| *** sripriya has quit IRC | 00:13 | |
| *** dkushwaha__ has quit IRC | 00:17 | |
| *** yifei has joined #tacker | 00:17 | |
| *** saju_m has joined #tacker | 00:28 | |
| *** foutatoro has quit IRC | 00:31 | |
| *** saju_m has quit IRC | 00:53 | |
| *** gongysh has joined #tacker | 00:58 | |
| trozet | sridhar_ram: ping? | 00:59 |
|---|---|---|
| *** s3wong has quit IRC | 01:06 | |
| gongysh | sridhar_ram, hi | 01:17 |
| sridhar_ram | trozet: gongysh: pong | 01:19 |
| trozet | sridhar_ram: did you see my earlier dilemma about the tosca stuff | 01:19 |
| gongysh | sridhar_ram, you said: Yes, without advsvc Tacker won't be able to create cross-tenant nics. This is something we need to preserve as we improve Tacker's multi-tenancy in the upcoming cycles. | 01:20 |
| gongysh | what do you mean cross-tenant nics? | 01:20 |
| sridhar_ram | trozet: I haven't fully caught up to it yet.. can you summarize ? | 01:21 |
| trozet | sridhar_ram: the tosca parser expects that the VNFs and CPs defined in the forwarding path are also listed in the VNFFGD | 01:22 |
| trozet | sripriya: so i think we can either A) require a user to copy the VNFD template into the VNFFGD template, B) automatically look up the VNFD before sending it to tosca parser and include in the parser arg | 01:22 |
| trozet | <trozet> sridhar_ram:^ | 01:22 |
| sridhar_ram | gongysh: ability to spin VNFs is tenants other than "nfv" tenant and networks | 01:22 |
| sridhar_ram | trozet: hmm..can this (grabbing dependent VNFD template) into FFGD silently in the backend ? | 01:24 |
| sridhar_ram | trozet: (B) I guess | 01:24 |
| sridhar_ram | gongysh: basically mimic the old servicevm concept | 01:25 |
| trozet | sridhar_ram: yeah the only problem there is its going to need to just be a best effort type of action, because at that point we don't know if the VNFFGD is even correct, so we are trying to read it before we have validated it | 01:25 |
| gongysh | sridhar_ram, I am not quite understand. currently, all the resources (vnf's VDU vms, flavors, image) in openstack is under VIM's project context. | 01:26 |
| sridhar_ram | trozet: i see. what is the earliest we can error out due to syntax issues ? | 01:26 |
| trozet | sridhar_ram: well if it has yaml syntax problems, it will fail pretty early | 01:27 |
| trozet | sridhar_ram: but if its just incorrect tosca syntax, we won't know until it hits the translator | 01:27 |
| sridhar_ram | trozet: which will happen in ffg-create ? | 01:28 |
| trozet | sridhar_ram: i can put a block before hte translator to say, hey try to read the forwarding path VNFs and CPs, then add it to what is going to be validated | 01:28 |
| trozet | sridhar_ram: no this will happen at VNFFGD create | 01:28 |
| gongysh | sridhar_ram, I mean, if my VIM1's project is 'gongysh', but my tacker's project is sridhar, the sridhar tacker user will create all the resources under openstack 'gongysh' project. | 01:28 |
| trozet | sridhar_ram: this will be not really "checking that the VNFDs exist" it will just be tricking the tosca translator | 01:28 |
| sridhar_ram | trozet: given where we are i think it is okay .. to push this a bit on the user to keep things sane. We can improve in a later iteration | 01:28 |
| *** sripriya has joined #tacker | 01:29 | |
| trozet | sridhar_ram: ok makes sense, right now i'm trying to just get a dummy one to pass the translator | 01:29 |
| sridhar_ram | trozet: okay.. | 01:29 |
| sridhar_ram | folks - family is calling, will be back in 60mins | 01:30 |
| openstackgerrit | Stephen Wong proposed openstack/tacker: VNFFG service chain Neutron networking-sfc driver https://review.openstack.org/347568 | 01:33 |
| *** s3wong has joined #tacker | 01:34 | |
| s3wong | trozet: uploaded new patch | 01:36 |
| trozet | s3wong: cool thanks | 01:36 |
| s3wong | trozet: still missing port-chain update, but hopefully can unblock you there | 01:37 |
| trozet | s3wong: ok. I'm working on the tosca part then will look at your patch | 01:38 |
| s3wong | trozet: cool --- let me know if it doesn't work. The networking-sfc part should be good... though I can't run all the functional tests I want as I haven't set up a multi-node setup | 01:39 |
| trozet | s3wong: ok will do | 01:39 |
| s3wong | trozet: Thanks! | 01:40 |
| trozet | s3wong: thanks for pushing it :) | 01:40 |
| s3wong | trozet: speaking of which --- why do you want to add symmetrical in flow-classifier related driver interfaces? | 01:40 |
| s3wong | trozet: reverse fc should be created only if a chain has symmetrical flag set | 01:41 |
| trozet | s3wong: right | 01:41 |
| s3wong | trozet: (we are still only doing one fc uuid) | 01:41 |
| *** openstack has joined #tacker | 01:46 | |
