| *** sripriya has quit IRC | 00:01 | |
| openstackgerrit | vishwanath jayaraman proposed openstack/tacker: Move functional tests from commonservices directory https://review.openstack.org/367697 | 00:06 |
|---|---|---|
| *** sripriya has joined #tacker | 00:07 | |
| *** diga has joined #tacker | 00:07 | |
| *** vishwanathj is now known as vishwanathj_zzz | 00:08 | |
| *** vishwanathj_zzz is now known as vishwanathj_ | 00:08 | |
| *** KanagarajM_ has joined #tacker | 00:10 | |
| *** KanagarajM has quit IRC | 00:10 | |
| diga | trozet: | 00:12 |
| diga | trozet: Hi | 00:13 |
| *** uck has quit IRC | 00:15 | |
| bryan_att | sridhar_ram: any way to inject an SSH key into a VM via the template? I tried using the "key_name" property but the mitaka/stable version appears not to support that. | 00:18 |
| sridhar_ram | bryan_att: unfortunately that went in Newton as well :( | 00:18 |
| bryan_att | sridhar_ram: how were you able to do anything with VMs if you couldn't SSH to them? (at least with normal cloud images which have password SSH disabled) | 00:19 |
| bryan_att | I guess you were creating custom images for testing? | 00:20 |
| sridhar_ram | sripriya: qq, would a newton tacker + mitaka keystone work from vim-register point of view ? | 00:20 |
| sridhar_ram | bryan_att: many vnfs we tried support restapi, CLI and netconf | 00:21 |
| bryan_att | sridhar_ram: OK, let me see what I can find in the meantime, or I can customize an ubuntu server so I can log into it. I need SSH so that I can push setup scripts. | 00:22 |
| sripriya | sridhar_ram: d you mean registerting an openstack mitaka vim? | 00:23 |
| sripriya | sridhar_ram: or to run tacker tacker service itself? | 00:24 |
| *** uck has joined #tacker | 00:24 | |
| *** uck has quit IRC | 00:29 | |
| sridhar_ram | bryan_att: Ok... we also want to cover blank vm + vnfc s/w option (like ur vIMS).. hence we started adding ssh key, secgroup, etc in the newer release | 00:41 |
| bryan_att | sridhar_ram: OK, I'll get there soon enough. Next OPNFV release will be Newton based. In the meantime I'll figure it out. | 00:41 |
| sridhar_ram | sripriya: is later feasible? | 00:42 |
| sripriya | sridhar_ram: i dont think so, with common packages (oslo, middleware, etc.) used across multiple services, version incompatibility will cause problems for tacker newton to be installed with mitaka services | 00:45 |
| *** gongysh has quit IRC | 00:45 | |
| *** sripriya has quit IRC | 00:46 | |
| openstackgerrit | dharmendra kushwaha proposed openstack/python-tackerclient: Add support for multi delete https://review.openstack.org/349806 | 00:46 |
| sridhar_ram | sripriya: thought so, was trying to be greedy :) | 00:47 |
| sridhar_ram | sripriya: however newton tacker connecting to mitaka OpenStack would be supported...we shd point check this | 00:47 |
| *** gongysh has joined #tacker | 01:11 | |
| *** KanagarajM_ has quit IRC | 01:13 | |
| *** KanagarajM has joined #tacker | 01:13 | |
| *** gongysh has quit IRC | 01:41 | |
| *** dkushwaha__ has joined #tacker | 01:41 | |
| *** s3wong has quit IRC | 01:48 | |
| *** gongysh has joined #tacker | 02:01 | |
| *** diga has quit IRC | 02:32 | |
| *** dkushwaha__ has quit IRC | 02:38 | |
| *** amotoki has quit IRC | 02:49 | |
| *** vishnoianil has quit IRC | 02:52 | |
| *** amotoki has joined #tacker | 03:49 | |
| *** achatterjee has joined #tacker | 03:55 | |
| *** amotoki has quit IRC | 04:16 | |
