| *** bobh has quit IRC | 00:15 | |
| openstackgerrit | Merged openstack/python-tackerclient: Updated from global requirements https://review.openstack.org/254577 | 00:42 |
|---|---|---|
| *** gongysh has joined #tacker | 00:50 | |
| *** sripriya has quit IRC | 01:25 | |
| *** tbh has joined #tacker | 01:30 | |
| *** u_kozat has quit IRC | 01:36 | |
| *** gongysh has quit IRC | 01:39 | |
| *** lhcheng has joined #tacker | 01:44 | |
| *** lhcheng has quit IRC | 01:49 | |
| *** santoshk has quit IRC | 01:56 | |
| *** tbh has quit IRC | 02:05 | |
| sridhar_ram | s3wong: I just replied http://lists.openstack.org/pipermail/openstack-dev/2015-December/081921.html | 02:29 |
| s3wong | sridhar_ram: cool | 02:29 |
| s3wong | sridhar_ram: note that I am not for or against working with them --- just think that we should reply | 02:29 |
| sridhar_ram | s3wong: sure.. | 02:30 |
| *** sridhar_ram has quit IRC | 02:41 | |
| *** s3wong has quit IRC | 02:55 | |
| *** sripriya has joined #tacker | 03:11 | |
| *** sripriya_ has joined #tacker | 03:21 | |
| *** sripriya has quit IRC | 03:24 | |
| *** gongysh has joined #tacker | 03:32 | |
| *** prashantD has quit IRC | 03:32 | |
| *** gongysh has quit IRC | 03:37 | |
| *** gongysh has joined #tacker | 03:52 | |
| *** lhcheng has joined #tacker | 03:58 | |
| *** cartik has joined #tacker | 04:01 | |
| *** changzhi has joined #tacker | 04:03 | |
| *** lhcheng has quit IRC | 04:03 | |
| *** gongysh has quit IRC | 04:06 | |
| cartik | keystone not starting while trying to install tacker, here is the log | 04:07 |
| cartik | http://paste.openstack.org/show/481416/ | 04:07 |
| sripriya_ | cartik: there may be a process already running on port 5000, kill the listening process and restart keystone | 04:09 |
| *** gongysh has joined #tacker | 04:13 | |
| cartik | thanks sripriya_ , let me try | 04:16 |
| *** gongysh has quit IRC | 04:17 | |
| *** tbh has joined #tacker | 04:19 | |
| *** gongysh has joined #tacker | 04:22 | |
| *** amotoki has quit IRC | 04:25 | |
| *** cartik has quit IRC | 05:39 | |
| *** gongysh_ has joined #tacker | 07:23 | |
| *** gperal has joined #tacker | 07:46 | |
| *** karimb has joined #tacker | 08:13 | |
| *** egonzalez has joined #tacker | 08:14 | |
| *** mbound_ has joined #tacker | 08:23 | |
| *** lhcheng has joined #tacker | 08:38 | |
| openstackgerrit | gongysh proposed openstack/tacker: Remove RPC related stuff https://review.openstack.org/255732 | 08:41 |
| *** gongysh has quit IRC | 08:42 | |
| openstackgerrit | gongysh proposed openstack/tacker: Remove RPC related stuff https://review.openstack.org/255732 | 08:43 |
| openstackgerrit | gongysh proposed openstack/tacker: Remove neutron agents codes https://review.openstack.org/255737 | 08:48 |
| openstackgerrit | gongysh proposed openstack/tacker: Remove nova notification settings https://review.openstack.org/255743 | 08:55 |
| *** tbh has quit IRC | 09:06 | |
| openstackgerrit | gongysh proposed openstack/tacker: Remove Service instance, context resources https://review.openstack.org/250156 | 09:09 |
| *** zeih has joined #tacker | 09:18 | |
| *** tbh has joined #tacker | 09:22 | |
| *** openstackgerrit has quit IRC | 09:32 | |
| *** openstackgerrit has joined #tacker | 09:33 | |
| *** gperal has quit IRC | 09:37 | |
