BRIDGE: https://zoom.us/j/661303200?pwd=TFdRd0c2MTJUem8xa252UGJHTE1Mdz09

Passcode: 209247

We will start our meetings by mentioning the project's Antitrust Policy, which you can find linked from the LF and project websites. The policy is important where multiple companies, including potential industry competitors, are participating in meetings. Please review and if you have any questions, please contact your company legal counsel. Members of the LF may contact Andrew Updegrove at the firm Gesmer Updegrove LLP, which provides legal counsel to the LF.

Attendance is taken purely upon #info in Zoom Chat 

Zoom Chat Log 

06:59:56 From  Fernando (Fred) Oliveira  to  Everyone:
    #info Fred Oliveira, Verizon
07:00:17 From  Catherine Lefevre  to  Everyone:
    #info Catherine lefevre, ATT
07:00:32 From  Ciaran Johnston (Ericsson)  to  Everyone:
    #info Ciaran Johnston, Ericsson
07:00:33 From  Alla Goldner  to  Everyone:
    #info Alla Goldner, Amdocs
07:00:40 From  Yuanhong Deng (China Mobile)  to  Everyone:
    #info proxy Yuanhong Deng, China Mobile
07:00:48 From  Ranny HAIBY (Samsung)  to  Everyone:
    #info Ranny Haiby, Samsung
07:01:23 From  Timo Perala (Nokia)  to  Everyone:
    #info Timo Perala, Nokia
07:01:50 From  Michal JAGIELLO (T-Mobile PL)  to  Everyone:
    #info proxy Michal Jagiello, DT
07:01:53 From  bin.yang@windriver.com  to  Everyone:
    #info Bin Yang, Wind River
07:02:19 From  Martin Vezeau (Bell Canada)  to  Everyone:
    #info Martin Vezeau, Bell Canada
07:02:30 From  Shashikanth v.h.  to  Everyone:
    #info proxy Seshu, Huawei
07:03:08 From  Dong Wang (China Telecom)  to  Everyone:
    #info Dong Wang, China Telecom
07:03:44 From  Jason Hunt  to  Everyone:
    #info Jason Hunt, IBM
07:04:16 From  Eric Debeau  to  Everyone:
    #info Eric Debeau, Orange
07:04:33 From  Srini Addepalli (Intel)  to  Everyone:
    #info Srini Addepalli, Intel
07:07:06 From  Kenny PAUL (LFN)  to  Everyone:
    #topic Istanbul M4
07:10:12 From  Sai Seshu  to  Everyone:
    #info Seshu, huawei
07:20:13 From  Kenny PAUL (LFN)  to  Everyone:
    Eric's line keeps dropping out.
07:20:46 From  Eric Debeau  to  Everyone:
    Sylvain will connect
07:20:52 From  Kenny PAUL (LFN)  to  Everyone:
    Silence suppression issues maybe?
07:24:06 From  William Reehil (AT&T)  to  Everyone:
    https://gerrit.onap.org/r/c/oom/+/123671
07:24:25 From  Kenny PAUL (LFN)  to  Everyone:
    Last AAI issue appears to be an SO connection to MariaDB - unsure why it would be an AAI issue
07:30:53 From  Sai Seshu  to  Everyone:
    Ok
07:31:47 From  ONAP Meeting 2  to  Everyone:
    For AAI patch, SO issue may be transient. Unfortunately we were not lucky on this patch which got a lot of « gate internal issues »
07:31:50 From  William Reehil (AT&T)  to  Everyone:
    aai merge conflict resolved in oom
07:32:19 From  ONAP Meeting 2  to  Everyone:
    So let’s wait for the next gate to really check and hopefully we’ll be able to merge :)
07:32:55 From  ONAP Meeting 2  to  Everyone:
    It  was Morgan’s comment
07:33:08 From  ONAP Meeting 2  to  Everyone:
    (For the documentation part on use case)
07:33:32 From  Sai Seshu  to  Everyone:
    I second the thought of so blocking the gating... Let's see on how to proceed, also we will have other pods updates as per the testing needs
07:35:42 From  William Reehil (AT&T)  to  Everyone:
    Yeah I was wondering if the issue was transient or not, hopefully the next build succeeds
07:43:08 From  Kenny PAUL (LFN)  to  Everyone:
    Link to page of GCs (in the context of when they were approved)https://lf-onap.atlassian.net/wiki/x/-Ub7
07:45:29 From  Kenny PAUL (LFN)  to  Everyone:
    "when" == which release
07:52:02 From  Eric Debeau  to  Everyone:
    Keep moving
07:55:37 From  Catherine Lefevre  to  Everyone:
    3 parts
07:56:27 From  Catherine Lefevre  to  Everyone:
    for the conditions:
07:56:35 From  Catherine Lefevre  to  Everyone:
    GR Status by 9/23
07:56:58 From  Catherine Lefevre  to  Everyone:
    Project tasks and use case tasks by 9/30?
