| *** XueFeng has quit IRC | 00:59 | |
| *** yanyanhu has joined #senlin | 01:13 | |
| *** ruijie has joined #senlin | 01:29 | |
| *** guoshan has joined #senlin | 01:30 | |
| *** guoshan has quit IRC | 01:35 | |
| *** guoshan has joined #senlin | 01:36 | |
| *** shu-mutou-AWAY is now known as shu-mutou | 01:53 | |
| *** guoshan has quit IRC | 02:42 | |
| *** yuanying has quit IRC | 02:47 | |
| *** guoshan has joined #senlin | 02:54 | |
| *** zhurong has joined #senlin | 03:03 | |
| *** zhurong has quit IRC | 03:13 | |
| openstackgerrit | RUIJIE YUAN proposed openstack/senlin master: lb_policy support CLUSTER_RECOVER(1) https://review.openstack.org/438889 | 03:24 |
|---|---|---|
| *** wllabs has joined #senlin | 03:34 | |
| wllabs | how to user node_operation | 03:34 |
| wllabs | i don't find any operation command in the client | 03:35 |
| *** yuanying has joined #senlin | 03:49 | |
| openstackgerrit | Merged openstack/senlin master: Refresh resource status before doing node_recover https://review.openstack.org/438379 | 04:03 |
| *** guoshan has quit IRC | 04:11 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-senlinclient master: Updated from global requirements https://review.openstack.org/439364 | 04:16 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/senlin master: Updated from global requirements https://review.openstack.org/439233 | 04:17 |
| yuanbin | Qiming, I see you fix bug about image or flavor for python-openstacksdk, but i use python-openstacksdb 0.9.5 version, the floavr apear The resource could not be found, use 0.9.10 the question not exist | 04:23 |
| openstackgerrit | XueFeng Liu proposed openstack/senlin master: Revise cluster/node check action records design https://review.openstack.org/435822 | 04:23 |
| yuanbin | Qiming, Sorry, I think I can be solved. | 04:46 |
| *** guoshan has joined #senlin | 05:10 | |
| Qiming | yuanbin, you are supposed to use the newer version | 05:26 |
| Qiming | python-openstacksdk>=0.9.13 | 05:26 |
| ruijie | hi Qiming | 05:27 |
| ruijie | you talked about deploying k8s with Senlin | 05:28 |
| ruijie | then what's the role of Senlin here? | 05:28 |
| Qiming | cluster manager | 05:29 |
| Qiming | managing a cluster of physical or virtual nodes on which kubernetes is used to manage containers/pods/services | 05:31 |
| *** guoshan has quit IRC | 05:32 | |
| ruijie | I see, so, we keep managing docker directly and also support managing k8s cluster to support docker? | 05:35 |
| Qiming | yes | 05:38 |
| *** guoshan has joined #senlin | 05:38 | |
| Qiming | the reason I'm looking into that is ... | 05:38 |
| Qiming | there are many users already deployed openstack for managing VMs and other peripheral resources | 05:39 |
| Qiming | most of these users are also considering managing containers as light-weight execution environments | 05:39 |
| Qiming | one option for them is to set aside some physical machines and deploy kubernetes to manage those containers | 05:40 |
| Qiming | that is fine. | 05:40 |
| Qiming | one drawback is that resources cannot be shared | 05:41 |
| Qiming | another one is the company/organization has to maintain two resource pools, one for VM, one for containers, two groups of people ... | 05:42 |
| Qiming | yet another one is more subtel | 05:42 |
| Qiming | subtle | 05:42 |
| Qiming | when you got more workloads running containers, you have to manually add nodes (physical/virtual) to the k8s pool | 05:42 |
| Qiming | when VM requests go up, you have to do the migration in the reverse direction ... | 05:43 |
| Qiming | it is ... very inconvenient at least | 05:43 |
| Qiming | for those openstack users who also wants to manage containerized workloads, we can provide them an integrated solution | 05:44 |
| Qiming | run k8s on a VM cluster managed by senlin | 05:44 |
| Qiming | when container workloads increase, senlin can help automatically expand the underlying VM cluster | 05:44 |
| Qiming | when the workloads decrease, senlin can help shrink the VM cluster automatically | 05:45 |
| Qiming | and we have VM level aut-healing, load-balancing etc for free | 05:45 |
| Qiming | it would be a great value-add for people to enjoy all the benefits of both OpenStack and kubernetes | 05:46 |
| ruijie | yes Qiming, the cloud foundry has the same problem | 05:48 |
| ruijie | but the load for controlling modules are usually not high | 05:48 |
| Qiming | right | 05:49 |
| Qiming | it is always about the worker nodes | 05:49 |
| *** guoshan has quit IRC | 05:53 | |
| *** guoshan has joined #senlin | 05:59 | |
| *** guoshan has quit IRC | 06:02 | |
| ruijie | Qiming, so the controlling modules of k8s also run in one cluster of Senlin? | 06:03 |
| Qiming | it depends | 06:03 |
| Qiming | you can create an orphan node for the controllers, which may not be recommended | 06:04 |