| openstackgerrit | gongysh proposed openstack/tacker: Remove RPC server and client related stuff https://review.openstack.org/255732 | 09:48 |
| *** gperal has joined #tacker | 09:48 | |
| *** igordcard has joined #tacker | 09:49 | |
| openstackgerrit | gongysh proposed openstack/tacker: Remove RPC server and client related stuff https://review.openstack.org/255732 | 09:50 |
| openstackgerrit | gongysh proposed openstack/tacker: Remove neutron agents codes https://review.openstack.org/255737 | 09:50 |
| *** igordcard has quit IRC | 09:58 | |
| *** igordcard has joined #tacker | 09:59 | |
| *** igordcard_ has joined #tacker | 10:04 | |
| *** igordcard has quit IRC | 10:07 | |
| openstackgerrit | Aqsa proposed openstack/tacker: Update local.conf.example for OpenWRT image pre-installation https://review.openstack.org/255780 | 10:24 |
| *** lhcheng has quit IRC | 10:33 | |
| *** igordcar1_ has joined #tacker | 10:50 | |
| *** igordcar1_ has quit IRC | 10:50 | |
| *** igordcar1_ has joined #tacker | 10:50 | |
| *** igordcard_ has quit IRC | 10:50 | |
| *** igordcar1_ has quit IRC | 10:50 | |
| *** changzhi has quit IRC | 11:01 | |
| *** igordcard_ has joined #tacker | 11:20 | |
| *** mbound_ has quit IRC | 12:45 | |
| *** openstackgerrit has quit IRC | 12:47 | |
| *** openstackgerrit has joined #tacker | 12:48 | |
| *** dmellado has quit IRC | 12:57 | |
| *** dmellado has joined #tacker | 12:57 | |
| openstackgerrit | gongysh proposed openstack/tacker: Unify the driver config opts https://review.openstack.org/255160 | 13:01 |
| *** gongysh has joined #tacker | 13:02 | |
| *** karimb has quit IRC | 13:02 | |
| *** tbh has quit IRC | 13:19 | |
| *** karimb has joined #tacker | 13:19 | |
| *** gongysh has quit IRC | 14:13 | |
| *** bobh has joined #tacker | 14:33 | |
| *** zeih has quit IRC | 15:21 | |
| *** dandruta has joined #tacker | 15:27 | |
| *** zeih has joined #tacker | 15:33 | |
| *** tbh has joined #tacker | 16:01 | |
| *** zeih has quit IRC | 16:13 | |
| *** zeih has joined #tacker | 16:36 | |
| *** zeih has quit IRC | 16:40 | |
| *** lhcheng has joined #tacker | 16:52 | |
| *** sridhar_ram has joined #tacker | 16:55 | |
| *** gperal has quit IRC | 16:55 | |
| *** prashantD has joined #tacker | 16:59 | |
| *** zeih has joined #tacker | 17:04 | |
| *** zeih has quit IRC | 17:09 | |
| *** s3wong has joined #tacker | 17:10 | |
| *** vishwanathj has quit IRC | 17:23 | |
| *** vishwanathj has joined #tacker | 17:23 | |
| *** lhcheng_ has joined #tacker | 17:26 | |
| *** lhcheng has quit IRC | 17:29 | |
| *** karimb has quit IRC | 17:34 | |
| *** egonzalez has quit IRC | 17:52 | |
| sridhar_ram | s3wong: can you help to merge the bot requirement - https://review.openstack.org/#/c/255565/? | 17:59 |
| s3wong | sridhar_ram: sure | 18:04 |
| *** s3wong has quit IRC | 18:05 | |
| openstackgerrit | Merged openstack/tacker: Remove Service instance, context resources https://review.openstack.org/250156 | 18:05 |
| openstackgerrit | Merged openstack/tacker: Updated from global requirements https://review.openstack.org/255565 | 18:11 |
| *** dandruta has quit IRC | 18:14 | |
| *** karimb has joined #tacker | 18:27 | |
| *** sridhar_ram has quit IRC | 18:36 | |
| *** s3wong has joined #tacker | 18:36 | |
| *** karimb has quit IRC | 18:38 | |
| *** sripriya_ has quit IRC | 18:39 | |
| *** lhcheng_ is now known as lhcheng | 18:48 | |