| s3wong | trozet: at least that's how I would implement it if I take on the task at the networking-sfc side | 01:46 |
| trozet | s3wong: i would suggest just override the method with whatever you feel is necessary, then lets go fix the abstract driver | 01:46 |
| s3wong | trozet: abstract driver wouldn't be set in stone anyway | 01:46 |
| s3wong | trozet: we still need to add the NSH related parameters at some point --- seems like the support will be coming soon at networking-sfc | 01:46 |
| trozet | s3wong: yah | 01:47 |
| *** uck has joined #tacker | 01:48 | |
| s3wong | trozet: perhaps moving forward, we would like to hide all those details in some sort of dictionary, so we don't need to constantly extend driver interface | 01:48 |
| s3wong | trozet: again, post-Newton | 01:48 |
| s3wong | trozet: anyway. Have fun. Time for me to go home | 01:49 |
| trozet | s3wong: sure. ttyl | 01:49 |
| s3wong | trozet: ttyl | 01:49 |
| *** s3wong has quit IRC | 01:49 | |
| openstackgerrit | vishwanath jayaraman proposed openstack/tacker: VNF monitoring event capture: Functional test https://review.openstack.org/366226 | 01:52 |
| gongysh | sripriya, hi, I want to remove auth_url unique limit to support any tacker users can register its own VIM for one openstack. | 02:07 |
| gongysh | sripriya, what do u think? | 02:07 |
| sripriya | gongysh: this goes back to the multi tenant support discussion we had earlier around midcycle | 02:12 |
| sripriya | gongysh: it is easy to remove the constraint on auth url | 02:13 |
| sripriya | gongysh: and it needs “shared” attribute to be removed as well | 02:13 |
| sripriya | gongysh: the only bit of concern is with the tenant mismatch between tacker and the remote vim itself | 02:13 |
| sripriya | gongysh: i’m referring to the tenant where a vnf is deployed with the tenant hosting the actual instance resource | 02:14 |
| gongysh | sripriya, I know that. the local tenant and remote tenant can be different or the same. | 02:15 |
| gongysh | sripriya, I don't think it is a concern. If tacker and VIM are on the same keystone, they can be the same, or they should be different. | 02:17 |
| sripriya | gonysh: also, today— to support auto flavor creation the user is expected to have ‘admin’ role on remote tenant, this will end up with users being able to access other instances as well, so this is not necessarily providing multi tenant support | 02:18 |
| sripriya | gongysh: even when they are on same keystone, they can be on different tenant | 02:18 |
| gongysh | sripriya, yes, I said they can be the same. | 02:19 |
| sripriya | gongysh: i think it is good to take this discussion on gerrit with a spec | 02:20 |
| gongysh | sripriya, ok I will draft a spec | 02:20 |
| sripriya | gongysh: this will help us collect inputs and take the right approach to provide multi tenant support | 02:21 |
| sripriya | gongysh: thanks | 02:21 |
| *** avishnoi has quit IRC | 02:23 | |
| *** tung_doan has joined #tacker | 02:23 | |
| trozet | sripriya: i was able to work around the problem, by defining my own tosca.nodes.nfv.FP.Tacker | 02:28 |
| trozet | sridhar_ram:^ | 02:28 |
| sripriya | trozet: did you see https://github.com/openstack/tosca-parser/blob/master/toscaparser/extensions/nfv/TOSCA_nfv_definition_1_0.yaml | 02:36 |
| trozet | sripriya: yeah | 02:36 |
| sripriya | trozet: so you override the original tosca.nodes.nfv.FP and provide your own node types for it? | 02:37 |
| sripriya | trozet: how are you handling the criteria node type? | 02:38 |
| trozet | sripriya: yeah it looks like if you specify "requirements" in the node type, tosca parser automatically tries to figure out relationships | 02:38 |
| trozet | sripriya: this is what I have right now | 02:38 |
| *** vishwana_ is now known as vishwanathj_zzz | 02:38 | |
| openstackgerrit | Tung Doan proposed openstack/tacker: Alarm monitor: All in one https://review.openstack.org/365435 | 02:39 |
| trozet | sripriya: https://paste.fedoraproject.org/423716/raw/ | 02:39 |
| trozet | sripriya: i'm expanding on it to list hte exact criteria types allowed | 02:40 |
| *** KanagarajM has joined #tacker | 02:41 | |
| sripriya | trozet; ok, so you dont need the ‘requirements’ node type the vnffgd template | 02:42 |
| sripriya | trozet: https://review.openstack.org/#/c/344522/24/tacker/tests/unit/vm/infra_drivers/heat/data/vnffgd_template.yaml@13 | 02:42 |
| trozet | sripriya: right, that is the workaround. Don't use requirements or tosca parser will search for relationships | 02:42 |
| trozet | sripriya: yeah i'm going to have to change the other parts of hte code, this impacts it a bit | 02:43 |
| trozet | sripriya: new template https://paste.fedoraproject.org/423719/raw/ | 02:43 |