07:57:57 From  Catherine Lefevre  to  Everyone:
    of M4
07:58:45 From  Eric Debeau  to  Everyone:
    You must check the Doc remaining open tickets. One of thme should be closed
08:01:04 From  Sylvain Desbureaux (Orange)  to  Everyone:
    Vid needs portal
08:01:19 From  Sylvain Desbureaux (Orange)  to  Everyone:
    I don’t know ^_^
08:02:26 From  Sai Seshu  to  Everyone:
    Yes, but can run on h release pods
08:02:50 From  Kenny PAUL (LFN)  to  Everyone:
    #vote does the TSC approves the Istanbul release as passing the M3 milestone with the condition that the      Global requirements tasks status is understood by Sept. 23,       Project and Use Case tasks are to be closed by   Sept 30     Doc tasks are closed by Sept 30?
08:03:42 From  Sai Seshu  to  Everyone:
    Isnt it m4?
08:03:52 From  Kenny PAUL (LFN)  to  Everyone:
    arrrg!
08:04:00 From  Sai Seshu  to  Everyone:
    :)
08:04:15 From  Kenny PAUL (LFN)  to  Everyone:
    #vote does the TSC approves the Istanbul release as passing the M4 milestone with the condition that the      Global requirements tasks status is understood by Sept. 23,       Project and Use Case tasks are to be closed by   Sept 30     Doc tasks are closed by Sept 30?
08:05:02 From  Sai Seshu  to  Everyone:
    #vote +1
08:05:06 From  Catherine Lefevre  to  Everyone:
    #vote +1
08:05:09 From  Fernando (Fred) Oliveira  to  Everyone (in Waiting Room):
    #vote +1
08:05:10 From  Ranny HAIBY (Samsung)  to  Everyone:
    #vote +1
08:05:17 From  Jason Hunt  to  Everyone:
    #vote +1
08:05:32 From  Sai Seshu  to  Kenny PAUL (LFN)(Direct Message):
    I got to drop, Sashi will be my proxy for the rest of the call... Thanks
08:05:36 From  Timo Perala (Nokia)  to  Everyone:
    #vote +1
08:05:48 From  Kenny PAUL (LFN)  to  Sai Seshu(Direct Message):
    ok
08:05:48 From  Catherine Lefevre  to  Everyone:
    Doc tasks = tasks assigned to the Document project
08:06:02 From  Sai Seshu  to  Kenny PAUL (LFN)(Direct Message):
    Thanks Kenny..
08:06:23 From  Eric Debeau  to  Everyone:
    #vote +1
08:07:02 From  Alla Goldner  to  Everyone:
    #vote +1
08:07:49 From  Dong Wang (China Telecom)  to  Everyone:
    #vote +1
08:08:51 From  David McBride  to  Everyone:
    OC-760
    DOC-761
    USECASEUI-596
    USECASEUI-597
    USECASEUI-602
    CCSDK-3444
    INT-1965
    INT-1966
    MSB-612
    MSB-614
    MULTICLOUD-1389
    MULTICLOUD-1390
    MULTICLOUD-1392
    OOM-2807
    SDNC-1596
    SO-3751
    SO-3756
    DCAEGEN2-2900
    DCAEGEN2-2898
    VFC-1877
    CPS-606
    MSB-610
    REQ-914
    REQ-902
08:09:22 From  David McBride  to  Everyone:
    The first item was cut off.  It should be DOC-760
08:09:40 From  Kenny PAUL (LFN)  to  Everyone:
    ^^list of actual Project & Usecase tasks . Thanks David.
08:11:16 From  Kenny PAUL (LFN)  to  Everyone:
    Will leave the vote open for 2 more mins.
08:13:46 From  bin.yang@windriver.com  to  Everyone:
    #vote +1
08:14:10 From  Kenny PAUL (LFN)  to  Everyone:
    #endvote
08:18:17 From  Kenny PAUL (LFN)  to  Everyone:
    #AGREED the TSC approves the Istanbul release as passing the M4 milestone with the condition that the           - Global requirements tasks status is understood by Sept. 23,            - The following Project and Use Case tasks are to be closed by   Sept 30:      DOC-760     DOC-761     USECASEUI-596     USECASEUI-597     USECASEUI-602     CCSDK-3444     INT-1965     INT-1966     MSB-612     MSB-614     MULTICLOUD-1389     MULTICLOUD-1390     MULTICLOUD-1392     OOM-2807     SDNC-1596     SO-3751     SO-3756     DCAEGEN2-2900     DCAEGEN2-2898     VFC-1877     CPS-606     MSB-610     REQ-914     REQ-902
08:19:24 From  Toine SIEBELINK (EST)  to  Everyone:
    Practical question: can PTLs go ahead and creating Istanbul branches in gerrit at this stage?