| *** sripriya has joined #tacker | 18:57 | |
| *** sridhar_ram has joined #tacker | 19:17 | |
| *** sridhar_ram has quit IRC | 20:02 | |
| *** prashantD has quit IRC | 20:18 | |
| *** sridhar_ram has joined #tacker | 20:22 | |
| *** prashantD has joined #tacker | 20:24 | |
| *** karimb has joined #tacker | 20:26 | |
| *** bobh has quit IRC | 20:31 | |
| *** trozet has joined #tacker | 21:21 | |
| *** vishwanathj has quit IRC | 21:34 | |
| *** vishwanathj has joined #tacker | 21:34 | |
| *** u_kozat has joined #tacker | 21:40 | |
| openstackgerrit | Sripriya Seetharam proposed openstack/tacker: Fix mgmt_url 'NoneType' error can't be decoded https://review.openstack.org/256104 | 21:43 |
| *** karimb has quit IRC | 21:48 | |
| *** lhcheng_ has joined #tacker | 21:59 | |
| *** lhcheng has quit IRC | 22:03 | |
| *** u_kozat_ has joined #tacker | 22:09 | |
| *** karimb has joined #tacker | 22:11 | |
| *** u_kozat has quit IRC | 22:12 | |
| *** karimb has quit IRC | 22:14 | |
| *** karimb has joined #tacker | 22:25 | |
| *** gongysh has joined #tacker | 22:28 | |
| gongysh | sridhar_ram, hi | 22:30 |
| sridhar_ram | gongysh: hi | 22:34 |
| gongysh | you said we will remove the device , device template terms in tacker at mitaka cycle? | 22:34 |
| sridhar_ram | gongysh: yes | 22:35 |
| gongysh | ok | 22:35 |
| sridhar_ram | gongysh: that was the decision in one of the team meeting few months back | 22:35 |
| sridhar_ram | gongysh: we need to add a "deprecated" notice in stable/liberty code and remove that code in Mitaka | 22:36 |
| gongysh | sridhar_ram, after https://review.openstack.org/#/c/252824/ is merged, I think we should make our functional test voting. | 22:36 |
| sridhar_ram | gongysh: well, agree but lets monitor for 1 or 2 weeks before switch on voting | 22:37 |
| gongysh | ok, | 22:37 |
| sridhar_ram | gongysh: btw, AWESOME job nailing it all the way in Nova! | 22:37 |
| sridhar_ram | gongysh: saw you already have one +2 | 22:38 |
| gongysh | Yes, alex_xu, the current core, and garyk the neutron core are all my friends. | 22:38 |
| sridhar_ram | gongysh: it goes to say, not many people create VMs with multiple nics | 22:39 |
| sridhar_ram | .. in openstack | 22:39 |
| gongysh | sridhar_ram, right. | 22:39 |
| *** karimb has quit IRC | 22:39 | |
| sripriya | agree. gongysh: great job fixing that. we need that fix so badly... | 22:39 |
| sridhar_ram | gongysh: good to know.. | 22:39 |
| gongysh | sridhar_ram, one more thing I noticed the our vnfd template is not tosca tempalte | 22:40 |
| sridhar_ram | I know Sripriya chased Nova folks in Tokyo.. to the extent they got scared of her ;-) | 22:40 |
| sridhar_ram | gongysh: what you mean ? | 22:40 |
| gongysh | sridhar_ram, http://docs.oasis-open.org/tosca/tosca-nfv/v1.0/csd02/tosca-nfv-v1.0-csd02.pdf | 22:41 |
| gongysh | sridhar_ram, I cannot feel our way to define the vnfd is the way defined in this document. | 22:42 |
| sridhar_ram | gongysh: oh yes, we implemented a simple parser within tacker.. but the plan was always to cut over to a proper tosca parser. that's what bobh is doing | 22:42 |
| sridhar_ram | gongysh: keep in mind, tosca-nfv standard is also evolving in parallel.. | 22:43 |
| *** trozet has quit IRC | 22:43 | |
| gongysh | sridhar_ram, currently, our vnfd template is in casual format, we can have our own contents. | 22:43 |