| sripriya | trozet: i see, i wonder why they added that in the tosca translator | 02:44 |
| sripriya | trozet: for FP | 02:44 |
| trozet | sripriya: yeah the groups part looks fine, i can leave that just as its regular type | 02:44 |
| sripriya | trozet: ok the new template aligns with the custom node types | 02:44 |
| trozet | sripriya: yeah gonna try to finish this and post a new patch before i go to bed | 02:45 |
| sripriya | trozet: sure, but any idea why they added requirements for FP? | 02:45 |
| trozet | sripriya: well the way we are using it does not exactly match | 02:46 |
| sripriya | trozet: hmmm | 02:46 |
| trozet | sripriya: i think sridhar_ram said there was redundancy when we reviewed the spec, so we modified the forwarder/capability stuff a little | 02:46 |
| sripriya | trozet: okay… | 02:47 |
| trozet | sripriya: iirc the "capability" was supposed to be 'forwarder' or something to allow forwarding capability | 02:47 |
| trozet | sripriya: and the forwarder was the CP or something, i can't exactly remember | 02:48 |
| sripriya | trozet: this is vice versa to what we have in template :-) | 02:49 |
| trozet | sripriya: yeah I would need to go back and look at the TOSCA docs to remember | 02:49 |
| sripriya | trozet: i also see in tosca translator that capability appears to be child node of forwarder | 02:50 |
| trozet | sripriya: yeah i remember there was some syntax problem in the early tosca docs there | 02:51 |
| trozet | sripriya: like it wasn't even proper yaml syntax | 02:51 |
| sripriya | trozet: ok | 02:51 |
| sripriya | trozet: does group need custome defs as well? | 02:53 |
| trozet | sripriya: nah that one is ok | 02:53 |
| sripriya | trzoet: i dont see the translator code handling all properties defined in the template | 02:53 |
| openstackgerrit | Merged openstack/tacker: Better handle vim domain exception https://review.openstack.org/351431 | 02:53 |
| sripriya | trozet: properties is fine, i guess the ‘members’ part | 02:54 |
| trozet | sripriya: yeah the members part - it seems to work if Forwarding_Path1 just exists in the node_templates | 02:55 |
| openstackgerrit | Merged openstack/tacker: Update the sample vnfd template in getting started guide https://review.openstack.org/363501 | 02:56 |
| sripriya | trozet: ok | 02:56 |
| sripriya | trozet: otherwise it complains uknown node type | 02:56 |
| *** KanagarajM has quit IRC | 03:06 | |
| *** hparekh has joined #tacker | 03:50 | |
| *** sripriya has quit IRC | 03:56 | |
| openstackgerrit | venkatamahesh proposed openstack/tacker: [py35] Fix for jenkins-gate-py35 error https://review.openstack.org/366656 | 04:00 |
| openstackgerrit | venkatamahesh proposed openstack/tacker: [py35] Fix for jenkins-gate-py35 error https://review.openstack.org/366656 | 04:02 |
| *** KanagarajM has joined #tacker | 04:08 | |
| *** haiwei has quit IRC | 04:13 | |
| *** uck has quit IRC | 04:18 | |
| *** neel has joined #tacker | 04:19 | |
| *** amotoki has joined #tacker | 04:23 | |
| *** links has joined #tacker | 04:28 | |
| openstackgerrit | Tim Rozet proposed openstack/tacker: Implements VNFFG into NFVO https://review.openstack.org/344522 | 04:36 |
| trozet | sridhar_ram: around? | 04:37 |
| *** janki has joined #tacker | 04:38 | |
| *** diga has joined #tacker | 04:38 | |
| *** sripriya has joined #tacker | 04:46 | |
| *** n-harada has joined #tacker | 04:50 | |
| *** hparekh_ has joined #tacker | 04:51 | |
| *** gongysh has quit IRC | 04:52 | |
| *** hparekh has quit IRC | 04:54 | |
| *** mfedosin has joined #tacker | 04:55 | |
| *** sripriya has quit IRC | 05:03 | |
| *** avishnoi has joined #tacker | 05:07 | |
| *** sripriya has joined #tacker | 05:07 | |
| *** tbh has joined #tacker | 05:11 | |
| dkushwaha | sripriya, ping | 05:14 |
| sripriya | dkushwaha: pong | 05:14 |
| dkushwaha | sripriya, as per your comment on https://review.openstack.org/#/c/349806, i am updating it. So just before commiting the code, i need a confirmation | 05:15 |
| dkushwaha | sripriya, please look on http://paste.openstack.org/show/568935/ | 05:15 |
| *** mfedosin has quit IRC | 05:15 | |
| dkushwaha | sripriya, I have updated the code, and the new output would be like in ^^ | 05:16 |
| dkushwaha | sripriya, please confirm if it looks ok. | 05:18 |
| *** Vijayendra has joined #tacker | 05:19 | |
| *** neel has quit IRC | 05:19 | |
| *** neel has joined #tacker | 05:21 | |
| *** Vijayendra has quit IRC | 05:25 | |
| sripriya | dkushwaha: would it be the same kind of message for other resources too? | 05:31 |
| *** links has quit IRC | 05:31 | |
| *** Vijayendra has joined #tacker | 05:38 | |
| dkushwaha | sripriya, as mentioned for vnfd, it would be same for others | 05:43 |
| dkushwaha | sripriya, just different for vnf only | 05:43 |