08:20:51 From  Catherine Lefevre  to  Everyone:
    @Toine, it is up to you - considering that you will need to maintain master & branch
08:22:08 From  Toine SIEBELINK (EST)  to  Everyone:
    ok, clear, thanks @Catherine
08:31:09 From  Jason Hunt  to  Everyone:
    good stuff… thanks
08:31:49 From  Kenny PAUL (LFN)  to  Everyone:
    TSC ELECTIONS - NOMINATIONS CLOSE Sept 21st!!!!!     https://lists.onap.org/g/onap-tsc/message/8095
08:33:22 From  Kenny PAUL (LFN)  to  Everyone:
    ONAP Community Awards - nominations  close tomorrow:  https://lf-onap.atlassian.net/wiki/x/3X37

Zoom auto-transcript service - These are often translated incorrectly and can be misleading. They are NOT Authoritative!   Information as to why .
They are included here as a time stamp cross-reference for the recording only!  The notes above this line and the actual recordings are authoritative. 

07:03:36 transcription service has been enabled.
07:03:40 Recording has been enabled.
07:03:45 Please keep yourself muted.
07:03:46 When you come in.
07:03:48 If you are on a phone you can use star six. If you happen to send me a private chat message it will become part of the public record when I cut and paste everything into the minutes.
07:03:58 Jason I see you. Thank you, sir.
07:04:05 As always, we'll start off our meeting by mentioning our anti trust policy notice this policy is important, where we have multiple companies, potentially, including potential industry competitors participating in these meetings.
07:04:19 Please review it, you have any questions you can contact your company's legal counsel, members of the lF may also contact and dropped growth from Gizmo up the grove LLP which provides legal counsel to the Linux Foundation.
07:04:33 Thank you, Eric.
07:04:42 stop sharing that.
07:04:40 And I will share this rainy Thank you.
07:04:49 Good morning, good afternoon, good evening to everyone. Good night, wherever you may be. I'm feeling irrationally cheerful this morning.
07:05:02 And I hope everyone else's irrationally cheerful to go through and talk about the agenda will.
07:05:11 We've got him for scheduled today so that'll take up a good chunk of our time. release status.
07:05:22 We'll talk about rail Inge.
07:05:23 A couple of quick BTL updates.
07:05:26 Congratulations to near and Jana, who's a new dc converter. I want to say thanks to Peter who was stepping down from that role.
07:05:41 Discussion of D map, and the architectural evolution of it is on the agenda will cover PSC elections community awards and our usual
07:05:55 housekeeping items anything else anyone would like to add.
07:06:03 Okay, then I will stop sharing and turn it over to the capable hands of the own app release manager Mr David McBride thunderous applause.
07:06:19 Thanks Kenny.
07:06:23 All right. Well, I think everyone.
07:06:28 If nothing else, based on my numerous emails.
07:06:34 Realize that M for scheduled for today.
07:06:37 Let's just take a quick look at the schedule to remind everyone, where we're at. cm for feature freeze.
07:06:47 September 16.
07:06:50 Our first release candidate comes up next in mid October.
07:07:00 So in terms of the milestone status.
07:07:07 We've actually done pretty well i think
07:07:13 it's pull that up.
07:07:25 Okay, so for projects, we're at about 86% complete for the requirement owners, we're at about 95% complete.
07:07:42 So that's, that's actually pretty good.
07:07:48 Couple of things I wanted to point out one in particular
07:07:54 are some important documentation tasks. So, I believe, I think, Thomas is on holiday. Right now, I'm Andrea. Are you available.
07:08:14 Andreas is on holiday as well, is filling in for him.
07:08:20 Okay.
07:08:23 So, we have.
07:08:26 So, one of our are a couple of our, sorry, let me start over. One of our important tasks for him for his related documentation, and that is that
07:08:40 the documentation team goes in and verifies that the repository changes, and the preliminary documentation tasks for am three and four have been completed.
07:08:57 So since our documentation team is available.
07:09:03 We've been unable to complete those tasks. So the best thing that I can think of right now is to defer those tasks to the release candidate milestone.
07:09:20 Other other milestones that are still being worked on are about what you would expect.
07:09:34 So variety of completing the preliminary documentation.
07:09:39 Completing the assignment of jury issues to the release, etc.
07:09:46 Yep.
07:09:48 Someone have a question
07:09:50 or just random shouting.