| sripriya | sridhar_ram: lol, they were not scared of me, they just ignored me :-) with a "This is not a priority for us" look. glad we are fixing this as a "priority" for Tacker | 22:43 |
| sridhar_ram | gongysh: .. the plan is to take the attrs we define in tacker - say for things like numa, cpu-pinning and incorporate back into tosca-nfv | 22:44 |
| sridhar_ram | gongysh: exactly .. it is tosca-lite :) | 22:44 |
| sridhar_ram | gongysh: but we will be move to formal CSD soon w/ bobh's effort.. | 22:44 |
| sridhar_ram | gongysh: even then .. we will co-exist with some tacker specific attrs | 22:45 |
| sridhar_ram | gongysh: that is the reality | 22:45 |
| gongysh | sridhar_ram, cool, previously, I do things following standards, now we can infect back standands. | 22:45 |
| sridhar_ram | sripriya: yep | 22:45 |
| sridhar_ram | gongysh: exactly, the tables had been turned... standards are useless unless there are solid opensource implementations of them | 22:46 |
| gongysh | back to placement. | 22:46 |
| sridhar_ram | gongysh: sure | 22:46 |
| sridhar_ram | gongysh: I've a question to u, btw.. where do you include the strings to announce API deprecation.. ? | 22:48 |
| gongysh | one way is to specify the numa, cpu, memory stuff as vdu's properties. and we will translate into heat NOVA::FLAVOR resources to create flavor dynanically for each vdu. | 22:48 |
| sridhar_ram | gongysh: ack, that's one straight fwd way .. | 22:49 |
| gongysh | sridhar_ram, for API depre, current popular way is to use microversion. | 22:49 |
| gongysh | or to bump the API version number | 22:50 |
| gongysh | and then anounce the old one is dying | 22:50 |
| gongysh | placement: the dynamical flavor way for each VDU will create many flavors in nova. | 22:51 |
| sridhar_ram | will hold by deprecation questions.. lets continue w/ placement.. | 22:52 |
| gongysh | placement: so I want to define an element FLAVOR in vnfd template, and reference it in vdu element. | 22:52 |
| gongysh | In nova, the placement related feature is aggregate and flavor | 22:54 |
| vishwanathj | gongysh, sridhar_ram, hi | 22:54 |
| gongysh | vishwanathj, hi | 22:54 |
| sridhar_ram | gongysh: well, TOSCA uses host capabilities to describe all the VM properties.. | 22:55 |
| vishwanathj | the discussion sounds interesting eventhough I have to leave in 5 minutes to pick up my son | 22:55 |
| sridhar_ram | gongysh: we should try not use explicit FLAVOR element in vnfd template.. | 22:55 |
| sridhar_ram | let me give an example in a paste | 22:56 |
| gongysh | sridhar_ram, that is my concern. | 22:56 |
| gongysh | sridhar_ram, I don't want to explicit FLAVOR either, that is too nova way | 22:56 |
| sridhar_ram | gongysh: vishwanathj: see this http://paste.openstack.org/show/481565/ | 22:56 |
| sridhar_ram | in TOSCA these host properties are expressed has constraints / capabilities | 22:57 |
| vishwanathj | sridhar_ram, is that how TOSCA NFV defines them? | 22:57 |
| sridhar_ram | for things like numa, cpu-pinning .. we need to define new properties under host properties | 22:58 |
| sridhar_ram | vishwanathj: yes | 22:58 |
| sridhar_ram | vishwanathj: gongysh: keep in much some of this is derived from TOSCA simple profile.. | 22:58 |
| gongysh | sridhar_ram, so it seems we have to filter the flavor in nova, and then create flavor if not found when they are implemented. | 22:59 |