| sripriya | dkushwaha: ah ok, i see that | 05:44 |
| sripriya | dkushwaha: can the message be along similar lines of https://review.openstack.org/#/c/314716/3/tackerclient/tacker/v1_0/__init__.py@536 | 05:45 |
| sripriya | dkushwaha: otherwise looks fine | 05:45 |
| *** links has joined #tacker | 05:46 | |
| dkushwaha | sripriya, yes we can, but to keep code simpler and generic for for all others too, I would perfer it as "All vnf(s) accepted to delete successfully" or "Successfully accepted to delete vnf(s):" (as per the scanerio) | 05:49 |
| openstackgerrit | venkatamahesh proposed openstack/tacker: Clean up the imports according to hacking standards https://review.openstack.org/359608 | 05:51 |
| dkushwaha | sripriya, any thought on that? | 05:51 |
| sripriya | dkushwaha: the message with accepted doesnt seem right (implies whether all or just few were accepted for deletion) versus, the suggestion on delete initated versus deleted imply it is not yet deleted completely | 05:53 |
| dkushwaha | sripriya, when all accepted it will print " All vnf(s) accepted to delete successfully" and in case when few are accepted among all it will print with id/name which are accepted . Like as: "Successfully accepted to delete vnf(s): v1, v2, v3" | 05:57 |
| dkushwaha | sripriya, isn't what we want? | 05:58 |
| sripriya | dkushwaha: no, a general message for vnfs | 05:59 |
| openstackgerrit | Tung Doan proposed openstack/tacker: Alarm monitor: All in one https://review.openstack.org/365435 | 06:03 |
| *** Ravikiran_K has joined #tacker | 06:04 | |
| dkushwaha | sripriya, Actually I didn't get your point. I will push the code on gerrit. kindly comment there. | 06:11 |
| openstackgerrit | dharmendra kushwaha proposed openstack/python-tackerclient: Add support for multi delete https://review.openstack.org/349806 | 06:13 |
| *** saju_m has joined #tacker | 06:15 | |
| openstackgerrit | dharmendra kushwaha proposed openstack/python-tackerclient: Add support for multi delete https://review.openstack.org/349806 | 06:16 |
| *** gongysh has joined #tacker | 06:20 | |
| openstackgerrit | dharmendra kushwaha proposed openstack/python-tackerclient: Body can not be None while creating vnfd https://review.openstack.org/346557 | 06:27 |
| *** saju_m has quit IRC | 06:34 | |
| *** hparekh_ has quit IRC | 06:35 | |
| openstackgerrit | venkatamahesh proposed openstack/tacker: Clean up the imports according to hacking standards https://review.openstack.org/359608 | 06:37 |
| *** gongysh has quit IRC | 06:37 | |
| *** saju_m has joined #tacker | 06:46 | |
| *** tbh has quit IRC | 06:52 | |
| *** sripriya has quit IRC | 06:56 | |
| *** hparekh has joined #tacker | 07:17 | |
| openstackgerrit | Tung Doan proposed openstack/tacker: Alarm monitor: All in one https://review.openstack.org/365435 | 07:24 |
| *** tamilhce has joined #tacker | 07:25 | |
| *** openstackgerrit has quit IRC | 07:33 | |
| *** openstackgerrit has joined #tacker | 07:34 | |
| *** tamilhce has quit IRC | 07:43 | |
| *** tamilhce has joined #tacker | 07:44 | |
| *** dkushwaha has left #tacker | 08:12 | |
| *** gongysh has joined #tacker | 08:48 | |
| gongysh | sridhar_ram, hi | 09:02 |
| gongysh | KanagarajM, hi | 09:11 |
| *** saju_m has quit IRC | 09:11 | |
| openstackgerrit | Kanagaraj Manickam proposed openstack/tacker: VNF scaling: Functional test https://review.openstack.org/363573 | 09:25 |
| *** mbound has joined #tacker | 09:30 | |
| *** diga has quit IRC | 09:43 | |
| *** mfedosin has joined #tacker | 09:45 | |
| openstackgerrit | dharmendra kushwaha proposed openstack/tacker: tacker resource and test_api_v2 refactoring https://review.openstack.org/317749 | 10:08 |
| openstackgerrit | dharmendra kushwaha proposed openstack/tacker: Refactor Tacker unit tests for services/vm https://review.openstack.org/367258 | 10:08 |
| *** achatterjee has quit IRC | 10:10 | |
| *** neel has quit IRC | 10:12 | |
| *** gongysh has quit IRC | 10:21 | |
| *** amotoki has quit IRC | 10:48 | |
| *** KanagarajM has quit IRC | 10:56 | |
| *** gongysh has joined #tacker | 11:37 | |
| *** neel has joined #tacker | 11:39 | |
| *** KanagarajM has joined #tacker | 11:43 | |
| *** links has quit IRC | 11:52 | |
| openstackgerrit | Lu lei proposed openstack/tacker: Correcting spelling mistakes in test_vnfm_param.py https://review.openstack.org/351723 | 11:55 |
| *** hparekh has quit IRC | 12:04 | |
| *** neel has quit IRC | 12:22 | |
| *** amotoki has joined #tacker | 12:48 | |
| *** KanagarajM has quit IRC | 13:03 | |
| *** KanagarajM has joined #tacker | 13:03 | |
| *** bobh has joined #tacker | 13:08 | |
| *** jraju has joined #tacker | 13:12 | |
| *** Vijayendra has quit IRC | 13:16 | |