| *** Vijayendra has joined #tacker | 04:21 | |
| *** tbh has joined #tacker | 05:05 | |
| *** neel has joined #tacker | 05:13 | |
| openstackgerrit | Merged openstack/tacker-horizon: Adds audit support for VNF, VIM and VNFD resources https://review.openstack.org/349755 | 05:17 |
| *** janki has joined #tacker | 05:21 | |
| *** lulei has quit IRC | 05:22 | |
| *** sripriya has joined #tacker | 05:23 | |
| openstackgerrit | dharmendra kushwaha proposed openstack/python-tackerclient: Add support for multi delete https://review.openstack.org/349806 | 05:27 |
| *** vishwanathj_ is now known as vishwanathj_zzz | 05:28 | |
| dkushwaha | gongysh, ping | 05:36 |
| dkushwaha | gongysh, please check my response on https://review.openstack.org/#/c/346557/ | 05:36 |
| *** kvmahesh has joined #tacker | 05:38 | |
| *** Ravikiran_K has joined #tacker | 05:41 | |
| *** amotoki has joined #tacker | 05:43 | |
| *** kvmahesh has quit IRC | 05:49 | |
| *** santoshk has joined #tacker | 05:56 | |
| santoshk | KanagarajM, Hi | 05:56 |
| *** vishnoianil has joined #tacker | 06:06 | |
| *** amotoki has quit IRC | 06:19 | |
| *** veena_ has joined #tacker | 06:20 | |
| openstackgerrit | Merged openstack/python-tackerclient: Add support for multi delete https://review.openstack.org/349806 | 06:28 |
| *** amotoki has joined #tacker | 06:28 | |
| *** sripriya has quit IRC | 06:46 | |
| *** amotoki has quit IRC | 06:48 | |
| janki | tung_doan, ping | 06:48 |
| *** KanagarajM has quit IRC | 06:49 | |
| *** KanagarajM has joined #tacker | 06:49 | |
| *** amotoki has joined #tacker | 06:52 | |
| *** santoshk has quit IRC | 06:56 | |
| *** amotoki has quit IRC | 07:02 | |
| *** amotoki has joined #tacker | 07:14 | |
| *** veena_ has quit IRC | 07:22 | |
| *** tung_doan has quit IRC | 07:32 | |
| *** vishwana_ has joined #tacker | 07:42 | |
| *** vishwana_ has quit IRC | 07:42 | |
| *** vishwana_ has joined #tacker | 07:43 | |
| *** neel has quit IRC | 07:43 | |
| *** vishwanathj_zzz has quit IRC | 07:44 | |
| *** KanagarajM_ has joined #tacker | 07:53 | |
| *** KanagarajM has quit IRC | 07:53 | |
| *** mbound has joined #tacker | 07:54 | |
| *** mbound has quit IRC | 07:59 | |
| openstackgerrit | JiWei proposed openstack/tacker: Fix order of arguments in assertIs https://review.openstack.org/367812 | 08:14 |
| *** tung_doan has joined #tacker | 08:15 | |
| tung_doan | janki: pong | 08:16 |
| janki | tung_doan, hey, I tried to solve that password issue | 08:20 |
| tung_doan | janki: oh.. nice.. i already had a way to fix.. but it's great to hear from you | 08:21 |
| janki | tung_doan, ohh cool :) | 08:21 |
| tung_doan | janki: really want to know your solution.. appreciate it | 08:24 |
| janki | tung_doan, I did it os module. | 08:24 |
| janki | tung_doan, how about you? | 08:24 |
| *** vishwana_ has quit IRC | 08:25 | |
| *** vishwanathj_zzz has joined #tacker | 08:25 | |
| tung_doan | janki: i still used username-password method.. but we dont need to hard-code them | 08:26 |
| janki | tung_doan, I did like this http://pastebin.com/LqYuXb73. Yours? | 08:27 |
| tung_doan | I just save username and pass in tacker.conf. but didnot try it.. did it work ok? | 08:30 |
| tung_doan | janki: with your changes | 08:31 |
| *** lulei has joined #tacker | 08:39 | |
| janki | tung_doan, yes, my changes work okay | 08:40 |
| janki | tung_doan, share your changes. I will try | 08:41 |
| *** neel has joined #tacker | 08:46 | |
| *** KanagarajM has joined #tacker | 08:47 | |
