haah | 02:54 |
shaohe_feng | zhenguo: Have you addressed the DB updated with Node UUID? | 02:56 |
zhenguo | shaohe_feng: yes, I have removed the revert method in Scheduler Task | 02:58 |
shaohe_feng | zhenguo: Good. let me try, the building status will changed to error at last on my evn. | 02:58 |
zhenguo | shaohe_feng: but I find another issue, if multiple creating requests come, they will get the same node, as we don | 02:59 |
zhenguo | shaohe_feng: as we don't have a mechnism to handle access control to the node cache | 02:59 |
shaohe_feng | shaohe_feng: so we need to fix it. | 03:00 |
zhenguo | shaohe_feng: We need to add a lock before accessing to node cache | 03:00 |
zhenguo | shaohe_feng: Yes, please have a look at this https://review.openstack.org/#/c/412049/ | 03:00 |
zhenguo | shaohe_feng: I will test it first | 03:00 |
* zhenguo brb | 03:04 | |
shaohe_feng | zhenguo: How do test the concurrent creating? | 03:05 |
zhenguo | shaohe_feng: you can issue some creating request at the same time | 03:05 |
*** Jeffrey4l has quit IRC | 03:05 | |
*** Jeffrey4l has joined #openstack-nimble | 03:05 | |
zhenguo | shaohe_feng: maybe in a loop | 03:06 |
shaohe_feng | zhenguo: OK. a script. | 03:08 |
shaohe_feng | zhenguo: where did you config the taskflow storage backend? | 03:27 |
shaohe_feng | zhenguo: I did not find it in mysql DB. | 03:28 |
zhenguo | shaohe_feng: I don't config it anywhere, just the default value | 03:28 |
*** yuntongjin has quit IRC | 04:26 | |
*** Jeffrey4l has quit IRC | 04:53 | |
shaohe_feng | zhenguo: what's the default, you tell me it can persistent. | 05:09 |
shaohe_feng | zhenguo: strange | 05:10 |
zhenguo | shaohe_feng: I'm not sure, but on my test, when restart the engine service, it seems continue the flow, but I didn't test it deeply | 05:10 |
shaohe_feng | zhenguo: the Doc said, if no backed provide, it will be in-memory. | 05:10 |
zhenguo | shaohe_feng: weird, we can test it again | 05:11 |
shaohe_feng | zhenguo: So it is configure the backend as persistent. | 05:11 |
zhenguo | shaohe_feng: not sure | 05:11 |
shaohe_feng | zhenguo: http://paste.openstack.org/show/592752/ | 05:12 |
shaohe_feng | zhenguo: you can see the NOTE(harlowja) | 05:12 |
shaohe_feng | zhenguo: I know the storage default id in-memory | 05:13 |
shaohe_feng | zhenguo: but flow engine configure it as persistent as default | 05:13 |
shaohe_feng | zhenguo: Though, we do not know where does it persistent. :) | 05:14 |
zhenguo | shaohe_feng: seems it is | 05:14 |
shaohe_feng | zhenguo: so nimble restart, and the engine initiate, find the peresitent task flow status | 05:17 |
shaohe_feng | zhenguo: It will continue the former interupted task. | 05:18 |
shaohe_feng | zhenguo: This is really good, the task can go on even restart. | 05:19 |
shaohe_feng | zhenguo: and we worried about this edge case before. | 05:19 |
shaohe_feng | zhenguo: now no worry about it. | 05:19 |
zhenguo | shaohe_feng: I'm not quite sure about that, need to do more test | 05:21 |
shaohe_feng | zhenguo: I just guess. | 05:42 |
shaohe_feng | zhenguo: we can confirm it. It is good to mention it this weekend meeting up to promote nimble. | 05:45 |
*** yuntongjin has joined #openstack-nimble | 05:47 | |
openstackgerrit | Zhenguo Niu proposed openstack/nimble: Following up patch of 5a89faf496dcb3046d149d11f4d76b399ee63702 https://review.openstack.org/412293 | 06:06 |