07:09:53 Okay, random shouting.
07:09:55 No problem.
07:10:01 So, but we have, we have gotten up over 85%. And so I think it's worth.
07:10:11 We're going to look at status of containers and so on.
07:10:14 So the question comes down to do we want to delay the milestone further, or is this rate of completion, or level of completion sufficient that we can declare the milestone achieved with conditionally achieved based on completing the romaine tasks by the
07:10:43 by the release candidate milestone.
07:10:46 I think one thing that plays into that is the status of the, the Docker container delivery. And
07:11:00 so, excuse me, I, I took a shot at using Morgan's new tool and send out an email and, Of course, immediately.
07:11:12 There were a number of problems with that, I will unfortunately I was using, I was using a link from the demo that Morgan us some time ago I didn't realize that it was date specific so.
07:11:25 So, all the, the, the link that I sent out was reflecting the information from about three weeks ago so obviously that's, that's good. So,
07:11:40 Morgan maybe help me out here How do I how do I navigate to the latest.
07:11:49 No, I get to it from the stage.
07:11:53 He's not on the bridge. Oh, okay.
07:11:59 Hey, I can help you.
07:12:01 Okay.
07:12:01 Yeah and I get there from here.
07:12:04 Maybe I can help you.
07:12:06 I believe that you seen the light. Well, can you repeat what are we looking forward for the Docker version.
07:12:16 Oh, well, what I wanted to do is get to Morgan's Docker version evolution tool. I had a URL that I recorded from his demo, but it's date specific but I'm not sure how to, how to navigate to the, the tool with the current data.
07:12:37 Okay, I believe that.
07:12:42 No, it's not this one.
07:12:45 David.
07:12:51 Okay Where should I start was a was a test on this page it's not supposed to go.
07:12:58 But I think you you already missing you. Presently the just before.
07:13:05 And yeah, it was well I was showing up to bed.
07:13:14 Sorry, some interference here.
07:13:18 Yeah I had, I have this link.
07:13:21 But this link. This is a it links to an old version of the data.
07:13:29 So this is from August 23, this is was from my notes yeah it looks like there's a date in the URL. Yeah. So how do I get to the current data. I guess what I'm asking
07:13:49 is the conversion.
07:13:55 I guess I could just manipulate the date.
07:13:59 Maybe.
07:14:00 Or maybe not.
07:14:08 How about change the URL to without the dates and see if you go on the top level, yeah that that's what I get really good, go to the debt just cut the stuff off before daily before on up that follows own app daily pod master so get rid of right yeah well
07:14:27 that and you should go to some place reason you shouldn't be able to search some things, right, that's what I did earlier and people told me no you can't get there from here today.
07:14:40 Ah, so the statement that's the testing but just now go back one, just go slowly go down.
07:14:49 Maybe.
07:14:51 Sorry. Yeah, Erica finger. Yeah, Eric you keep dropping in and out.
07:14:56 So we're not hearing complete sentences.
07:15:10 leaks into main page for you and not below us.
07:15:14 So I tried to I explained that probably wrong. You just have to add a slash index dash versions that HTML and then you have the version of the, of the doctor but tomorrow morning there will be a link so it would be easier to jump into the different pages
07:15:32 of stability graph, and the version of the doctor. Well, that's okay.
07:15:34 Morgan conveniently and he has email this morning, listed the list of the ones that are still missing so.
07:15:43 So we're still looking for containers from a chai.
07:15:51 And then MSP I think I, I saw an email regarding MSP. This morning I think that's in process vid but it isn't maintained. And now are Do we still require container from bed do to dependencies or do we not need this.
07:16:13 Anyone know.
07:16:17 And then finally, the NF SDK. So, according to Morgan's reading of the status.
07:16:27 This is what's currently missing.
07:16:39 And if any of the details or representatives of any of these projects want to make any comments that these, feel free.
07:16:48 So, go ahead.
07:16:52 Okay, well we have two options either. I didn't see any dependency on with fast or we stumble.
07:17:00 So, I know we if we needed. We do like we do. Usually, we take the latest documentation which is available.
07:17:10 And if there is an issue with you will we fix it.
07:17:13 But since it is not part of his computer officially. If there is a need, let's take the container of only.
07:17:27 Okay.
07:17:30 If someone else was going to say something. Yeah, I'm gonna comment on Annie I so I've had my om, you know, container bump.
07:17:38 Commit open for over two weeks now.
07:17:42 And originally there was some failures and we fix those.
07:17:47 And then, the only thing that was failing was so it didn't seem anything that had to do with AI.
07:17:53 So you know it's been like that for a couple of days, and then someone just merge something that put it into conflict, so I'm trying to fix that.