| *** KanagarajM_ has quit IRC | 08:49 | |
| *** KanagarajM_ has joined #tacker | 08:51 | |
| *** KanagarajM has quit IRC | 08:54 | |
| *** amotoki has quit IRC | 09:04 | |
| *** mfedosin has joined #tacker | 09:06 | |
| *** mbound has joined #tacker | 09:14 | |
| *** tbh has quit IRC | 09:22 | |
| *** KanagarajM__ has joined #tacker | 09:29 | |
| *** KanagarajM_ has quit IRC | 09:33 | |
| *** neel has quit IRC | 09:41 | |
| *** Vijayendra has quit IRC | 09:48 | |
| *** Vijayendra has joined #tacker | 09:48 | |
| openstackgerrit | Tung Doan proposed openstack/tacker: Alarm monitor: All in one https://review.openstack.org/365435 | 10:27 |
| *** neel has joined #tacker | 10:30 | |
| *** tung_doan is now known as tung_doan_afk | 10:32 | |
| *** mfedosin has quit IRC | 10:34 | |
| *** mfedosin has joined #tacker | 10:36 | |
| *** neel has quit IRC | 10:56 | |
| *** tbh has joined #tacker | 10:59 | |
| *** lulei has quit IRC | 11:00 | |
| *** mfedosin has quit IRC | 11:05 | |
| *** tung_doan_afk is now known as tung_doan | 11:21 | |
| *** KanagarajM__ has quit IRC | 11:23 | |
| openstackgerrit | dharmendra kushwaha proposed openstack/tacker: Cleanup unused file hooks.py https://review.openstack.org/367953 | 11:31 |
| *** dkushwaha has quit IRC | 11:33 | |
| *** Vijayendra has quit IRC | 11:48 | |
| *** gongysh has quit IRC | 12:00 | |
| *** mfedosin has joined #tacker | 12:12 | |
| *** bobh has joined #tacker | 12:19 | |
| *** bobh has quit IRC | 12:30 | |
| *** mfedosin has quit IRC | 12:36 | |
| *** mfedosin has joined #tacker | 12:44 | |
| *** tung_doan is now known as tung_doan_afk | 12:58 | |
| *** gongysh has joined #tacker | 13:11 | |
| trozet | sridhar_ram: ping? | 13:48 |
| trozet | gongysh: ping? | 13:51 |
| openstackgerrit | Janki Chhatbar proposed openstack/tacker: Alarm monitor: All in one https://review.openstack.org/365435 | 13:56 |
| openstackgerrit | Janki Chhatbar proposed openstack/tacker: Alarm monitor: All in one https://review.openstack.org/365435 | 13:57 |
| *** bobh has joined #tacker | 14:01 | |
| *** mfedosin has quit IRC | 14:02 | |
| *** mfedosin has joined #tacker | 14:04 | |
| openstackgerrit | Janki Chhatbar proposed openstack/tacker: Alarm monitor: All in one https://review.openstack.org/365435 | 14:06 |
| *** bobh has quit IRC | 14:06 | |
| *** janki has quit IRC | 14:07 | |
| *** bobh has joined #tacker | 14:14 | |
| *** tamilhce has joined #tacker | 14:58 | |
| *** tamilhce has quit IRC | 15:02 | |
| *** tung_doan_afk is now known as tung_doan | 15:09 | |
| *** tbh_ has joined #tacker | 15:09 | |
| tung_doan | gonggysh: ping | 15:09 |
| *** tbh_ has quit IRC | 15:09 | |
| *** gongysh has quit IRC | 15:13 | |
| openstackgerrit | bharaththiruveedula proposed openstack/tacker-specs: Tacker VNFC support https://review.openstack.org/339798 | 15:14 |
| *** mbound has quit IRC | 15:22 | |
| *** vishwanathj_zzz is now known as vishwanath | 15:27 | |
| *** vishwanath is now known as vishwanathj | 15:27 | |
| *** amotoki has joined #tacker | 15:27 | |
| *** gongysh has joined #tacker | 15:30 | |
| *** uck has joined #tacker | 15:44 | |
| *** yifei has quit IRC | 15:48 | |
| *** openstackgerrit has quit IRC | 15:49 | |
| *** openstackgerrit has joined #tacker | 15:50 | |
| -openstackstatus- NOTICE: New setuptools release appears to have a circular import which is breaking many jobs - check for ImportError: cannot import name monkey. | 15:55 | |
