*** Qiming has quit IRC | 00:48 | |
*** zzxwill has joined #senlin | 01:09 | |
*** zzxwill has quit IRC | 01:23 | |
*** zzxwill has joined #senlin | 01:24 | |
*** Qiming has joined #senlin | 01:24 | |
openstackgerrit | Qiming Teng proposed openstack/senlin: Remove a TODO item https://review.openstack.org/294356 | 01:28 |
---|---|---|
*** heyongli has joined #senlin | 01:44 | |
*** heyongli has quit IRC | 01:44 | |
*** heyongli has joined #senlin | 01:44 | |
*** Yanyanhu has joined #senlin | 01:49 | |
*** heyongli has quit IRC | 01:52 | |
openstackgerrit | Merged openstack/senlin: Remove a TODO item https://review.openstack.org/294356 | 01:52 |
*** heyongli has joined #senlin | 01:52 | |
*** zzxwill has quit IRC | 01:52 | |
*** zzxwill has joined #senlin | 01:54 | |
*** heyongli has quit IRC | 01:54 | |
*** heyongli has joined #senlin | 01:55 | |
*** heyongli has quit IRC | 01:56 | |
*** heyongli has joined #senlin | 01:57 | |
*** heyongli has quit IRC | 02:04 | |
*** heyongli has joined #senlin | 02:05 | |
openstackgerrit | Merged openstack/senlin: Fix problems with add timer and improve periodic tasks https://review.openstack.org/293897 | 02:13 |
*** heyongli has quit IRC | 02:15 | |
*** heyongli has joined #senlin | 02:15 | |
*** heyongli has quit IRC | 02:25 | |
*** heyongli has joined #senlin | 02:25 | |
*** heyongli has quit IRC | 02:35 | |
*** heyongli has joined #senlin | 02:36 | |
*** heyongli has quit IRC | 02:44 | |
*** heyongli has joined #senlin | 02:45 | |
*** heyongli has quit IRC | 02:45 | |
*** heyongli has joined #senlin | 02:46 | |
*** zzxwill has quit IRC | 02:50 | |
openstackgerrit | Qiming Teng proposed openstack/python-senlinclient: Fix event list sorting https://review.openstack.org/294373 | 02:51 |
*** heyongli has quit IRC | 02:52 | |
*** zzxwill has joined #senlin | 02:54 | |
xuhaiwei | Hi, everyone, just find something interesting, many openstack projects have their own icons, does Senlin have one? Or do we need one? | 03:01 |
*** Yanyanhu has quit IRC | 03:08 | |
*** Yanyanhu has joined #senlin | 03:09 | |
*** elynn has joined #senlin | 03:11 | |
xuhaiwei | found it in the Senlin introduction ppt | 03:14 |
xuhaiwei | is it a formal one? | 03:14 |
Qiming | it can be | 03:19 |
Qiming | :) | 03:19 |
Qiming | if there are some artists to contribute a better one, that would be great | 03:20 |
elynn | Hi can we support lbaasv2 in lb profile? | 03:27 |
elynn | lb policy | 03:28 |
zzxwill | Hi xuhaiwei, could you please help send me the Senlin introduction ppt? I might find some guy for the icon. | 03:36 |
*** zzxwill has quit IRC | 03:46 | |
*** elynn has quit IRC | 03:48 | |
xuhaiwei | zzxwill, I am back now | 04:06 |
xuhaiwei | how to send to you? | 04:06 |
*** zzxwill has joined #senlin | 04:07 | |
xuhaiwei | Qiming, I think the ppt you sent to me can be shared to others, right? | 04:07 |
xuhaiwei | zzxwill, just be back now | 04:07 |
*** shu-mutou has joined #senlin | 04:07 | |
xuhaiwei | how to send to you, zzxwill | 04:08 |
*** elynn has joined #senlin | 04:21 | |
*** zzxwill has quit IRC | 04:26 | |
*** elynn has quit IRC | 04:26 | |
*** elynn has joined #senlin | 04:26 | |
*** shu-mutou has quit IRC | 04:44 | |
*** zzxwill has joined #senlin | 04:58 | |
zzxwill | xuhaiwei zzxwill@gmail.com Thanks. | 04:59 |
Qiming | elynn, lb policy is based on lbaasv2 api | 05:02 |
elynn | good to know that. | 05:03 |
openstackgerrit | Qiming Teng proposed openstack/python-senlinclient: Fix event list sorting https://review.openstack.org/294373 | 05:06 |
xuhaiwei | zzxwill, the mail is sent to you | 05:07 |
*** zzxwill has quit IRC | 05:14 | |
*** zzxwill has joined #senlin | 05:20 | |
zzxwill | Received the mail, thx:) xuhaiwei. | 05:22 |
elynn | Hi Qiming , if I attach a lb policy to cluster, how can I get the VIP address? | 05:24 |