| openstackgerrit | dharmendra kushwaha proposed openstack/python-tackerclient: Add support for multi delete https://review.openstack.org/349806 | 13:27 |
| *** Vijayendra has joined #tacker | 13:30 | |
| *** KanagarajM has quit IRC | 13:31 | |
| *** dkushwaha__ has joined #tacker | 13:33 | |
| *** KanagarajM has joined #tacker | 13:33 | |
| dkushwaha__ | KanagarajM, ping | 13:37 |
| *** gongysh has quit IRC | 13:38 | |
| dkushwaha__ | KanagarajM, Could you please review the patch https://review.openstack.org/#/c/349806/ | 13:38 |
| *** mbound_ has joined #tacker | 13:39 | |
| *** mbound has quit IRC | 13:41 | |
| *** uck has joined #tacker | 13:51 | |
| *** janki has quit IRC | 13:58 | |
| *** amotoki has quit IRC | 14:01 | |
| *** amotoki has joined #tacker | 14:02 | |
| *** amotoki has quit IRC | 14:17 | |
| vishwanathj_zzz | sridhar_ram hi | 14:18 |
| *** vishwanathj_zzz is now known as vishwanathj | 14:18 | |
| *** amotoki has joined #tacker | 14:28 | |
| *** amotoki has quit IRC | 14:33 | |
| *** janki has joined #tacker | 14:37 | |
| *** uck has quit IRC | 14:53 | |
| *** tamilhce_ has joined #tacker | 14:54 | |
| trozet | sridhar_ram: ping? | 14:55 |
| *** tamilhce has quit IRC | 14:57 | |
| *** jraju has quit IRC | 14:59 | |
| *** KanagarajM has quit IRC | 15:05 | |
| *** uck has joined #tacker | 15:05 | |
| *** uck has quit IRC | 15:05 | |
| *** tung_doan_ has joined #tacker | 15:05 | |
| *** mbound_ has quit IRC | 15:18 | |
| *** amotoki has joined #tacker | 15:29 | |
| *** tbh has joined #tacker | 15:33 | |
| *** amotoki has quit IRC | 15:34 | |
| *** tbh has quit IRC | 15:40 | |
| *** janki has quit IRC | 15:56 | |
| *** tbh has joined #tacker | 15:56 | |
| trozet | sridhar_ram: something works :) https://paste.fedoraproject.org/423973/raw/ | 15:57 |
| *** janki has joined #tacker | 15:58 | |
| *** uck has joined #tacker | 16:00 | |
| *** mbound has joined #tacker | 16:19 | |
| *** uck has quit IRC | 16:23 | |
| *** uck has joined #tacker | 16:23 | |
| *** mbound has quit IRC | 16:24 | |
| *** janki has quit IRC | 16:28 | |
| *** amotoki has joined #tacker | 16:31 | |
| *** amotoki has quit IRC | 16:31 | |
| *** amotoki has joined #tacker | 16:31 | |
| *** KanagarajM has joined #tacker | 16:39 | |
| *** tamilhce_ has quit IRC | 16:44 | |
| *** tbh has quit IRC | 16:50 | |
| *** Vijayendra has quit IRC | 16:56 | |
| *** s3wong has joined #tacker | 17:28 | |
| *** sripriya has joined #tacker | 17:28 | |
| openstackgerrit | Stephen Wong proposed openstack/tacker: VNFFG service chain Neutron networking-sfc driver https://review.openstack.org/347568 | 17:40 |
| *** KanagarajM has quit IRC | 17:52 | |
| *** KanagarajM has joined #tacker | 17:52 | |
| *** LamT_ has joined #tacker | 17:54 | |
| *** bobh_ has joined #tacker | 18:01 | |
| *** bobh has quit IRC | 18:04 | |
| sridhar_ram | trozet: glad to see that! | 18:04 |
| *** vishwanathj is now known as vishwanathj_ | 18:06 | |
| vishwanathj_ | sridhar_ram would appreciate if you could review https://review.openstack.org/#/c/366226/5 to see if its good to merge | 18:07 |
| sridhar_ram | vishwanathj_: sure, will do | 18:07 |
| tung_doan | sripriya: ping | 18:07 |
| sripriya | tung_doan: pong | 18:08 |
| tung_doan | sripriya: it looks like that auth_token does't make sense since vim_auth is supported: https://github.com/openstack/tacker/blob/master/tacker/vnfm/monitor.py#L248 | 18:08 |
| sripriya | tung_doan: yes | 18:09 |
| tung_doan | sripriya: i just wonder whether it should be removed or other ways to get token here? | 18:10 |
| sripriya | tung_doan: i dont think we are using it though | 18:11 |
| tung_doan | sripriya: whether it should have an option here: https://github.com/openstack/tacker/blob/master/tacker/common/clients.py | 18:14 |
| sripriya | tung_doan: it is good to get out of password based authentication and move to token or key based authentication for vims | 18:16 |
| sripriya | tung_doan: let us not digress towards it now. we can take it up as a separate enhancement. | 18:19 |
| tung_doan | sripriya: ack... IMO, it's difficult for requests which do not come from CLI..right? | 18:19 |
| sripriya | tung_doan: for the tacker API requests, you can stil pass token in your requests | 18:21 |
| sripriya | tung_doan: is that your question? | 18:21 |
| tung_doan | sripriya: correct | 18:21 |
| sripriya | tung_doan: the tacker API accepts token as well as password based requests | 18:22 |
| sripriya | tung_doan: similar to all other openstack services | 18:22 |
| tung_doan | sripriya: good news for me.. thanks :) | 18:23 |
| sripriya | tung_doan: sure, let me know if you need further help in this. | 18:23 |
| tung_doan | sripriya: thanks | 18:29 |
| *** santoshk has joined #tacker | 18:33 | |
| *** mfedosin has quit IRC | 18:41 | |