| *** ChanServ changes topic to "New setuptools release appears to have a circular import which is breaking many jobs - check for ImportError: cannot import name monkey." | 15:55 | |
| *** amotoki_ has joined #tacker | 16:02 | |
| *** gongysh has quit IRC | 16:03 | |
| *** amotoki_ has quit IRC | 16:03 | |
| *** amotoki has quit IRC | 16:05 | |
| *** mbound has joined #tacker | 16:22 | |
| *** mbound has quit IRC | 16:28 | |
| *** janki has joined #tacker | 16:42 | |
| *** yifei has joined #tacker | 16:46 | |
| *** ChanServ changes topic to "Discussion on OpenStack NFV Orchestration | Docs @ http://docs.openstack.org/developer/tacker/ | Wiki @ https://wiki.openstack.org/wiki/Tacker | Logged @ http://eavesdrop.openstack.org/irclogs/%23tacker/" | 16:59 | |
| -openstackstatus- NOTICE: setuptools 27.1.2 addresses the circular import | 16:59 | |
| *** uck has quit IRC | 17:09 | |
| *** yifei has quit IRC | 17:10 | |
| *** vishnoianil has quit IRC | 17:22 | |
| *** janki has quit IRC | 17:30 | |
| *** sripriya has joined #tacker | 17:30 | |
| sridhar_ram | trozet: pong | 17:39 |
| *** tbh has quit IRC | 17:47 | |
| openstackgerrit | Merged openstack/python-tackerclient: Removing test cases for already removed methods https://review.openstack.org/361027 | 17:47 |
| *** uck has joined #tacker | 17:49 | |
| *** mfedosin has quit IRC | 17:51 | |
| *** prashantD has joined #tacker | 17:55 | |
| trozet | sridhar_ram: i'm trying to test everything together on my setup | 17:59 |
| trozet | sridhar_ram: but the register vim part is new to me | 17:59 |
| trozet | might have questions | 17:59 |
| *** s3wong has joined #tacker | 18:01 | |
| trozet | sripriya: ping? | 18:03 |
| sridhar_ram | trozet: sure Tim. I'm point checking the feature as well.. | 18:10 |
| trozet | sridhar_ram: ok i was just able to create the VIM, i guess it's working...it says "REACHABLE" | 18:11 |
| sridhar_ram | trozet: perfect, that's good | 18:11 |
| trozet | sridhar_ram: i wasnt sure what domain name was in the vim.conf...i just put the region for keystone | 18:11 |
| sridhar_ram | trozet: btw, left few minor comments in the patchset | 18:12 |
| trozet | sridhar_ram: yeah saw that, thanks | 18:12 |
| *** vishnoianil has joined #tacker | 18:15 | |
| *** sripriya has quit IRC | 18:16 | |
| trozet | sridhar_ram: ok able to try to create VNFFG now, hit a bug already :) | 18:17 |
| sridhar_ram | trozet: finding bugs at this stage is absolute goodness :) | 18:19 |
| trozet | sridhar_ram: haha yeah looks like i forgot to do a try/exception | 18:20 |
| sridhar_ram | trozet: I see | 18:21 |
| trozet | sridhar_ram: actually there is some kind of bug in my filter | 18:29 |
| trozet | sridhar_ram: vnfm_plugin.get_vnfds(context, {'name': vnfd}, fields=['id']) | 18:30 |
| trozet | sridhar_ram: for filters (2nd arg) is that the correct way to write it? | 18:30 |
| trozet | sridhar_ram: oh crap, the vnfd needs to be in a list | 18:45 |
| *** uck_ has joined #tacker | 18:49 | |
| *** uck has quit IRC | 18:49 | |
| tung_doan | sridhar_ram: ping | 19:26 |
| *** sripriya has joined #tacker | 19:28 | |
| sripriya | trozet: pong | 19:28 |
| trozet | hey sripriya | 19:28 |
| trozet | sripriya: was asking earlier about registering a VIM, but figured it out with an example from unit test | 19:29 |
| sripriya | trozet: cool, continuing to review your patch | 19:29 |
| trozet | sripriya: good | 19:31 |
| trozet | sripriya: i'm testing the whole thing now in a full openstack setup | 19:31 |