*** shu-mutou-AFK has joined #senlin | 05:26 | |
Qiming | oh, seems no way | 05:26 |
Qiming | Yanyanhu ? | 05:26 |
Qiming | seems we need to add an interface somehow | 05:27 |
Yanyanhu | hi, sorry | 05:28 |
Yanyanhu | elynn, this is problem | 05:28 |
Yanyanhu | now the vip address is not exposed... | 05:29 |
elynn | oh... sad to hear that... | 05:30 |
Qiming | this kind of data can be exposed via cluster.data | 05:30 |
elynn | I'm thinking if we can expose it in cluster data ? | 05:30 |
Qiming | if we want to keep the interface as simple as it is | 05:30 |
Yanyanhu | maybe we should store vip info into some places, e.g. metadata of lb policy | 05:31 |
elynn | Qiming, ha, we are in the same direction. | 05:31 |
Qiming | yes, elynn | 05:31 |
elynn | Since one policy can attach to multiple cluster | 05:31 |
Yanyanhu | hmm, cluster.data is better | 05:31 |
*** zzxwill has quit IRC | 05:32 | |
*** shu-mutou-AFK has quit IRC | 05:32 | |
Yanyanhu | or cluster_policy.data? | 05:32 |
Qiming | each policy binding creates a new binding object | 05:34 |
elynn | sounds good to me. | 05:34 |
xuhaiwei | if the IP is stored in cluster data, when the policy is detached, the cluster data will be updated? | 05:34 |
Yanyanhu | ah, sorry, it should be stored in cluster.data | 05:34 |
Yanyanhu | xuhaiwei, yes | 05:34 |
Qiming | yes, xuhaiwei, that is for sure | 05:34 |
Yanyanhu | since now all lb policy attachment related data are stored in cluster.data | 05:35 |
xuhaiwei | so the IP should be saved there too | 05:35 |
Yanyanhu | yes, I think so | 05:36 |
Yanyanhu | let me check the code to confirm it | 05:36 |
Qiming | currently, lb id is saved in binding data | 05:37 |
Yanyanhu | Qiming, yes, I was wrong | 05:37 |
Yanyanhu | again | 05:37 |
Qiming | that can be changed | 05:37 |
elynn | So we store lb id in cluster's data too? | 05:37 |
elynn | in the future? | 05:37 |
Qiming | but I think VIP is a cluster attribute | 05:37 |
xuhaiwei | excuse me, what is the binding data? | 05:38 |
Yanyanhu | xuhaiwei, cluster_policy object | 05:38 |
Yanyanhu | it's unique for each attachement between a cluster and a policy | 05:38 |
xuhaiwei | thanks, Yanyanhu | 05:38 |
Yanyanhu | Qiming, this is because we only allow one lb policy been attached to a cluster | 05:38 |
Qiming | no, one lb policy can be attached to mulitiple cluster | 05:39 |
Qiming | lb policy itself is only a specification | 05:39 |
Yanyanhu | yes, but one cluster can only be attached one lb policy? | 05:39 |
Qiming | the data it initialized is saved per-cluster-binding | 05:40 |
Qiming | Yanyanhu, don't think so | 05:40 |
Qiming | you can attach more than one, in theory | 05:40 |
Qiming | in aws, loadbalancer is a list for an asg | 05:40 |
xuhaiwei | no conflicts happen? | 05:40 |
Qiming | two lbs won't see each other | 05:40 |
Yanyanhu | oh, I thought only scaling policy can be attached like this | 05:41 |
Yanyanhu | yes, attaching more than one lb policies to a cluster makes sense | 05:41 |
Qiming | well, we didnt change singleton value for lb yet | 05:41 |
Yanyanhu | for cases, each node has more than one network interfaces | 05:41 |
Qiming | so we need to set lb policy's singleton to False | 05:42 |
Yanyanhu | yes, I think so | 05:42 |
Qiming | and store data visible to users in cluster.data | 05:42 |
Qiming | when you are showing a cluster | 05:42 |
xuhaiwei | two lb policies control two nics of one node? am I wrong? | 05:43 |
Qiming | you will get 'loadbalancers': [{'id': 'lb1', 'vip': '1.2.3.4'}] from its data | 05:43 |
Yanyanhu | Qiming, if so, we need distinguish the data from different polices | 05:43 |
Yanyanhu | ok | 05:44 |
xuhaiwei | Qiming, if two lb policies attachment is allowed, the 'id' should be a list | 05:44 |