| *** Ravikiran_K has quit IRC | 18:44 | |
| sridhar_ram | vishwanathj_: done | 18:47 |
| bryan_att | sridhar_ram: sripriya: FYI I got my Tacker install script working on OPNFV. Starting to further develop the Hello World blueprint - I need to know how to specify security group association, and if it's possible to create a custom security group (like a network) in the template. | 18:48 |
| sridhar_ram | bryan_att: glad to hear the progress. | 18:49 |
| sridhar_ram | bryan_att: we don't have sec-grp association yet, i believe we have a RFE request out to add that. Let me check its status. | 18:50 |
| bryan_att | thanks - so for now the groups have to be added after-the-fact. That's OK. | 18:51 |
| sridhar_ram | bryan_att: a patchset is out here https://review.openstack.org/#/c/334256/ (from your fellow ATT colleague :) | 18:51 |
| sripriya | bryan_att: cool | 18:51 |
| openstackgerrit | Merged openstack/tacker: VNF monitoring event capture: Functional test https://review.openstack.org/366226 | 18:53 |
| *** vishwanathj_ is now known as vishwanathj | 19:00 | |
| *** mbound has joined #tacker | 19:00 | |
| vishwanathj | sridhar_ram thanks for the merge ... I noted your point about the enhancement, good suggestion | 19:01 |
| sridhar_ram | vishwanathj: urw, btw opened this RFE so that we don't lose track of that.. https://bugs.launchpad.net/tacker/+bug/1621592 | 19:02 |
| openstack | Launchpad bug 1621592 in tacker "RFE: verify specific sequence of monitoring events" [Low,New] | 19:02 |
| vishwanathj | sridhar_ram I am thinking of capturing a launchpad bug to track the enhancement | 19:02 |
| vishwanathj | oh cool | 19:02 |
| openstackgerrit | Merged openstack/python-tackerclient: Allow auth url without port for vim registration https://review.openstack.org/363483 | 19:05 |
| *** prashantD has joined #tacker | 19:07 | |
| trozet | s3wong: ping? | 19:08 |
| bryan_att | sridhar_ram: I'm trying to find a reliable way to find the openstack server instance(s) for a tacker VNF. Any advice? I need the server instance IDs so I can add security groups, as that's an unsupported feature. | 19:26 |
| *** prashantD has quit IRC | 19:26 | |
| sridhar_ram | bryan_att: we just introduced that capability in the master (newton release) | 19:29 |
| sridhar_ram | https://www.irccloud.com/pastebin/6LjNEcgh/tacker-vnf-resource-list | 19:29 |
| bryan_att | ok, that's why that command doesn't work.... I'm on mitaka/stable | 19:30 |
| bryan_att | Are there any backward compatibility issues with running newton tacker on a mitaka openstack base install? | 19:31 |
| *** avishnoi has quit IRC | 19:34 | |
| bryan_att | sridhar_ram: I guess I could associate the server by the management IP address, which should be unique for each server in the VNF, right? (list all instances and find the one(s) that have the IP address associated with the VNF) | 19:34 |
| sridhar_ram | https://www.irccloud.com/pastebin/RpwNEOSn/ | 19:35 |
| sridhar_ram | bryan_att: for mitaka i would suggest to a 2-step process to get the same information. Use the following combination of tacker vnf-show and heat resource-list ... | 19:35 |
| sridhar_ram | bryan_att: if you've a flat network looking up using mgmt-ip could work | 19:36 |
| bryan_att | thanks, I'll try that. I do have a flat network but can't count on that in the test design. | 19:36 |
| *** n-harada has quit IRC | 19:40 | |
| openstackgerrit | Tung Doan proposed openstack/tacker: Alarm monitor: All in one https://review.openstack.org/365435 | 19:43 |
| openstackgerrit | Sripriya Seetharam proposed openstack/tacker: Derive infra driver using VIM type https://review.openstack.org/367599 | 19:50 |
| trozet | s3wong: let me knwo when you get back | 20:00 |
| *** vishnoianil has joined #tacker | 20:04 | |
| bryan_att | sridhar_ram: I have an issue adding security groups - neutron says I need port_security_enabled. How do I spec that in the blueprint? https://www.irccloud.com/pastebin/xaGfU2XF/ | 20:14 |
| openstackgerrit | Sripriya Seetharam proposed openstack/tacker: Derive infra driver using VIM type https://review.openstack.org/367599 | 20:19 |
| openstackgerrit | vishwanath jayaraman proposed openstack/tacker-horizon: Adds audit support for VNF, VIM and VNFD resources https://review.openstack.org/349755 | 20:28 |
| vishwanathj | amotoki Hi | 20:28 |
| vishwanathj | amotoki I am looking for horizon related guidance .... I am trying to put a refresh button in a tab .... this is in reference to patchset https://review.openstack.org/349755 | 20:30 |
| openstackgerrit | Sripriya Seetharam proposed openstack/tacker: Rename heat unit tests to openstack https://review.openstack.org/367623 | 20:41 |
| s3wong | trozet: ping | 20:42 |