zhenguo | shaohe_feng: sorry, seems the node_uuid issue is because I don't save the instance object, I just added a following up patch to fix it and add NetworkError in no reschedule exception types | 06:08 |
zhenguo | shaohe_feng: please test with the patch | 06:08 |
shaohe_feng | zhenguo: I guess, you did not save the instance. | 06:28 |
zhenguo | shaohe_feng: hah | 06:28 |
shaohe_feng | zhenguo: So sorry, I'm planed to go throug the code last weekend. | 06:29 |
zhenguo | shaohe_feng: but the Schedule revert is also a reason, so I removed it | 06:29 |
shaohe_feng | zhenguo: so many personal affairs | 06:29 |
shaohe_feng | zhenguo: really sorry. | 06:30 |
shaohe_feng | zhenguo: have a baby, can work works on home. | 06:30 |
shaohe_feng | zhenguo: I have disscuss with Rai. | 06:31 |
zhenguo | shaohe_feng: np :) | 06:31 |
shaohe_feng | zhenguo: He tell me some requirement of CMCC | 06:31 |
shaohe_feng | zhenguo: I will send a email. | 06:31 |
zhenguo | shaohe_feng: oh, many thanks | 06:32 |
shaohe_feng | zhenguo: send the mail | 06:42 |
shaohe_feng | zhenguo: He discuss with CMCC | 06:42 |
shaohe_feng | zhenguo: and CMCC's requirement. | 06:42 |
zhenguo | shaohe_feng: Thanks | 06:47 |
openstackgerrit | Zhenguo Niu proposed openstack/nimble: Fix node NotFound issue when getting an instance https://review.openstack.org/412314 | 06:58 |
* zhenguo brb | 07:09 | |
openstackgerrit | liusheng proposed openstack/python-nimbleclient: Support retuning detailed info of "instance list" command https://review.openstack.org/412326 | 07:29 |
zhenguo | liusheng, shaohe_feng: do you think we need a taskflow when deleting an instance? | 07:33 |
liusheng | zhenguo: I think it is needed | 07:33 |
zhenguo | liusheng: ok, but how to revert if we failed to delete some resources | 07:34 |
zhenguo | liusheng: I mean, if we fail to delete some resources, how to handle the instance | 07:34 |
liusheng | zhenguo: we need to clean info, set state, during the process, and may revert | 07:34 |
liusheng | zhenguo: hmm, may set it to "error" ? | 07:35 |
zhenguo | liusheng: seems we should always be successful to delete an instance | 07:35 |
zhenguo | liusheng: you mean we should fail the deleting request if there's something wrong with deleting other resources? | 07:36 |
liusheng | zhenguo: yeah, but in which situation the resource related to the instance will be deleted failed ? | 07:39 |
zhenguo | liusheng: like when there's something wrong when you remove the port from neutron | 07:39 |
liusheng | zhenguo: just asked Zhenyu, he said in Nova, if resource related instance deleted failed, Nova will only record warning log .. | 07:40 |
zhenguo | liusheng: seems reasonable | 07:41 |
liusheng | zhenguo: does taskflow can also be used to display progress ? | 07:41 |
zhenguo | liusheng: not sure, shaohe_feng mentioned it can provide progress | 07:42 |
liusheng | zhenguo: cool | 07:47 |
zhenguo | liusheng: hah | 07:48 |
liusheng | zhenguo: may we can refer to other projects which use taskflow ? | 07:48 |
zhenguo | liusheng: ye | 07:49 |
zhenguo | liusheng: but now we use like what cinder does | 07:49 |
liusheng | zhenguo: so copy cinder!, hah | 07:50 |
zhenguo | liusheng: but cinder only use taskflow for creating, and not provide progress | 07:51 |
zhenguo | liusheng: what you mean by progress | 07:51 |
zhenguo | liusheng: it can log every task's status and revert reason | 07:51 |