Yanyanhu | so maybe we should follow this principle: storing data visible for user into cluster.data while storing data invisible for user into cluster_policy.data | 05:45 |
Qiming | it is a list, check the dict again | 05:45 |
Qiming | +1, Yanyanhu | 05:45 |
Yanyanhu | ok, I see | 05:45 |
xuhaiwei | ok, got it, 'loadbalancers' is a list | 05:45 |
elynn | How about data: {'loadbalancers': {'lb_id': {'vip': 8.8.8.8}}} | 05:45 |
Qiming | elynn, that's fine | 05:46 |
elynn | if loadbalancers is a list, it's not easy for heat to get specific VIP | 05:46 |
Qiming | right, you will have to iterate | 05:46 |
Qiming | two-leveled map improves usability | 05:47 |
Yanyanhu | Qiming, cluster.data has not been used as of now? | 05:47 |
Qiming | Yanyanhu, I think so | 05:47 |
Yanyanhu | hmm, need a design for its format | 05:47 |
Qiming | We have been reserving that 'data' field for cluster-level attributes | 05:48 |
Yanyanhu | for any possible attrs we want to expose to end user | 05:48 |
Yanyanhu | yes | 05:48 |
Qiming | loadbalancers is definitely a good candidate | 05:48 |
elynn | yes | 05:48 |
Yanyanhu | right | 05:48 |
Yanyanhu | ok, will think about it and propose a patch for it | 05:49 |
Qiming | great | 05:49 |
Qiming | I like elynn's proposal about data format | 05:50 |
Yanyanhu | let me add a item into mitaka workitem etherpad | 05:50 |
elynn | Cooooool~ thank you Yanyanhu | 05:50 |
Yanyanhu | ok, it looks good to me as well | 05:50 |
Yanyanhu | no problem :) | 05:50 |
Qiming | thanks, elynn, this is an important feature | 05:50 |
Yanyanhu | a start point of cluster level attr generalization :) | 05:51 |
Qiming | emm ... not there yet | 05:51 |
Yanyanhu | yes, not that far :) | 05:51 |
Qiming | this kind of a data is more of a constant | 05:52 |
*** xuhaiwei has quit IRC | 05:52 | |
Qiming | when we were talking about cluster wide attribute collection, it is more about dynamic ones that will demand several service calls | 05:52 |
Yanyanhu | I see | 05:54 |
openstackgerrit | Qiming Teng proposed openstack/senlin: Add size checking when adding nodes https://review.openstack.org/294403 | 06:01 |
openstackgerrit | Qiming Teng proposed openstack/senlin: Documentation for the region placement policy https://review.openstack.org/294063 | 06:05 |
openstackgerrit | Qiming Teng proposed openstack/senlin: Add size checking when deleting nodes https://review.openstack.org/294408 | 06:19 |
*** zzxwill has joined #senlin | 06:29 | |
*** zzxwill has quit IRC | 06:30 | |
*** zzxwill has joined #senlin | 06:37 | |
*** elynn has quit IRC | 06:50 | |
*** zzxwill has quit IRC | 06:50 | |
openstackgerrit | Qiming Teng proposed openstack/senlin: Set update_at to new value after resize https://review.openstack.org/294424 | 06:54 |
*** elynn has joined #senlin | 07:08 | |
*** elynn has quit IRC | 07:12 | |
*** elynn has joined #senlin | 07:12 | |
*** elynn_ has joined #senlin | 07:20 | |
*** elynn has quit IRC | 07:20 | |
openstackgerrit | Qiming Teng proposed openstack/senlin: Optimize event deserialization https://review.openstack.org/294426 | 07:28 |
*** elynn__ has joined #senlin | 07:29 | |
*** elynn_ has quit IRC | 07:32 | |
*** zzxwill has joined #senlin | 07:42 | |
*** elynn__ has quit IRC | 07:46 | |
*** zzxwill has quit IRC | 07:47 | |
*** zzxwill has joined #senlin | 07:47 | |
*** zzxwill has quit IRC | 07:54 | |
*** shu-mutou has joined #senlin | 07:58 | |
*** zzxwill has joined #senlin | 08:12 | |
*** zzxwill has quit IRC | 08:14 | |
Qiming | team, need your helps on reviewing some patch back logs | 08:15 |
Qiming | thanks | 08:15 |
Yanyanhu | got it | 08:16 |
*** zzxwill has joined #senlin | 08:19 | |
Qiming | a lot of dirty works need to be done to cut an RC1 and create a stable/mitaka branch | 08:22 |
Qiming | don't know how | 08:22 |