| trozet | sripriya: already hit a few bugs | 19:31 |
| sripriya | trozet: nice, on which distro? | 19:31 |
| trozet | sripriya: stable/mitaka | 19:31 |
| sripriya | trozet: mitaka? | 19:32 |
| trozet | sripriya: i remember sridhar_ram emailing about refactoring vnfm and some changes there. Do you know off hand if the VNF DB was changed from template_id to vnfd_id? | 19:32 |
| sripriya | trzoet: yes it was refactored | 19:32 |
| sridhar_ram | tung_doan: pong | 19:33 |
| trozet | sripriya: ok, my code is trying to filter for template_id | 19:33 |
| trozet | sripriya: so thats probably why things are breaking | 19:33 |
| sripriya | trozet: ^ | 19:33 |
| sripriya | trozet: i saw that yest. and wondered how it was working | 19:33 |
| trozet | sripriya: well in unit tests i mock the VNFM | 19:33 |
| trozet | sripriya: so tahts why those pass | 19:34 |
| tung_doan | sridhar_ram: Hi sridhar, actually next week is mid-autumn festival in Korea. Therefore, my work should be finished until 14th Sept. | 19:34 |
| sripriya | trozet: hmm that needs to be fixed too then | 19:34 |
| tung_doan | Currently, gonggysh and Janki are helping me to review my patch... Hope we can wrap this patch soon... | 19:34 |
| trozet | sripriya: yeah need to change that to vnfd_Id | 19:34 |
| sripriya | trozet: cool | 19:34 |
| sripriya | trozet: on my earlier question, why mitaka? | 19:34 |
| trozet | sripriya: thats what my OPNFV installer uses | 19:35 |
| trozet | sripriya: not using devstack | 19:35 |
| sripriya | trozet: the networking-sfc doesnt work on newton? | 19:35 |
| trozet | sripriya: no i just know this setup, and its dependable so I can be sure the rest of the stuff works | 19:35 |
| sripriya | trozet: sorry need some context, so you integrated your code to work with tacker mitaka then? | 19:36 |
| trozet | sripriya: no , i just installed tacker off my patch in master | 19:37 |
| trozet | sripriya: the rest of openstack is stable/mitaka | 19:37 |
| sripriya | trozet: are all installed on the same system? | 19:38 |
| trozet | sripriya: well its 3 nodes, 1 controller, and 2 computes | 19:38 |
| trozet | but yeah | 19:38 |
| sripriya | trozet: and you register the openstack instance as a vim and deploy vnffg | 19:39 |
| trozet | sripriya: yeah, after i created 2 VNFs | 19:39 |
| sripriya | trozet: i’m surprized you did not run into issues with common packages versions between tacker master and mitaka openstack services | 19:40 |
| trozet | sripriya: yeah it seems ok | 19:40 |
| sripriya | trozet: ah ok good to know, | 19:41 |
| sripriya | sridhar_ram: your ask yest. of tacker newton with openstack mitaka keystone works as per trozet. | 19:41 |
| sripriya | trozet: one last question which linux distro are you using? | 19:42 |
| trozet | sripriya: centos ftw! | 19:47 |
| trozet | sridhar_ram, sripriya: btw the vnf resource list stuff works, even with client-> server :) | 19:47 |
| sripriya | trozet: nice! | 19:48 |
| *** uck_ has quit IRC | 19:55 | |
| trozet | sripriya: VimInUseException: VIM e62b2468-6377-4a87-8d68-4cfcf39e1bdb is still in use by VNF | 20:01 |
| sripriya | trozet: are you trying to delete a VIM? | 20:02 |
| trozet | sripriya: that is returned when i do get_vim()? | 20:02 |
| sripriya | trozet: no | 20:02 |
| trozet | sripriya: https://paste.fedoraproject.org/425000/raw/ | 20:03 |
| trozet | def _get_vim(self, context, vim_id): | 20:03 |