liusheng | zhenguo: any related api to show the progress of action ? | 07:52 |
zhenguo | liusheng: maybe you can find from the email send from shaohe_feng | 07:52 |
openstackgerrit | Zhenguo Niu proposed openstack/nimble: Following up patch of 5a89faf496dcb3046d149d11f4d76b399ee63702 https://review.openstack.org/412293 | 07:53 |
liusheng | zhenguo: hmm, may need to dig the code of taskflow | 07:57 |
zhenguo | liusheng: yes, but no rush | 07:58 |
zhenguo | liusheng: I find you update the tempest patch again, do you have some new clues? | 07:59 |
liusheng | zhenguo: no :( | 07:59 |
liusheng | zhenguo: just handle the merge conflict | 07:59 |
zhenguo | liusheng: hah, it's really a tricky one | 08:00 |
openstackgerrit | liusheng proposed openstack/python-nimbleclient: Support retuning detailed info of "instance list" command https://review.openstack.org/412326 | 08:06 |
openstackgerrit | Zhenguo Niu proposed openstack/nimble: Fix node NotFound issue when getting an instance https://review.openstack.org/412314 | 08:06 |
liusheng | zhenguo: seems we don't have time to care nimbleclient :( | 08:07 |
zhenguo | liusheng: hah | 08:08 |
zhenguo | liusheng: I can test it later, maybe we can just merge the patches directly :P | 08:09 |
zhenguo | liusheng: and seems it's easy to change the CLI to 'openstack baremetal server' | 08:11 |
liusheng | zhenguo: about that, one way we can only change the command, another way, we replate the "instance" word with server, which way you prefer ? | 08:12 |
zhenguo | liusheng: I totally agree with replacing instance with server | 08:13 |
zhenguo | liusheng: and I like the short command :P | 08:13 |
liusheng | zhenguo: ok, I will do that, sir | 08:15 |
zhenguo | liusheng: thanks | 08:15 |
liusheng | zhenguo: so how about "type" ? lo | 08:18 |
liusheng | zhenguo: flavor ? | 08:18 |
zhenguo | liusheng: I think it's ok to use flavor | 08:18 |
zhenguo | liusheng: and for nimble API, do you think we should also replace 'instance' with 'server' and 'type' with 'flavor' | 08:19 |
liusheng | zhenguo: I am ok with that, what't shaohe's opinion ? @shaohe_feng | 08:23 |
shaohe_feng | liusheng: I'm here. | 08:27 |
shaohe_feng | so many message. | 08:27 |
liusheng | zhenguo: what do you think to replace "instance" with "server", and replace "instance type" with "flavor" ? | 08:28 |
shaohe_feng | liusheng: sounds good. | 08:28 |
shaohe_feng | zhenguo: server is more meaningful thant instance. | 08:28 |
liusheng | shaohe_feng: hah | 08:28 |
zhenguo | liusheng: as nova use server and flavor in his url as well | 08:28 |
zhenguo | shaohe_feng: yes | 08:28 |
shaohe_feng | liusheng: zhenguo: "Unknown column 'instance_type_extra_specs_1.instance_type_uuid' in 'field list' | 08:29 |
zhenguo | and users has already get used to use server and flavor | 08:29 |
liusheng | shaohe_feng: it is main because "baremetal compute instance" it longer than "baremetal server" | 08:29 |
shaohe_feng | zhenguo: yes. | 08:29 |
zhenguo | shaohe_feng: something wrong? | 08:30 |
liusheng | shaohe_feng: oh, I have change the db schema, you may need to clean db and run nimble-dbsync upgrade | 08:30 |
shaohe_feng | liusheng: server here does mean "compute instance" | 08:30 |
shaohe_feng | zhenguo: any change for the DB recently? | 08:30 |
shaohe_feng | liusheng: OK. got it. | 08:30 |
liusheng | shaohe_feng: openstack baremetal compute instance xxx --> openstack baremetal server xxx | 08:30 |