07:18:02 But
07:18:04 you know that I didn't see any reason why a eyes.
07:18:13 Shouldn't be merged. And it just keeps timing out It seems like so someone can comment on what's going on.
07:18:23 Silvana or restart.
07:18:35 Just, just as this community my desk when we become concerned.
07:18:40 So far, I batch. The issue is that at least one or two components, I don't remember that.
07:18:52 Starting on the trading so we cannot communicate. The same for is DC with a new version which is the season.
07:18:58 Maybe some troubles that, but that's those are good at em fours and goodies to provide a try for the new version and the good is ready to test it and as ear for it we see that there is an issue somewhere doesn't prevent anybody I guess to update after
07:19:20 the review on Garrett, with a new book of actions that fix the issues that we have seen that so good of employees to to propose decamps in.
07:19:36 And usually look at is ok we manage it and if not, we try to find the solution. And between info and FC.
07:19:41 You're,
07:19:44 you're dropping out or SC, we
07:19:53 deserve.
07:19:57 up to the project can write us have good Wi Fi
07:20:21 issues that have been found in in ze.
07:20:21 Just know.
07:20:22 Okay, we're missing a lot of what you're saying to post to be done.
07:20:29 We've we missed quite a bit of that because your connection is dropping now.
07:20:36 But if I, if I got the gist of what you're saying.
07:20:50 Here, you're saying that these
07:20:52 drugs are just.
07:20:55 Okay.
07:20:58 I've seen it up.
07:21:01 I would put in the chat.
07:21:04 Okay.
07:21:07 If I understood correctly so Bond was saying that the fact that these are not currently available is not necessarily gating for him for
07:21:22 that these issues can be resolved.
07:21:26 Between them for and the release candidate was definitely to set up.
07:21:32 Yeah.
07:21:35 The problem we need to pick up the pen Why's this thing, right, up till 2am full. So, if the latest Docker is not available, the project team will not be able to kick off that white listing activities.
07:22:08 So maybe we need to set up a bridge or whatever to understand because I'm hearing I'm trying to translate what I'm hearing here. So from am project team, they believe they have fix all the issues issue and now we need to focus on the issue with the So,
07:22:13 but from the om team. They believe there are still an issue in ANSI. So, in order to evoke miscommunication I would suggest that maybe we have a quick call.
07:22:28 In order to see where we really stand.
07:22:38 Other comments.
07:22:41 Maybe comments in the chat, let's see.
07:22:44 I don't see any common in the chat
07:22:52 at all empty me to work so So Bill Are you okay because I hope I've completed what you were trying to express. Yeah.
07:23:04 The problem is so so we need to have an AI so brainstorming, to understand how we move for.
07:23:12 Yep, that makes sense.
07:23:14 Yeah, cuz the last that so like in the last one that actually worked because there's a lot of them just keep getting a boarded.
07:23:21 But the last one that worked. It seems like the issue. The only issue was that so couldn't connect to their Maria dB.
07:23:29 But shouldn't have anything to do with us, so we just need to figure out, you know, what's holding this up.
07:23:37 And you look at this feedback in the case of yeah yeah okay it's gonna get it.
07:23:45 Okay, so, I think you did the right thing. So, maybe as an action.
07:23:51 I'm retried to go to the page with Kenny just put a slash so Glasgow MTM to work together.
07:24:05 I'm trying to go to the wiki myself for that I have Bible freezing.
07:24:11 And thank you beautiful to give it.
07:24:18 So, if I love to do myself know we can hear you.
07:24:22 I said I love to do myself. No, we can hear you. Okay, so somebody is able to hide on the wiki page, I'm still trying to fix my laptop.
07:24:29 Okay yeah I was, yeah I was in the process of typing something into the, you got it Kenny, you okay.
07:24:35 So I think that's what we have to do AI is so in om team to work together.
07:24:42 And also I invite people to look at the game it for additional information MSB.
07:24:50 Is there somebody, somebody who was on the team.
07:24:53 I think it is be was part of the video.
07:24:56 But we don't know if we expect new doctor, we didn't receive dr Elson given.
07:25:10 Does it mean that Ms b we still need to use drilling, can we have a confirmation from the energy, as, as the team.
07:25:11 So, MSBZC Silva, as a push the an updated version yesterday. So, we are waiting for getting results.
07:25:23 Okay, I have. There's.
07:25:26 There is a note from Jim Kwan says five hours ago.
07:25:33 Sorry MSP is sped patch to om for reviewing a moment to go.
07:25:41 So I assume that's the same one that you're looking at. So long.
07:25:46 Yes, 124286.
07:25:52 Yeah, yeah.
07:25:54 We have a.