| sridhar_ram | the way it stands TOSCA-NFV profile is derived from TOSCA Simple Profile.. so some of the namespace (like artifacts for VM image are not in NFV but in simple profile) | 22:59 |
| vishwanathj | ok, I will make sure to revisit the TOSCA simple profile again | 22:59 |
| sridhar_ram | gongysh: yes.. | 22:59 |
| sridhar_ram | there is a debate in tosca whether to make nfv profile "standalone" (w/o being derived from simple profile). but that is lot of work... | 23:00 |
| gongysh | sridhar_ram, cool, I will have to revisit the TOSCA NFV file too. | 23:01 |
| sridhar_ram | gongysh: See discussion in tbh's https://bugs.launchpad.net/tacker/+bug/1516193 and his auto-resource spec https://review.openstack.org/250291 | 23:01 |
| openstack | Launchpad bug 1516193 in tacker "Automatic Flavor Creation based on VNFD Template" [Medium,Invalid] - Assigned to bharaththiruveedula (bharath-ves) | 23:01 |
| vishwanathj | gongysh, it might be a good idea for me and you to revisit it before our tacker meeting next week so that next week meeting can be fruitful | 23:01 |
| gongysh | sridhar_ram, after that I will try to provide a design spec in tacker spec. | 23:02 |
| sridhar_ram | gongysh: current thinking is we can create as many flavors as possible for each VDU .. | 23:02 |
| gongysh | sridhar_ram, yes, that is a first step, we can always improve. | 23:02 |
| sridhar_ram | gongysh: vishwanathj: do you guys see any issue in creating many flavors ? | 23:02 |
| sridhar_ram | gongysh: yep | 23:02 |
| gongysh | sridhar_ram, that will make administrator crazy. | 23:03 |
| sridhar_ram | roughly - number of VDUs in our VNFD catalog .. could go to hundreds.. | 23:03 |
| sridhar_ram | gongysh: well, administrators should see them or care of about them.. | 23:04 |
| sridhar_ram | gongysh: it is owned by tacker (will be name appropriately)... | 23:04 |
| sridhar_ram | gongysh: also if it gets created in tacker tenant / user .. it will stay hidden there | 23:04 |
| vishwanathj | Need to investigate if flavors created by tacker tenant will be hidden to other tenants | 23:04 |
| sridhar_ram | hmm.. I'm assuming a glance like --visiblity flag | 23:05 |
| sridhar_ram | vishwanathj: agree | 23:05 |
| gongysh | sridhar_ram, flavor can be created not to be shared. | 23:05 |
| sridhar_ram | gongysh: that's good .. either case we can't avoid creating many, many flavors .. it depends on VDU types coming into our catalog. | 23:06 |
| sridhar_ram | gongysh | 23:06 |
| vishwanathj | I think that there should be a way to come up with pre-created flavors for NFVs | 23:07 |
| sridhar_ram | vishwanathj: again, the arg is what are you optimizing for ? | 23:07 |
| sridhar_ram | vishwanathj: flavor create is a simple db operation ? | 23:07 |
| sridhar_ram | anyway.. this will be taken care in tbh | 23:08 |
| vishwanathj | we have to consider that just because a flavor is created, it does not mean it is usable or an instance can be instantiated.... | 23:08 |
| sripriya | sridhar_ram: gongysh: vishwanathj: by default flavor is visible to all tenants | 23:08 |
| vishwanathj | apologies, I have to leave now....feeling bad that I have to leave when the discussion is getting interesting | 23:09 |