zhenguo | shaohe_feng: yes, as liusheng mentioned, he has changed the DB, | 08:30 |
liusheng | shaohe_feng: to make the uuid as primary key | 08:31 |
shaohe_feng | liusheng: zhenguo: let me update the DB. | 08:31 |
zhenguo | liusheng, shaohe_feng: I like 'openstack baremetal server create' hah | 08:31 |
liusheng | shaohe_feng: but please note, since I modified the initial upgrade script, now there is only one db upgrade script | 08:32 |
liusheng | shaohe_feng: you may need to manully clean db firstly | 08:32 |
zhenguo | yes, you may just deleted the nimble db | 08:33 |
shaohe_feng | zhenguo: liusheng: so I can not use "nimble-dbsync upgrade" | 08:34 |
shaohe_feng | right? | 08:34 |
zhenguo | shaohe_feng: yes | 08:34 |
liusheng | shaohe_feng: yes | 08:34 |
zhenguo | shaohe_feng: as we just modify the initial revision | 08:34 |
zhenguo | shaohe_feng: no new revision, so no upgrade | 08:34 |
shaohe_feng | zhenguo: got it. | 08:35 |
*** yuntongjin has quit IRC | 08:38 | |
liusheng | zhenguo: I am afraid there will be many conflicts if I cook the replacing change, since I have several patches in review, hah | 08:40 |
zhenguo | liusheng: so we can merge them first, hah | 08:40 |
shaohe_feng | liusheng: zhenguo: should mysql create nimble database with utf8 | 08:40 |
zhenguo | shaohe_feng: seems yes | 08:41 |
shaohe_feng | zhenguo: oh, $ openstack baremetal compute type list | 08:50 |
shaohe_feng | zhenguo: my drop the nimbe DB manully, | 08:51 |
zhenguo | shaohe_feng: so no default type | 08:51 |
shaohe_feng | zhenguo: so the type disappere. | 08:51 |
shaohe_feng | zhenguo: yes, devstack help me to create one. | 08:52 |
zhenguo | shaohe_feng: you can check the type uuid from ironic node's properties, than create the type again | 08:52 |
zhenguo | shaohe_feng: with the uuid we set to ironic node | 08:52 |
zhenguo | s/than/then | 08:52 |
shaohe_feng | zhenguo: OK. | 08:52 |
zhenguo | shaohe_feng: I'm now restacking, hah | 08:53 |
* zhenguo brb | 08:54 | |
openstackgerrit | liusheng proposed openstack/nimble: Use oslo context instead of the native context implementation https://review.openstack.org/412350 | 08:56 |
shaohe_feng | zhenguo: why are you restacking? | 09:10 |
zhenguo | shaohe_feng: my env is too old, and I want to refresh it, lol | 09:10 |
zhenguo | shaohe_feng: I think restack is nore easy to manually upgrade | 09:11 |
shaohe_feng | zhenguo: hope full everything is OK> | 09:11 |
zhenguo | shaohe_feng: hah | 09:11 |
shaohe_feng | zhenguo: yes, I think so. there maybe something error if manually upgrade by my experience | 09:12 |
zhenguo | shaohe_feng: yes, so I choose to restack it | 09:12 |
shaohe_feng | zhenguo: I just need to check the instance_type from ironic node's properties. | 09:12 |
shaohe_feng | zhenguo: no type uuid in it. | 09:13 |
zhenguo | shaohe_feng: no instance_type in node's properties? | 09:13 |
shaohe_feng | zhenguo: just instance_type in node's properties | 09:13 |
shaohe_feng | zhenguo: no instance_type uuid. | 09:13 |
zhenguo | shaohe_feng: yes, it's instance_type, you mean no value of instance_type key? | 09:14 |
shaohe_feng | zhenguo: has value. | 09:14 |
shaohe_feng | zhenguo: I got you. | 09:14 |
zhenguo | shaohe_feng: yes, it's instance_type's value | 09:15 |
shaohe_feng | zhenguo: shaohe_feng: you can check the type uuid from ironic node's properties, than create the type again | 09:15 |