| trozet | if not self.is_vim_still_in_use(context, vim_id): | 20:03 |
| trozet | return self.get_vim(context, vim_id) | 20:03 |
| trozet | sripriya: is that intentional? or a bug? | 20:06 |
| sripriya | trozet: that is used internally by update_vim for updating vim only when vnf is not deployed on it | 20:07 |
| trozet | sripriya: ah i should use get_vim not _get_vim | 20:07 |
| sripriya | trzoet: in your code you need to be using the get_vim call instead of _get_vim | 20:07 |
| trozet | ok easy fix | 20:07 |
| sripriya | trozet: may be that method name should be different insted of _get_vim | 20:08 |
| trozet | sripriya: yeah | 20:08 |
| sridhar_ram | tung_doan: act, let's wrap up bef Sept 14th | 20:09 |
| sridhar_ram | tung_doan: *ack | 20:09 |
| tung_doan | I will spend the rest of this week and several days next week to fix and enhance code... so if possible, please help me to review | 20:09 |
| trozet | sridhar_ram, sripriya: huzzah! https://paste.fedoraproject.org/425007/ | 20:09 |
| trozet | (note this is without the networking-sfc driver, im just generating UUID for driver call) | 20:10 |
| sridhar_ram | tung_doan: will help w/ reviews | 20:10 |
| sripriya | trozet: super awesome | 20:10 |
| sridhar_ram | trozet: woot! | 20:10 |
| sridhar_ram | trozet: the uuids in vnf_mapping maps to existing active VNFs ? | 20:11 |
| trozet | | a8d3f96f-d34a-404c-a552-f0ec13ee41ea | testVNF2 | | 20:12 |
| trozet | | e0d23783-27bd-43ee-9f16-ae4d4506ba23 | testVNF1 | | 20:12 |
| trozet | sridhar_ram^ yeah i this vnffg i created with abstract types, so the plugin picked which ones to use | 20:12 |
| sridhar_ram | trozet: cool.. i believe you now ;-) | 20:12 |
| trozet | sridhar_ram, sripriya: more goodness https://paste.fedoraproject.org/425008/ | 20:13 |
| sridhar_ram | trozet: if you create the same vnffg those VNFs won't be used again ? | 20:13 |
| sridhar_ram | *those two VNFs | 20:13 |
| trozet | sridhar_ram: haha stop poking holes in it! | 20:13 |
| trozet | not a use case :) | 20:13 |
| trozet | sridhar_ram: no but seriously, the same VNFs can be used again in a chain | 20:13 |
| sridhar_ram | trozet: lol... couldn't hold my enthusiam | 20:13 |
| sridhar_ram | trozet: yes, with specific mapping .. but how abt w/ abstract types ? | 20:14 |
| trozet | sridhar_ram: it will reuse them | 20:14 |
| trozet | sridhar_ram: i actually dont know though if networking-sfc handles that, or if the driver handles that correctly | 20:15 |
| trozet | sridhar_ram: but from plugin perspective, you should be able to use VNFs in more than one chain | 20:15 |
| sripriya | trozet: only if there are more than one VNF for vnfd and random func. picked the same VNF again | 20:15 |
| trozet | sripriya: yeah that is a bug for sure | 20:15 |
| trozet | sripriya: so if you said VNF1, VNF1 (as vnfds) in your forwarding path and it picked the same one twice, that would be a bug I think | 20:16 |
| trozet | sripriya: and i think that bug exists in my code | 20:16 |
| sridhar_ram | trozet: :) | 20:16 |
| trozet | sridhar_ram, sripriya: right now the code would just use the first VNF instance for both, but if a user specifies 2 VNFDs explicitly in the path i think we should do 2 | 20:16 |
| trozet | sridhar_ram, sripriya: agree? | 20:17 |
| sripriya | trzoet: if you needed to deploy a new VNFFG with the same VNFD and it picked up the same VNF again that should work | 20:17 |
| sripriya | trozet: ^ | 20:17 |