| vishwanathj | sripriya, thanks for that data...also I have noticed that you can delete a flavor after an instance has been created with it | 23:10 |
| sridhar_ram | vishwanathj: well, yes.. you need to have compute hosts available / match for nova to schedule.. | 23:10 |
| gongysh | flavor-access-add Add flavor access for the given tenant. | 23:10 |
| gongysh | flavor-access-list Print access information about the given | 23:10 |
| gongysh | flavor. | 23:10 |
| gongysh | flavor-access-remove Remove flavor access for the given tenant. | 23:10 |
| gongysh | flavor-create Create a new flavor. | 23:10 |
| gongysh | flavor-delete Delete a specific flavor | 23:10 |
| gongysh | flavor-key Set or unset extra_spec for a flavor. | 23:10 |
| gongysh | flavor-list Print a list of available 'flavors' (sizes of | 23:10 |
| sripriya | vishwanathj: i see | 23:10 |
| gongysh | servers). | 23:10 |
| gongysh | flavor-show Show details about the given flavor. | 23:10 |
| vishwanathj | bye folks | 23:10 |
| *** vishwanathj has left #tacker | 23:10 | |
| gongysh | [fedora@tacker devstack]$ nova help flavor-create | 23:11 |
| gongysh | usage: nova flavor-create [--ephemeral <ephemeral>] [--swap <swap>] | 23:11 |
| gongysh | [--rxtx-factor <factor>] [--is-public <is-public>] | 23:11 |
| gongysh | <name> <id> <ram> <disk> <vcpus> | 23:11 |
| gongysh | Create a new flavor. | 23:11 |
| gongysh | Positional arguments: | 23:11 |
| gongysh | <name> Unique name of the new flavor. | 23:11 |
| gongysh | <id> Unique ID of the new flavor. Specifying 'auto' will | 23:11 |
| gongysh | generated a UUID for the ID. | 23:11 |
| gongysh | <ram> Memory size in MB. | 23:11 |
| gongysh | <disk> Disk size in GB. | 23:11 |
| gongysh | <vcpus> Number of vcpus | 23:11 |
| gongysh | Optional arguments: | 23:11 |
| gongysh | --ephemeral <ephemeral> Ephemeral space size in GB (default 0). | 23:11 |
| gongysh | --swap <swap> Swap space size in MB (default 0). | 23:11 |
| gongysh | --rxtx-factor <factor> RX/TX factor (default 1). | 23:11 |
| gongysh | --is-public <is-public> Make flavor accessible to the public (default | 23:11 |
| gongysh | true). | 23:11 |
| gongysh | sorry for the code way. | 23:11 |
| gongysh | Flavor can be created privately, and the access list can be controlled to. | 23:12 |
| gongysh | so the other tenants can be isolated to our tacker flavors. | 23:12 |
| sripriya | i guess --is -public will be True by default, which needs to set to False so that it is not shared | 23:13 |
| sridhar_ram | gongysh: another note, that tbh found, you can even delete a flavor even if the VM instantiated using that flavor is still running | 23:13 |
| gongysh | sridhar_ram, yes, the VM's flavor can be deleted even if VM is runing. | 23:14 |
| *** u_kozat_ has quit IRC | 23:42 | |
| sridhar_ram | radez: ping | 23:42 |
| sridhar_ram | gongysh: one help - can you please cherrypick the db-migration and service-instance/service-context cleanup fixed to stable/liberty branch ? | 23:44 |
| gongysh | yes | 23:44 |
| sridhar_ram | gongysh: our friends in tacker-sfc - trozet, radez - needs these fixes in stable/liberty | 23:45 |
| sridhar_ram | gongysh: thanks! | 23:45 |
| gongysh | cool, are they using the tacker project for production? | 23:45 |
| gongysh | sridhar_ram, what do u mean tacker-sfc? | 23:49 |
| gongysh | is it another project? | 23:49 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!