07:25:56 We had roughly 40 batches, since Monday.
07:26:01 So we try our best, but of course it is a trunk.
07:26:07 Because of that,
07:26:10 and we also had some issues with the version.
07:26:18 So, we lost some gates but yes, it's.
07:26:23 We had some issue for an enterprise user issue now is that the, that is a match conflicts. That means that we cannot test again, an NDA patch, so we I'm waiting for for bill to be Bay's dispatch improve latest master so we can retest.
07:26:43 Yeah, yeah I'm doing that now but I mean that just give it. So, yeah, yeah, yeah that's because we merge the big passion and nei a couple of days ago.
07:26:55 Yesterday, I guess, oh this morning this is yesterday guess lucky.
07:27:04 So, and again, gate is very is quite food so unfortunately will take some time to to review, or dispatches.
07:27:16 after the storm of the game and still sticking to meet him.
07:27:39 Via we we fully understand that there are a lot of things especially people have been submitted at the last minute so here we are just doing a kind of status checkpoint.
07:27:54 And we will we will go at the speed that as you can do right there. There is no point you know.
07:28:05 So, yeah, Just to to explain that first.
07:28:09 I was really story because it's 826 or more time than what we should expect.
07:28:16 And the second part is that you don't forget that. And for as far as I remember I'm for Mr.
07:28:28 Ai, or like me to be nice to have provided a guilty view on on with the new version.
07:28:46 But do we do, we don't mandate to have it merged. So that means that we can, we can walk on this review, especially by a great things of Yeshua and if, If we see the iceberg.
07:28:50 Even between and four and FC so I guess that for me, and an AI his team is clearly as delivered for him for everything they wanted, they needed to deliver sorry.
07:29:04 And, and now we will do our best to America as fast as possible.
07:29:08 But there is no.
07:29:11 Yeah, there shouldn't be any pressure on Anya, to see that they will miss a stumble of it is because they say they are right on time, as far as I understand that makes sense.
07:29:27 I think we are ugly, ugly, maybe as confused people by stating, if we don't have the Joker, we cannot kick off the pen worst thing.
07:29:40 But I do think my statement on was right away.
07:29:46 So, do your best.
07:29:47 And in the meanwhile the AI team will try to validate what they can with the former locker know we need to see if we look at the items we did here. So I think we have feedback.
07:30:06 No, we need to see if we look at the items we did here. So I think we have feedback. Mo end I right, and we have the two teams to work together.
07:30:13 balance issue right to to fix the problem he doesn't define me be.
07:30:20 We have the gate and we know in the pipeline, which we have a decision.
07:30:24 And then VNFSDK is it.
07:30:28 I don't think we would expect the container that I don't know if we have the VNFSDKOEVNLZK, I was thinking VB.
07:30:41 So I don't think and i guys he's on the call so say shoe can ask you to follow up about what we do, do we need to use the anomaly this channel.
07:30:53 Are we expecting a container for Istanbul.
07:30:56 Thank you for sure.
07:30:58 So, Kenny Can you put an action item.
07:31:08 Under VNFZK tissue will follow up with, with the NFL.
07:31:14 Alright, so, David.
07:31:28 I think for the progression evolution, we, we have a plan, or at least we know where we are, would you like to give us more insight about all the items.
07:31:29 I think related to him for. I think those are the the major points. And so I guess the,
07:31:39 I guess what I've heard is that we're saying that we continue to work on these issues between em for and RC, so we could move forward with a conditional approval of em for based on the completion of the other.
07:32:00 The remaining tasks. Although, actually there is one other thing that I wanted to ask about I know, I think it was last week.
07:32:09 I can't remember if it was so long or Christophe or someone else raised the issue of the use case documentation.
07:32:19 And there is a there is a milestone task for the requirement owners for completing preliminary documentation for their use case.
07:32:35 And
07:32:35 the feedback that I received last week was that their ad not, then there's been very little in the way of documentation updates. So I'm wondering if that has changed.
07:32:51 This week, because obviously the requirement owners have been, you know, closing the documentation task implying that the that the task has been completed.
07:33:13 Maybe that I don't know maybe that was from Andreas, or, but before we left on holiday I can't remember who made the comment.
07:33:22 But anyway, we can we can follow up in the, in the PDF.
07:33:27 I think it was more gone with him to Trump.
07:33:30 Okay.
07:33:31 All right.
07:33:33 Okay, well let's let's follow up on that in the, in the PDL meeting.
07:33:39 So, so I'd like to get some feedback from the TASC.
07:33:46 Are we okay with conditionally approving em for, or would you like to extend the schedule and give more time damn for weather.
07:33:59 But what people think.
07:34:05 David, would you mind to go back to the milestone status.