shaohe_feng | zhenguo: typo | 09:15 |
zhenguo | shaohe_feng: I have correct it after that | 09:15 |
zhenguo | s/than/then | 09:15 |
zhenguo | shaohe_feng: haha | 09:16 |
shaohe_feng | zhenguo: "type uuid" in it :) | 09:16 |
zhenguo | shaohe_feng: oh, yes | 09:16 |
shaohe_feng | zhenguo: so I'm looking for the UUID. :) | 09:16 |
zhenguo | shaohe_feng: lol, sorry for the misleading | 09:17 |
openstackgerrit | liusheng proposed openstack/nimble: Use oslo context instead of the native context implementation https://review.openstack.org/412350 | 09:17 |
zhenguo | shaohe_feng: I can take the floating ip association/disassociation task | 09:17 |
shaohe_feng | zhenguo: Haha. I should guess this is a typo. We will not store type uuid in ironic. | 09:17 |
shaohe_feng | zhenguo: greate. | 09:17 |
shaohe_feng | zhenguo: update the task track. | 09:18 |
zhenguo | shaohe_feng: we set the instance type name there, but seems there's a potential issue as we allow dupliate instance type name | 09:19 |
shaohe_feng | zhenguo: yes. we need to distingued the node by type. | 09:20 |
shaohe_feng | zhenguo: different tanent can name "small", but maybe different config. | 09:24 |
zhenguo | shaohe_feng: yes | 09:24 |
shaohe_feng | zhenguo: How scheduler to distiguish them? | 09:24 |
zhenguo | shaohe_feng: no way, | 09:24 |
zhenguo | shaohe_feng: so I want to change to use uuid | 09:25 |
shaohe_feng | zhenguo: yes. UUID is machine friendly | 09:26 |
shaohe_feng | zhenguo: then wo should not let name dupliated for same tanent. | 09:27 |
shaohe_feng | zhenguo: but it can be dupliated for different tanents. | 09:28 |
zhenguo | shaohe_feng: yes, I mean set instance type uuid to ironic node instead | 09:28 |
shaohe_feng | zhenguo: I have got you. :) | 09:29 |
zhenguo | shaohe_feng: hah | 09:29 |
zhenguo | shaohe_feng: my env is ok now, how about your's | 09:29 |
shaohe_feng | zhenguo: DB is OK. I can create instance. | 09:30 |
zhenguo | shaohe_feng: ok | 09:30 |
shaohe_feng | zhenguo: but it's error, it does not matter. I use it to test your create patch. | 09:31 |
zhenguo | shaohe_feng: ok, | 09:31 |
shaohe_feng | zhenguo: the instance status is error. | 09:31 |
shaohe_feng | zhenguo: IRC is not a fast way for chinese communication. | 09:31 |
zhenguo | shaohe_feng: lol | 09:32 |
zhenguo | shaohe_feng: maybe we need to practice more | 09:32 |
shaohe_feng | zhenguo: yes. | 09:32 |
shaohe_feng | zhenguo: when will come here to setup nimbe on the hpysical node. | 09:33 |
zhenguo | shaohe_feng: hah, nearly forget it | 09:33 |
shaohe_feng | zhenguo: welcome any time. | 09:34 |
shaohe_feng | zhenguo: No valid node is found with the request spec. | 09:34 |
zhenguo | shaohe_feng: I'm really busy these days, it's nearly the end of the year | 09:34 |
zhenguo | shaohe_fent: oh | 09:35 |
zhenguo | shaohe_feng: maybe it's better to restack | 09:35 |
shaohe_feng | zhenguo: Provisioning State" is "clean failed" | 09:35 |
shaohe_feng | zhenguo: How can I clean Provisioning State? | 09:35 |
shaohe_feng | zhenguo: just by ironic CLI? | 09:36 |
zhenguo | shaohe_feng: update db | 09:36 |
shaohe_feng | zhenguo: what's the "Provisioning State" is OK? | 09:40 |
shaohe_feng | zhenguo: NULL? | 09:40 |
zhenguo | shaohe_feng: available | 09:40 |
shaohe_feng | zhenguo: OK. manually change it. | 09:41 |