| s3wong | trozet: actually it is a pong | 20:42 |
| *** sripriya has quit IRC | 20:43 | |
| bryan_att | sridhar_ram: I hink the issue is in how Tacker sets the port_security_enabled value in the generated Heat template. How do I get this set to "true"? https://www.irccloud.com/pastebin/7eFx1ZA3/ | 20:44 |
| bryan_att | trozet or sripriya: if you have any ideas about the question I just asked sridhar_ram please let me know. | 20:45 |
| *** sripriya has joined #tacker | 20:51 | |
| *** tung_doan_ has quit IRC | 20:53 | |
| *** KanagarajM has quit IRC | 20:55 | |
| *** KanagarajM has joined #tacker | 20:58 | |
| trozet | s3wong: so i downloaded stable/mitaka networking-sfc, that should work with your driver, right? | 21:03 |
| s3wong | trozet: I think that is the one I tested | 21:04 |
| trozet | s3wong: ok good. Then also, is there a networking-sfc irc channel? | 21:04 |
| s3wong | trozet: that is the line I added in local.conf: enable_plugin networking-sfc git://git.openstack.org/openstack/networking-sfc stable/mitaka | 21:05 |
| s3wong | trozet: no | 21:05 |
| s3wong | trozet: presumably you can ping us on openstack-neutron; but TBH most of us isn't normally available on there :-) | 21:05 |
| s3wong | trozet: couple meetings ago there was a suggestion to add one, but don't think cathy_ ever did it | 21:06 |
| trozet | s3wong: ok so next question, after I python installed networking-sfc, what do i need to do to make it work | 21:07 |
| trozet | s3wong: do i need to change neutron.conf settings? Also, if I want to test that it just works by itself (create a chain, etc) where is the python client? | 21:08 |
| s3wong | trozet: networking-sfc CLIs are automatically added to neutron python client | 21:08 |
| s3wong | trozet: TBH I've only done devstack installation | 21:09 |
| s3wong | trozet: I actually don't know how stadium packages are loaded to Neutron | 21:09 |
| s3wong | trozet: but for sure I don't believe you need to modify neutron.conf | 21:10 |
| trozet | s3wong: do you know where hte client code lives? | 21:11 |
| s3wong | trozet: even our link https://wiki.openstack.org/wiki/Neutron/ServiceInsertionAndChaining didn't have anything on full OpenStack install, just devstack :-) | 21:11 |
| trozet | s3wong: oh wait it looks like just installing the networking-sfc part updated the client as well on my control node | 21:12 |
| trozet | s3wong: so i'm guessing if i also install that on my pc, then neutronclient will pick up the change | 21:12 |
| trozet | let me try | 21:12 |
| s3wong | trozet: and this isn't too useful:http://docs.openstack.org/developer/networking-sfc/installation.html | 21:12 |
| trozet | s3wong: hehe i saw that | 21:12 |
| s3wong | trozet: yeah, I would imagine the pip install does some magic | 21:12 |
| s3wong | trozet: the networking-sfc CLI codes are just extension of Neutron client classes; my guess is it invokes a load to install into neutron client | 21:13 |
| trozet | creating build/lib/networking_sfc/cli | 21:13 |
| trozet | copying networking_sfc/cli/__init__.py -> build/lib/networking_sfc/cli | 21:13 |
| trozet | copying networking_sfc/cli/flow_classifier.py -> build/lib/networking_sfc/cli | 21:13 |
| trozet | s3wong:^ | 21:13 |
| trozet | s3wong: yeah looks like it just copies them in | 21:13 |
| s3wong | trozet: should magically work | 21:13 |
| trozet | s3wong: cool i see them there now | 21:14 |
| s3wong | trozet: since these modules should all have the same PYTHONPATH | 21:14 |
| trozet | s3wong: ok so now i've got my patch and networking-sfc going | 21:14 |
| s3wong | trozet: finger cross | 21:14 |
| trozet | s3wong: so first im going to test without your driver and just return a random uuid | 21:14 |
| trozet | s3wong: then i'll migrate your driver into the openstack_driver and test the whole thing | 21:14 |
| s3wong | trozet: OK | 21:14 |
| s3wong | trozet: exciting... | 21:14 |
| trozet | s3wong: but at least hte VNFFGD create works, with tosca parsing | 21:14 |
| trozet | s3wong: from client -> servcer | 21:15 |
| s3wong | trozet: cool | 21:15 |
| trozet | s3wong: also, i know you are busy, but i took the patch out of WIP and updated it last night, in case you want to review | 21:15 |
| s3wong | trozet: so as long as plugin -> driver path works, we are home free :-) | 21:15 |
| *** tbh has joined #tacker | 21:16 | |
| s3wong | trozet: seems like LouisF and sridhar_ram will rely on our progress to have a demo for their summit talk :-) | 21:16 |
| trozet | s3wong: that would be awesome | 21:16 |
| s3wong | trozet: I am guessing you won't be going to BCN for O-Summit? | 21:17 |
| trozet | s3wong: nah going to skip that one | 21:22 |
| s3wong | trozet: me too --- I won't be going | 21:22 |