| trozet | sripriya: yeah that part should work now. The issue I am talking about is in a single VNFFG, your forwarding path is 2 VNFs, both of the same VNFD type, and you create with abstract types | 20:18 |
| trozet | sripriya: that means the plugin goes and decides which VNFs to use, and right nwo with the logic, it will pick the same VNF twice | 20:18 |
| sripriya | trozet: i see | 20:18 |
| trozet | sripriya: it should pick 2 different ones, right? | 20:18 |
| sripriya | trozet: suppose there are 2 VNFS, for the same VNFD it could pick both of them | 20:19 |
| trozet | sripriya: what if there was only 1 VNF, and 2 were specified of the same VNFD type in the forwarding path, we should throw exception? | 20:19 |
| sripriya | trozet: yeah it should pick 2 diff ones | 20:19 |
| trozet | sripriya: ok let me try to fix that | 20:19 |
| *** mbound has joined #tacker | 20:20 | |
| sripriya | trozet: but is that doable? | 20:20 |
| trozet | sripriya: doable code wise? | 20:20 |
| sripriya | trozet: where one VNF is used twice? | 20:20 |
| sripriya | trozet: on the driver side? | 20:21 |
| sripriya | trozet: not code | 20:21 |
| trozet | sripriya: well speaking not on networking-sfc driver, but on IETF SFC, yeah it is possible | 20:21 |
| sripriya | trozet: then shouldn’t we allow it? | 20:22 |
| trozet | sripriya: i don't think it is currently allowed in networking-sfc (they dont use IETF SFC just yet) | 20:22 |
| trozet | sripriya: but something I can follow up on | 20:22 |
| sripriya | trozet: okay | 20:22 |
| trozet | sripriya: for now I'll throw an exception if htere are not enough VNFs, and then I will look into that | 20:23 |
| sripriya | trozet: can we defer this decision to the driver or should this be handled in plugin itself | 20:23 |
| sripriya | trozet: okay | 20:23 |
| trozet | sripriya: that is another good point | 20:24 |
| trozet | sripriya: but i think if more than 1 is available, we should try to use a different one | 20:24 |
| trozet | sripriya: so let me at least add that | 20:24 |
| trozet | sripriya: and maybe not thrown any exception or do anythign right now for hte other case | 20:25 |
| sridhar_ram | trozet: just caught up.. at the outset, it seems using two VNFs of the same type in a FFG seems a bad idea | 20:25 |
| trozet | sridhar_ram: you think we should just ignore that case for now? | 20:25 |
| sridhar_ram | trozet: hmm... i think we shd error out | 20:25 |
| trozet | sridhar_ram: ok so if the same VNFD is listed in the path twice in the VNFFGD, we should throw exception on VNFFGD create, right? | 20:26 |
| sridhar_ram | trozet: yes, i can think of some convoluted cases where we might end up w/ such an ask.. but as a mainstream case we are better have erroring out | 20:27 |
| sridhar_ram | *better off | 20:27 |
| trozet | sridhar_ram: ok i will add that now then | 20:27 |
| trozet | sridhar_ram: so i guess where I am at now, let me fix that along with the comments in the review. The next step will be refactoring s3wong patch to be in openstack driver | 20:27 |
| sridhar_ram | trozet: let some ffg user explicitly hit this "restriction" and come to us with correct reasoning | 20:27 |
| trozet | sridhar_ram: but it looks like the plugin is working | 20:28 |
| sridhar_ram | trozet: super cool.. once the n-sfc refactoring into openstack driver happens.. we shd land all these patches and continue w/ test <-> fix cycle | 20:29 |
| sridhar_ram | sripriya: what do you think ? | 20:29 |