07:34:09 Just sure to go in for that.
07:34:14 Just sure to going for that. And, and also you're doing that wasn't. Was there an issue with set common portal that needs to be discussed in this context of its overall in the release but not necessarily the end for because portal has an unmaintained
07:34:28 is unmaintained.
07:34:31 So, is port so why don't you finish this and then we can cover the portal content topic.
07:34:41 David Can you go up and just a little bit to see
07:34:48 right away.
07:34:51 So,
07:35:00 Okay.
07:35:01 So do you know what type of items are not yet closed, is it related to pushing the code because I see we are still three item for multi cloud do for, so do we know what is left is the price testing.
07:35:21 So for the the the items in progress. This is what.
07:35:28 So he gets it spread out quite a bit so there's some that are preliminary documentation juror issues.
07:35:36 The om review so it's spread out quite a lot across the tasks.
07:35:43 Okay, but that leaves us being alone. Okay, so that's the that's the.
07:35:49 Those are the ones that are in progress. Here's, here are the ones that are open.
07:35:55 Okay.
07:36:02 And again, fairly spread out.
07:36:08 Now we need to go to see with you I. If we need to give the previous container or not.
07:36:16 Okay. And can you also give a view about what is left from the requirements perspective.
07:36:24 Are we to ask you just to recap,
07:36:28 in terms of that and for milestone tasks you mean, yeah, so, so we just have their two tasks. Then we have one for the CC VPN use case,
07:36:53 which is to document the integration test I think, Henry about it yesterday, see that.
07:37:05 So he said that he's closing this JIRA but it's still actually open so maybe he intended to close it and forgot to, I can follow up with him about that.
07:37:18 I don't suppose Henry is on the call today.
07:37:23 And then the other one is with CMPV to enhancements.
07:37:38 And again,
07:37:42 documenting the integration tasks and I think Christophe about this yesterday but haven't received a response.
07:37:52 Is anyone in touch with Christoph, because Mickey.
07:37:58 Emily recommended to me.
07:38:00 Okay, thank you.
07:38:02 Thank you. Okay, thank you for giving us more insight about what is left, because now we have the percentage and we we understand the area that has to be complete.
07:38:21 So, with the TLC have any additional questions to to make them judgment. If we keep going.
07:38:24 Or, if we do something so do we have an exception, with Catherine.
07:38:31 Catherine for you do that, could we look at what the global requirements look like your seem to skip that David.
07:38:43 Um, let's see, it should be, isn't it on your release charts.
07:38:49 Yeah, yeah, I should be able to yeah let me pull that up.
07:38:56 I thought it was on the same page we were just on.
07:39:23 So we still have some water in progress.
07:39:26 I'm not sure we somebody is updating this stage right yeah object so maybe you want to have a view.
07:39:37 Clicking on the requirement, or maybe you want to have a view about what projects are impacted and have committed. If we use the impact of component view.
07:39:50 right, to get on the same page is that you were showing that had the pie charts on it, because it is part of the release.
07:39:59 And sorry Catherine I'm stepping in, not as a TASC member but I've been wondering what's going on there.
07:40:05 Anybody can ask any question.
07:40:09 Ah, yeah, or the global requirements.
07:40:13 Are the global requirements.
07:40:22 That's a good question. I'm trying to recall what the one cylinder is, once the global requirement is met. Um, so, so go hang on here let me, I'll drop a link for you because I've got that set up as the
07:40:30 as the.
07:40:45 Well no, actually I just have it pointing to all the global requirements.
07:40:38 Once a global requirement has been met.
07:40:43 Because it is perpetual once it's been met, there's, it shouldn't go backwards.
07:40:50 So, Kenny what we've at least with tech com has been doing with global requirements, is that they're there were the global requirements but we create a separate ticket for each release so that you maintain those global requirements because you literally
07:41:05 can go backwards.
07:41:08 Global requirement, you know, so we create one per release, but so those are the ones we should be tracking yeah to verify that there's no no regressions okay because quite frankly in terms of security.
07:41:21 There is backward regression with every release there are packages have to be updated that's the biggest one.
07:41:27 Got it.
07:41:28 Yeah, I don't believe that the global requirements are the best practices are are tracked in the these requirement task pie charts, then how do we then.
07:41:45 But then, is the picture in a pose this question to the TASC does that give the TMC an incomplete picture of the release or not.
07:41:55 So, in terms of global requirements. If you go to the impact of component view.
07:42:03 Right. If you have an impact component you have a symbol that reflect the commitment of the team for this really.
07:42:14 But that's what one, right, understand your Vegas an issue for the one path is Park.
07:42:19 Under project status is the more.
07:42:25 And then, in fact, you does the commitment of the team for the release.