| Qiming | you can create a separate cluster for it | 06:04 |
| *** guoshan has joined #senlin | 06:04 | |
| ruijie | in our production environment, user always asked to let these modules run on one node mixed ... | 06:04 |
| Qiming | controllers and workers all run on a single physical machine? | 06:05 |
| ruijie | eg. 5 nodes to deploy controlling modules like routers, cloud controllers since cloud foundry has many modules. | 06:06 |
| ruijie | each of these modules has to be high available | 06:06 |
| Qiming | the node here is VM? | 06:09 |
| ruijie | yes Qiming | 06:09 |
| Qiming | that is fine | 06:09 |
| ruijie | like node1: router, cloud controller, node2: router, uaa, node3: cloud controller ... | 06:10 |
| Qiming | understood | 06:10 |
| ruijie | all these modules are deploy on VMs mixed, then is will be hard to provide auto-healing and auto-scaling | 06:10 |
| openstackgerrit | Qiming Teng proposed openstack/senlin master: Add filters to cluster_get_all_by_cluster() https://review.openstack.org/439405 | 06:12 |
| Qiming | I see | 06:12 |
| Qiming | what's the benefit of deploying load mixtures into each VM? | 06:13 |
| openstackgerrit | Qiming Teng proposed openstack/senlin master: Add filters to cluster_get_all_by_cluster() https://review.openstack.org/439405 | 06:16 |
| openstackgerrit | Qiming Teng proposed openstack/senlin master: Add filters to node_get_all_by_cluster() https://review.openstack.org/439405 | 06:17 |
| openstackgerrit | Qiming Teng proposed openstack/senlin master: Use node_get_all_by_cluster when possible https://review.openstack.org/439411 | 06:41 |
| openstackgerrit | Qiming Teng proposed openstack/senlin master: Improve node_ids_by_cluster() by adding filters https://review.openstack.org/439419 | 06:57 |
| *** XueFeng has joined #senlin | 07:00 | |
| XueFeng | yanyan, around? | 07:02 |
| openstackgerrit | Qiming Teng proposed openstack/senlin master: Use ids_by_cluster() object call when appropriate https://review.openstack.org/439423 | 07:09 |
| ruijie | sorry Qiming, just come back. Deploying load mixtures into each VM can reduce the resources costs... | 07:29 |
| Qiming | em ... not sure about it, ;) | 07:30 |
| ruijie | em, right. That depends .. | 07:33 |
| openstackgerrit | Nguyen Van Trung proposed openstack/senlin master: Change 'adanced' to 'advanced' https://review.openstack.org/439434 | 07:42 |
| *** yuanying has quit IRC | 08:01 | |
| *** yuanying has joined #senlin | 08:05 | |
| openstackgerrit | RUIJIE YUAN proposed openstack/senlin master: lb_policy support CLUSTER_RECOVER(1) https://review.openstack.org/438889 | 08:18 |
| *** XueFeng has quit IRC | 08:20 | |
| openstackgerrit | Merged openstack/senlin master: Change 'adanced' to 'advanced' https://review.openstack.org/439434 | 08:36 |
| *** root____3 has joined #senlin | 08:38 | |
| *** ruijie_ has joined #senlin | 08:45 | |
| *** ruijie has quit IRC | 08:48 | |
| *** yuanying has quit IRC | 08:51 | |
| *** yuanying has joined #senlin | 08:55 | |
| openstackgerrit | RUIJIE YUAN proposed openstack/senlin master: lb_policy support CLUSTER_RECOVER(1) https://review.openstack.org/438889 | 09:15 |
| *** yuanying has quit IRC | 09:16 | |
| *** yuanying has joined #senlin | 10:00 | |
| *** shu-mutou is now known as shu-mutou-AWAY | 10:05 | |
| openstackgerrit | RUIJIE YUAN proposed openstack/senlin master: lb_policy support CLUSTER_RECOVER(2) https://review.openstack.org/439515 | 10:13 |
| *** ruijie__ has joined #senlin | 10:19 | |
| *** ruijie_ has quit IRC | 10:20 | |
| *** yanyanhu has quit IRC | 10:20 | |
| openstackgerrit | RUIJIE YUAN proposed openstack/senlin master: lb_policy support CLUSTER_RECOVER(2) https://review.openstack.org/439515 | 10:22 |
| openstackgerrit | RUIJIE YUAN proposed openstack/senlin master: lb_policy support CLUSTER_RECOVER(1) https://review.openstack.org/438889 | 10:22 |
| *** Jeffrey4l has quit IRC | 10:30 | |
| *** guoshan has quit IRC | 10:30 | |
| *** root____3 has quit IRC | 10:43 | |
| *** Jeffrey4l has joined #senlin | 10:49 | |
| *** zhurong has joined #senlin | 14:01 | |
| openstackgerrit | Merged openstack/senlin master: Add filters to node_get_all_by_cluster() https://review.openstack.org/439405 | 14:02 |
| *** zhurong has quit IRC | 14:19 | |
| *** zhurong has joined #senlin | 14:21 | |
| *** zhurong has quit IRC | 15:31 | |
| *** ThiagoCMC has quit IRC | 16:14 | |
| *** Jeffrey4l has quit IRC | 17:59 | |
| *** wllabs has quit IRC | 19:18 | |
| *** openstackgerrit has quit IRC | 20:03 | |
| *** catintheroof has joined #senlin | 21:08 | |
| *** catintheroof has quit IRC | 21:10 | |
| *** catintheroof has joined #senlin | 21:10 | |
| *** catintheroof has quit IRC | 23:29 | |
| *** Guest27057 is now known as zigo | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!