| * trozet exudes giant sigh of relief | 20:30 | |
| trozet | sridhar_ram: also i just tried creating VNFFG with vnf_mapping provided, works | 20:31 |
| *** sripriya has quit IRC | 20:40 | |
| *** tung_doan is now known as tung_doan_afk | 20:43 | |
| *** sripriya has joined #tacker | 20:49 | |
| sripriya | sridhar_ram: good to land these patches asap else there will be another few cycles spent on resolving merge conflicts | 20:52 |
| *** uck has joined #tacker | 20:56 | |
| *** openstackstatus has quit IRC | 20:57 | |
| *** openstackstatus has joined #tacker | 21:00 | |
| *** ChanServ sets mode: +v openstackstatus | 21:00 | |
| *** uck has quit IRC | 21:00 | |
| trozet | sridhar_ram: lol so i added the thing to detect the same forwarder twice...and my unit test fail because it caught that my unit test example file has VNF1 listed twice: | 21:05 |
| trozet | path: | 21:05 |
| trozet | - forwarder: VNF1 | 21:05 |
| trozet | capability: CP11 | 21:05 |
| trozet | - forwarder: VNF1 | 21:05 |
| trozet | capability: CP12 | 21:05 |
| trozet | sridhar_ram, sripriya: actually now I realize listing the same VNF in the path is fine, as long as the CP is different | 21:07 |
| trozet | sridhar_ram, sripriya: if the same VNF is listed twice, I assume the first port is the ingress port and the 2nd port is egress | 21:08 |
| trozet | sridhar_ram, sripriya: so i guess really what we should look for is it listed more than twice, and not in order | 21:08 |
| *** uck has joined #tacker | 21:15 | |
| *** sripriya has quit IRC | 21:18 | |
| *** mbound has quit IRC | 21:19 | |
| *** sripriya has joined #tacker | 21:20 | |
| sripriya | trozet: yeah makes sense | 21:25 |
| sripriya | trozet: also, is it possible to create port pair out of ingress port of one vnf with egress port of another vnf | 21:26 |
| trozet | sripriya: no, i don't think that is possible | 21:26 |
| sripriya | trozet: i mean two different forwarders | 21:27 |
| sripriya | trozet: then in that case ingress and egres sport will be the same | 21:27 |
| sripriya | egress port* | 21:27 |
| trozet | sripriya: if you have a single VNF, and only one port listed, it is assumed that is both ingress and egress | 21:27 |
| sripriya | trozet: got it | 21:27 |
| sripriya | trozet: and if we have multiple VNFs with one port listed, it will be the same behavior again | 21:28 |
| trozet | sripriya: yeah | 21:28 |
| trozet | sripriya: i added a check for duplicate ports as well, for hte same VNF | 21:29 |
| trozet | sripriya: so if you have VNF1 CP1, VNF1 CP1, it will throw error | 21:29 |
| *** prashantD has quit IRC | 21:29 | |
| *** prashantD has joined #tacker | 21:30 | |
| sripriya | trozet: ok | 21:30 |
| sripriya | trozet: we have a good no. of combinations here that can happen, it is good to document these behaviors | 21:31 |
| *** bobh has quit IRC | 21:40 | |
| openstackgerrit | Tim Rozet proposed openstack/tacker: Implements VNFFG into NFVO https://review.openstack.org/344522 | 22:07 |
| *** prashantD has quit IRC | 22:17 | |
| *** prashantD has joined #tacker | 22:17 | |
| *** uck has quit IRC | 22:23 | |
| *** sripriya has quit IRC | 22:25 | |
| *** sripriya has joined #tacker | 22:25 | |
| *** sripriya has quit IRC | 22:41 | |
| *** bobh has joined #tacker | 22:54 | |
| *** bobh has quit IRC | 22:57 | |
| *** uck has joined #tacker | 23:12 | |
| *** uck has quit IRC | 23:17 | |
| *** vishwanathj has quit IRC | 23:52 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!