| trozet | s3wong: you going to be at the next one, in boston right? | 21:22 |
| s3wong | trozet: may be... | 21:22 |
| s3wong | trozet: you? | 21:23 |
| trozet | s3wong: probably, that one is pretty close to me :) | 21:30 |
| *** KanagarajM_ has joined #tacker | 21:32 | |
| *** KanagarajM has quit IRC | 21:32 | |
| *** sripriya has quit IRC | 21:35 | |
| *** sripriya has joined #tacker | 21:37 | |
| *** tbh has quit IRC | 21:42 | |
| sridhar_ram | s3wong: | 21:57 |
| sridhar_ram | s3wong: ping | 21:58 |
| trozet | sridhar_ram: ping? | 22:01 |
| *** bobh_ has quit IRC | 22:01 | |
| s3wong | sridhar_ram: pong | 22:08 |
| *** mbound has quit IRC | 22:08 | |
| *** KanagarajM_ has quit IRC | 22:22 | |
| *** KanagarajM has joined #tacker | 22:23 | |
| * sridhar_ram is back now .. got pulled away | 22:44 | |
| sridhar_ram | trozet: whats up? | 22:44 |
| sridhar_ram | s3wong: we need a devstack patch to do enable_plugin for networking-sfc ? | 22:44 |
| *** santoshk has quit IRC | 22:47 | |
| *** dkushwaha__ has quit IRC | 22:50 | |
| *** KanagarajM has quit IRC | 22:59 | |
| *** KanagarajM has joined #tacker | 23:00 | |
| s3wong | sridhar_ram: sorry, I too was being pulled away | 23:15 |
| s3wong | sridhar_ram: do you want to? | 23:15 |
| s3wong | sridhar_ram: I can certainly add it in our devstack sample local.conf | 23:16 |
| s3wong | sridhar_ram: it is needed (of course) for VNFFG to work with default driver | 23:16 |
| sridhar_ram | s3wong: yes.. please add it to your existing patchset | 23:16 |
| s3wong | sridhar_ram: OK | 23:16 |
| sridhar_ram | trozet: s3wong: i'm pulling the patchset to give this feature a spin.. | 23:17 |
| s3wong | sridhar_ram: great, more people testing it :-) | 23:17 |
| sridhar_ram | trozet: s3wong: what VMs do u use ? how to initial the pkt flow to "see" the chaining happening ? | 23:18 |
| s3wong | sridhar_ram: I got to be honest, I have NOT done traffic test | 23:18 |
| sridhar_ram | argh.. "how to you initiate the pkt flow .. " | 23:18 |
| sridhar_ram | s3wong: okay | 23:18 |
| s3wong | sridhar_ram: I only "functionally" test it to ensure the DB reflects what I set | 23:19 |
| s3wong | sridhar_ram: I actually haven't even looked into the OVS flow entries | 23:19 |
| sridhar_ram | s3wong: do you know how traffic tests happen in the n-sfc project ? | 23:19 |
| s3wong | sridhar_ram: cathy_ has a demo back in Tokyo | 23:20 |
| s3wong | sridhar_ram: it is multi-nodes though | 23:20 |
| s3wong | sridhar_ram: so pretty comprehensive | 23:20 |
| sridhar_ram | s3wong: i was thinking we could use two cirros "VNFs" and use a simple dest_port policy to see if traffic redirection happens | 23:21 |
| sridhar_ram | s3wong: for now, i'm interested in point checking the FFG feature.. | 23:21 |
| sridhar_ram | s3wong: looks the "traffic testers" (!!) would be better off trying the n-sfc testcase first and then .. trying the exact same scenario using TOSCA FFGD template | 23:22 |
| s3wong | sridhar_ram: sure --- the intent isn't testing networking-sfc, it is to test if the template can resolved down to correct OVS flows | 23:22 |
| s3wong | sridhar_ram: if you only have one node, the thing you miss is VxLAN tunnel (and hence the MPLS tag placed) | 23:23 |
| s3wong | sridhar_ram: but we don't care about that | 23:23 |
| sridhar_ram | s3wong: exactly.. don't care abt vxlan vs mpls | 23:23 |
| sridhar_ram | s3wong: let me know if you find a cheat sheat to test n-sfc using CLIs | 23:23 |
| s3wong | sridhar_ram: for traffic test with just networking-sfc, a description on set up and how it works is here:https://wiki.openstack.org/wiki/Neutron/ServiceInsertionAndChaining | 23:24 |
| sridhar_ram | s3wong: thanks, will check it out | 23:24 |
| sripriya | sridhar_ram: you can check this if it helps for cli commands http://docs.openstack.org/developer/networking-sfc/system_design%20and_workflow.html#boot-service-vms-and-attach-ports | 23:24 |
| sridhar_ram | sripriya: thanks, that helps! | 23:26 |
| sridhar_ram | s3wong: need one more help.. | 23:27 |
| sridhar_ram | s3wong: the parent of your patchset still points to the ffg abstract driver patchset which already merged... can you please rebase ur patchset ? | 23:27 |
| *** KanagarajM has quit IRC | 23:33 | |
| *** KanagarajM has joined #tacker | 23:33 | |
| *** gongysh has joined #tacker | 23:40 | |
| *** dkushwaha has joined #tacker | 23:45 | |
| *** KanagarajM has quit IRC | 23:49 | |
| *** KanagarajM has joined #tacker | 23:49 | |
| openstackgerrit | Sripriya Seetharam proposed openstack/tacker: Derive infra driver using VIM type https://review.openstack.org/367599 | 23:57 |
| s3wong | sridhar_ram: oh... right... it is merged, no need to have a parent | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!