08:05:46 <yasufum> #startmeeting tacker 08:05:47 <openstack> Meeting started Tue Feb 2 08:05:46 2021 UTC and is due to finish in 60 minutes. The chair is yasufum. Information about MeetBot at http://wiki.debian.org/MeetBot. 08:05:48 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 08:05:50 <openstack> The meeting name has been set to 'tacker' 08:07:14 <yasufum> There is no topic today on our etherpad 08:07:49 <yasufum> Do you have anything to discuss? 08:08:37 <LiangLu> I have 1 topic to discuss, would like to hear opinions from you, could I go on? 08:09:30 <yasufum> please go ahead 08:09:44 <LiangLu> https://etherpad.opendev.org/p/tacker-meeting#286 08:10:28 <LiangLu> sorry for write on Etherpad late.. This is a topic about Backward compatibility for patch https://review.opendev.org/c/openstack/tacker/+/768091 08:11:51 <LiangLu> in this patch we are trying to refactoring legacy action-driver code, to support SOL-001 based VNFLCM interfaces 08:12:39 <LiangLu> also we are trying to make sure user are using tacker's mgmt-driver(like kubernetes.py) in their package, by checking hash value. 08:14:00 <LiangLu> however, this patch will also make legacy action types like ACTION_SCALE_IN_VNF not run anymore since wo changed mgmt calling method in vnflcm_driver 08:15:57 <LiangLu> this makes incompletely backward compatibility.. And here what I want to discuss is, should this change keep legacy actions support? 08:19:45 <LiangLu> my opinions are: 1. we should or we are excepted to support SOL-001 based interfaces, 2. It is not a long time since legacy code released(in Victoria), so maybe it is a good timing to do a complete refactoring now. 08:20:09 <LiangLu> that's all 08:22:18 <yoshito-ito> I understand the point is; The scale operation in SOL-implementation calls legacy-MgmtDriver. This will be refactored to call your new MgmtDriver for SOL. 08:23:16 <yoshito-ito> I remember we discussed the design of new MgmtDriver for SOL-implementation in the previous PTG, and my opinion was to create completely new one. 08:23:42 <yoshito-ito> Because the methods for new MgmtDriver are different from legacy. 08:23:59 <yoshito-ito> So I agree. 08:24:20 <yoshito-ito> https://etherpad.opendev.org/p/Tacker-PTG-Wallaby 08:24:32 <yoshito-ito> L183 08:25:45 <yoshito-ito> takahashi-tsc is not here today? 08:26:39 <LiangLu> thanks for opinion! 08:28:03 <tsukasasa> takahashi-tsc is attending another meeting. 08:30:22 <yasufum> I am not opposite to your proposal 08:30:47 <takahashi-tsc> Hi, sorry for very late... 08:30:49 <yasufum> but you should revise your patch 08:30:52 <yasufum> hi 08:31:05 <yoshito-ito> hi takahashi-san 08:31:43 <yoshito-ito> please find https://etherpad.opendev.org/p/tacker-meeting#286 08:32:53 <takahashi-tsc> Thanks, I read meeting log and understand that we are talking about Mgmt Driver. 08:35:43 <yasufum> do you have any comments? 08:39:18 <takahashi-tsc> I remember that our conclusion in PTG is to create new one for SOL. In my opinion, it is better to divide. 08:43:22 <yasufum> LiangLu: What do you think? 08:46:04 <LiangLu> thanks for opinion, I think we don't need to fix patch to support legacy interfaces(is already modified in current patch) 08:46:35 <LiangLu> and also may I ask what else should we revise for patch? 08:47:24 <yasufum> You are proposing this topic because Itsuro suggested to discuss here, right? 08:47:53 <yasufum> If so, you should add an explanation in docs or so as he advised at least. 08:48:36 <yasufum> In addition, you do not make clear about backward compatibility 08:49:03 <yasufum> keep the legacy one, or discard, or so... 08:49:32 <yasufum> I do not understand your point discussed from your current patch actually 08:50:29 <yasufum> Itsuro also suggested to make a decision at this IRC meeting. I agree with him. 08:54:13 <LiangLu> thanks for suggestion, my opinion is to discard legacy codes, I will add this part to docs/commit message 08:56:07 <yasufum> takahashi-tsc: I am not sure about your comment, “it is better to divide”. It means the same as LiangLu? 08:58:16 <takahashi-tsc> Sorry for ambiguous, I said just to create new MgmtDriver feature for SOL. And "Discard legacy or not" seems another discussion 09:00:47 <yasufum> thanks 09:01:50 <takahashi-tsc> In my opinion, we should not discard legacy suddenly(e.g. W-release), so we should make some docs to explain. 09:02:07 <takahashi-tsc> And separately, we discuss and decide the strategy about legacy code. 09:02:17 <yoshito-ito> Let me correct the point of discussion 09:03:21 <yoshito-ito> LiangLu doesn't propose discarding of Legacy Mgmt-Driver 09:03:56 <yoshito-ito> He is just trying to drop the unintentional logic of calling legacy-MgmtDriver in the SOL scale operation. 09:05:32 <yoshito-ito> As we discussed in PTG, we agreed to have different MgmtDrivers for legacy and SOL. But we found that the scale operation supported in V-release has a logic calling legacy MgmtDriver. 09:05:36 <LiangLu> thanks for explanation! 09:06:13 <yoshito-ito> I'm not sure why that is there, but the point is that I think we can drop the unintentional logic related to legacy MgmtDriver in SOL scale operation. 09:07:08 <takahashi-tsc> Understood... I agree with LiangLu SOL operation should not call legacy-MgmtDriver. 09:09:21 <LiangLu> sorry for make things ambiguous, the conflict part(sol/legacy) is in vnflcm, calling scale operation, in the patch we are modifying to call SOL operation 09:10:41 <yasufum> I am still not sure about “3rd-party drivers” discussed while reviewing. 09:11:14 <yasufum> Is it not included in your scope of the change actually? 09:17:11 <yasufum> anyway, we don’t have time today. 09:17:48 <LiangLu> should we continue discuss on patch reviews? 09:17:55 <yasufum> LiangLu: please continue on gerrit if you have no comment now 09:18:16 <yasufum> or make it clear shortly if possible 09:19:39 <LiangLu> I think it is better to continue in reviews 09:19:59 <yasufum> OK, thanks 09:20:01 <LiangLu> sorry for taking long time.. 09:20:27 <yasufum> I would like to ask you to update your patch first. 09:20:37 <yasufum> Thanks everyone 09:20:49 <yasufum> I would like to close this meeting 09:20:52 <yasufum> bye 09:20:55 <yoshito-ito> bye 09:21:01 <masazumi-ota> bye 09:21:04 <takahashi-tsc> thanks 09:21:08 <yasufum> #endmeeting