openstackgerrit | liusheng proposed openstack/nimble: Use oslo context instead of the native context implementation https://review.openstack.org/412350 | 09:49 |
shaohe_feng | zhenguo: TRACE nimble.engine.manager raise exception.NoValidNode(_("No weighed nodes available")) | 09:54 |
shaohe_feng | zhenguo: I guess nimble scheduler just get the node info from cache | 09:55 |
shaohe_feng | zhenguo: so I update the ironic DB, I need to wait for nimbe to fresh the DB. | 09:55 |
shaohe_feng | zhenguo: right? | 09:55 |
zhenguo | shaohe_feng: yes | 09:55 |
zhenguo | shaohe_feng: or you can restart the engine service | 09:55 |
zhenguo | liusheng: why this happened https://bugs.launchpad.net/nimble/+bug/1650462 ? | 09:56 |
openstack | Launchpad bug 1650462 in Nimble "AttributeError will be raise when list available zones" [Undecided,New] - Assigned to Liusheng (liusheng) | 09:56 |
shaohe_feng | zhenguo: I have wait for some time. still same error. | 09:56 |
zhenguo | liusheng: I find this issue when testing the az list command | 09:56 |
shaohe_feng | zhenguo: let me restart the engine. | 09:57 |
zhenguo | shaohe_feng: ok | 09:57 |
*** kevinz has quit IRC | 09:57 | |
shaohe_feng | zhenguo: but I should not restart. | 09:57 |
shaohe_feng | zhenguo: any bug? | 09:57 |
liusheng | zhenguo: it is because the self.node_cache is a list previously, and the, it has been changed to a dict | 09:57 |
zhenguo | shaohe_feng: seems still some wrong with your ironic nodes | 09:57 |
liusheng | zhenguo: https://review.openstack.org/#/c/411655/ | 09:58 |
zhenguo | liusheng: but on my env it's a dict indeed, and I remember you have fixed it | 09:58 |
shaohe_feng | zhenguo: OK. | 09:58 |
shaohe_feng | zhenguo: restart not OK> | 09:58 |
shaohe_feng | zhenguo: 2016-12-19 05:50:54.812 TRACE nimble.engine.manager NoValidNode: No weighed nodes available │ | 09:58 |
liusheng | zhenguo: sorry, I didn't add the Closes-bug | 09:58 |
liusheng | zhenguo: yes | 09:58 |
zhenguo | liusheng: but not sure why I still have the issue | 09:59 |
liusheng | zhenguo: really | 09:59 |
liusheng | zhenguo: ? | 09:59 |
liusheng | zhenguo: it has been fixed in my env | 09:59 |
zhenguo | liusheng: yes, let me check | 09:59 |
zhenguo | liusheng, shaohe_feng: I have to go for dinner, will be back soon | 10:00 |
liusheng | zhenguo: enjoy :) | 10:01 |
zhenguo | liusheng: thanks | 10:01 |
shaohe_feng | zhenguo: enjoy. Will go home. and be on line at home. | 10:04 |
*** Jeffrey4l has joined #openstack-nimble | 10:24 | |
*** _liusheng has quit IRC | 11:37 | |
*** kevinz has joined #openstack-nimble | 12:09 | |
*** Jeffrey4l has quit IRC | 12:10 | |
*** Jeffrey4l has joined #openstack-nimble | 12:43 | |
openstackgerrit | Merged openstack/nimble: Following up patch of 5a89faf496dcb3046d149d11f4d76b399ee63702 https://review.openstack.org/412293 | 13:10 |
openstackgerrit | Merged openstack/nimble: Fix node NotFound issue when getting an instance https://review.openstack.org/412314 | 13:26 |
* zhenguo goes home | 13:27 | |
*** kevinz has quit IRC | 14:50 | |
openstackgerrit | Merged openstack/nimble: Add locks when accessing node cache https://review.openstack.org/412049 | 16:48 |
*** Jeffrey4l has quit IRC | 18:07 | |
*** openstackstatus has quit IRC | 19:13 | |
*** openstack has joined #openstack-nimble | 19:17 | |
*** l4yerffeJ_ has joined #openstack-nimble | 21:33 | |
*** l4yerffeJ has quit IRC | 21:35 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!