*** zzxwill has quit IRC | 08:22 | |
Yanyanhu | RC1? | 08:22 |
Qiming | yes, today is the deadline | 08:22 |
Yanyanhu | I see | 08:22 |
openstackgerrit | Yanyan Hu proposed openstack/senlin: Store random assigned VIP address to cluster data https://review.openstack.org/294450 | 08:24 |
Yanyanhu | leave for a while. Need to reboot my workstation | 08:25 |
*** Yanyanhu has quit IRC | 08:27 | |
*** zzxwill has joined #senlin | 08:29 | |
openstackgerrit | Merged openstack/senlin: Add size checking when adding nodes https://review.openstack.org/294403 | 08:31 |
openstackgerrit | Merged openstack/python-senlinclient: Fix event list sorting https://review.openstack.org/294373 | 08:36 |
*** zzxwill has quit IRC | 08:48 | |
openstackgerrit | Qiming Teng proposed openstack/senlin: Release notes for RC1 https://review.openstack.org/294462 | 08:54 |
openstackgerrit | Merged openstack/senlin: Set update_at to new value after resize https://review.openstack.org/294424 | 08:57 |
*** xuhaiwei has joined #senlin | 08:58 | |
xuhaiwei | hi QIming | 08:58 |
Qiming | yes | 08:58 |
xuhaiwei | about the weekly meeting for container cluster, when should we arrange it? | 08:59 |
Qiming | up to you | 08:59 |
Qiming | I'm mostly free | 08:59 |
xuhaiwei | should we do it by phone? | 08:59 |
Qiming | yep, by phone would be efficient | 09:00 |
xuhaiwei | ok | 09:00 |
xuhaiwei | we had it this Tuesday, what about the time for every week? | 09:00 |
Qiming | fine with me | 09:00 |
xuhaiwei | from 14:00~15:00 Beijing time ? | 09:01 |
Qiming | okay | 09:01 |
xuhaiwei | thanks | 09:01 |
*** Yanyanhu has joined #senlin | 09:01 | |
Qiming | Yanyanhu, the funtional test failure is weird | 09:04 |
Yanyanhu | 294408? | 09:04 |
Qiming | yep | 09:04 |
Yanyanhu | let me check it | 09:04 |
Qiming | dhellmann, online? | 09:05 |
Yanyanhu | Qiming, need to revise functional test I think since the request will be rejected in service layer rather than action layer now | 09:06 |
Yanyanhu | will work on it | 09:06 |
Qiming | the request should not be rejected? | 09:06 |
Yanyanhu | should be | 09:06 |
Yanyanhu | in current implementation, the request will be accepted | 09:07 |
Qiming | line 89 is trying to remove 2 nodes from the cluster | 09:07 |
Yanyanhu | although the action will fail | 09:07 |
Qiming | that should be fine | 09:07 |
Qiming | how could the target capacity become 0? | 09:08 |
Yanyanhu | oh, I didn't see the log in detail, I thought it was caused by the logic change of request rejection | 09:08 |
Yanyanhu | let me have a look | 09:08 |
Qiming | great, thanks | 09:08 |
Yanyanhu | no problem | 09:09 |
Qiming | seems there is a race condition | 09:13 |
Yanyanhu | you mean? | 09:14 |
Qiming | cluster's desired_capacity may be incorrectly updated | 09:14 |
Yanyanhu | you're right | 09:15 |
Yanyanhu | +------------------+--------------------------------------+ | 09:15 |
Yanyanhu | | Property | Value | | 09:15 |
Yanyanhu | +------------------+--------------------------------------+ | 09:15 |
Yanyanhu | | created_at | 2016-03-18T09:14:11 | | 09:15 |
Yanyanhu | | data | {} | | 09:15 |
Yanyanhu | | desired_capacity | 2 | | 09:15 |
Yanyanhu | | domain | None | | 09:15 |
Yanyanhu | functional test failed locally | 09:15 |
Yanyanhu | the cluster's desired_capacity is 2 now | 09:15 |
Qiming | after two nodes added ? | 09:16 |
Yanyanhu | I should add a check about cluster desired_capacity in that test | 09:17 |
Qiming | so we should not update desired_capacity in do_join() | 09:17 |
Yanyanhu | yes, I think so | 09:17 |
Qiming | it should be done in cluster_action | 09:17 |
Yanyanhu | yes | 09:17 |
Yanyanhu | really need to test cluster add/del node operations more completely | 09:19 |
Qiming | right | 09:19 |
Yanyanhu | they are not very common used operations currently | 09:19 |
Qiming | I'll work on fixing the desired_capacity update | 09:19 |
Qiming | please help improve func tests | 09:20 |
Yanyanhu | sure | 09:20 |
*** Yanyanhu has quit IRC | 09:30 | |
*** zzxwill has joined #senlin | 09:38 | |
*** zzxwill has quit IRC | 09:41 | |
*** zzxwill has joined #senlin | 09:42 | |
*** zzxwill has quit IRC | 09:50 | |
openstackgerrit | Qiming Teng proposed openstack/senlin: Fix desired_capacity computation https://review.openstack.org/294483 | 10:00 |
openstackgerrit | Merged openstack/senlin: Add size checking when deleting nodes https://review.openstack.org/294408 | 10:08 |
openstackgerrit | Qiming Teng proposed openstack/senlin: Fix desired_capacity computation https://review.openstack.org/294483 | 10:11 |
*** zzxwill has joined #senlin | 10:12 | |
openstackgerrit | Merged openstack/senlin: Documentation for the region placement policy https://review.openstack.org/294063 | 10:18 |
openstackgerrit | Merged openstack/senlin: Optimize event deserialization https://review.openstack.org/294426 | 10:20 |
*** zzxwill has quit IRC | 10:30 | |
*** shu-mutou is now known as shu-mutou-AFK | 10:34 | |
*** zzxwill has joined #senlin | 10:35 | |
*** zzxwill has quit IRC | 10:41 | |
*** zzxwill has joined #senlin | 10:42 | |
*** zzxwill has quit IRC | 10:49 | |
openstackgerrit | Qiming Teng proposed openstack/senlin: Store random assigned VIP address to cluster data https://review.openstack.org/294450 | 11:04 |
openstackgerrit | Merged openstack/senlin: Fix desired_capacity computation https://review.openstack.org/294483 | 11:12 |
openstackgerrit | Merged openstack/senlin: Store random assigned VIP address to cluster data https://review.openstack.org/294450 | 11:39 |
openstackgerrit | Merged openstack/senlin: Release notes for RC1 https://review.openstack.org/294462 | 11:45 |
*** Qiming has quit IRC | 11:58 | |
*** Yanyanhu has joined #senlin | 12:15 | |
*** Yanyanhu has quit IRC | 12:19 | |
*** Qiming has joined #senlin | 12:24 | |
openstackgerrit | Doug Hellmann proposed openstack/senlin: Update reno for stable/mitaka https://review.openstack.org/294565 | 12:34 |
openstackgerrit | Doug Hellmann proposed openstack/senlin: Update reno for stable/mitaka https://review.openstack.org/294565 | 12:35 |
dhellmann | Qiming : your RC1 is ready and your stable/mitaka branch is available. Please merge https://review.openstack.org/294565 and https://review.openstack.org/294564 to finish setting those up | 12:36 |
*** openstackgerrit has quit IRC | 12:48 | |
*** openstackgerrit has joined #senlin | 12:49 | |
Qiming | dhellmann, got it, thanks | 12:53 |
openstackgerrit | Merged openstack/senlin: Update reno for stable/mitaka https://review.openstack.org/294565 | 13:05 |
openstackgerrit | Yanyan Hu proposed openstack/senlin: Complete functional test of cluster_membership https://review.openstack.org/294588 | 13:33 |
*** zzxwill has joined #senlin | 14:01 | |
*** Kennan has quit IRC | 14:36 | |
*** Kennan has joined #senlin | 14:37 | |
openstackgerrit | zzxwill proposed openstack/python-senlinclient: Spelling mistakes on 'Clustering service command-line client' page https://review.openstack.org/294634 | 14:59 |
openstackgerrit | zzxwill proposed openstack/python-senlinclient: Spelling mistakes on 'Clustering service command-line client' page https://review.openstack.org/294634 | 15:27 |
*** Qiming has quit IRC | 16:27 | |
*** zzxwill has quit IRC | 16:41 | |
*** zzxwill has joined #senlin | 16:42 | |
*** zzxwill has quit IRC | 17:22 | |
*** zhenguo_ has quit IRC | 19:16 | |
*** zhenguo_ has joined #senlin | 19:17 | |
*** jdandrea has quit IRC | 20:21 | |
*** jdandrea has joined #senlin | 20:22 | |
*** dhellmann has quit IRC | 21:26 | |
*** dhellmann has joined #senlin | 21:26 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!