07:42:32 And if we go exactly for this page.
07:42:37 I think we have the commitment of all the theme to stay as it is with some notification or some project, as you can see in yellow.
07:42:46 So that's the basis.
07:42:47 And now it's up to the person funded the group of people or the use case owner who has submitted one day. There's a common to follow up with a different team.
07:43:01 So then, that so back to my.
07:43:06 I don't really know how to ask this question, or to phrase it properly. How do you track it as part of the decision of whether go forward with a release, whether to pass a milestone or not.
07:43:18 But it's faster.
07:43:21 It's faster for you because I believe all of this were submitted mostly by sitcom and you know sitcom need to provide the commendation to the CSC.
07:43:33 And if the, and they have also the exception process when they are the dialogue with the product team to accept or not.
07:43:43 If one team will raise a risk because we also have a sister, believe they cannot make it it's part of negotiation negotiation. In this context, between sitcom, and the project team, but if there is another use case owner for any best practice of global
07:44:04 requirement it's up to the use case owner.
07:44:06 And the person who submit to, to have this continuously, follow up with the project team, and to provide a commendation guaranteeing the global requirements.
07:44:19 Now, if it could be also if if it is not suitable. We are still. We are still learning from this new process.
07:44:27 If the people who are submitting global requirement best practice believe we need to announce the release of courses to track in a better way.
07:44:39 That's part of the lesson learned, and we could, we could take an action to improve the process contractor.
07:44:47 Yeah, that's what I was wondering is it shouldn't just be tracked the same as everything else is tracked for release, because it is, you know, the global the global requirements is that, you know, with each release you have to make sure that you're still
07:44:59 maintaining this. So wouldn't it just make sense to track those still within the page that has the nice pie charts on it, rather than someplace else because it seems to fall off.
07:45:16 Well, the problem, the problem Amy is that the tasks, I don't think makes sense for global requirements so.
07:45:22 So if we go back to, so there are times we do create tasks for SATCOM anyway creates tasks for each project for the global requirements. Right, that's what I was thinking.
07:45:33 That would be the best way to track it but when I look at the requirements so you know say 760 here. I'm trying to find the, the, the project links and I'm so go to 801.
07:45:55 It don't want is this is the dual stack I don't know what they've done that's not a sec com requirement so please just change the URL to say wreck dash 801.
07:46:05 Okay.
07:46:13 Now go down and you can see all of the blocking requirements.
07:46:18 And you see blocked by and there's five more links.
07:46:25 Right. And so those are all the tasks that have been created for the project for this particular one because I think the Java upgrade. And so there were five projects that had things to do it blocked the requirement of it.
07:46:40 This is 801 so this is specifically only for the assemble release, and it blocks the global requirement of 438.
07:46:50 But is this documented as a global requirement for 38 is the global requirement.
07:46:57 Okay.
07:47:00 801 is the release specific.
07:47:04 The really specific requirement to make sure you maintain the global requirement. And those the list of blocking, which was all of the project specific ones.
07:47:15 Those are the, those were the project tasks associated with it.
07:47:19 Okay.
07:47:23 All right.
07:47:25 That's why I'm thinking that sure 801 is that, you know, the, whatever that was I think it goes back to 463 probably, but it'll one is the project is the release specific.
07:47:38 Let me see I don't work can you scroll up to the top, whatever, eight oh, I don't remember if they don't want it as a task or an epic or what we made it but I don't think it's well now this for 438 is an epic, because for 38 is the goal.
07:47:55 Right.
07:47:58 Say, apologies that I don't remember what we created that task. 801 is a task.
07:48:05 So could we just keep tracking the global requirements as tasks for the release and then they also have sub tasks under them for each, or they have blocking tasks that are project based.
07:48:18 I'm just wondering if that's a possibility for you to do and I can work this offline because it's taking a lot of the time of this meeting.
07:48:25 Yeah, it's, it's a little.
07:48:30 It's something that I think would have to be kind of customized.
07:48:35 It's difficult to do you know as just sort of an automated query that you can copy and paste from release to release so it's something that would have to be set up.
07:48:49 You know in kind of a custom format that we could probably probably figure something out but I don't think that it can be tracked. you know, along with the other requirements.
07:49:02 Just because the, the tasks, tracked on the milestone status page for the, the, you know, the use cases and features and specifications are, for the most part applicable to global requirements.
07:49:21 So, but I agree with your point that, that I don't think that there's enough.
07:49:28 There's enough.
07:49:30 Visibility Yeah. Yeah, exactly. Yes, essentially, I mean other than the fact that I don't have that problem I know how to navigate this in the PTS do there's no, there's no visibility on the ring Yes.
07:49:42 Yeah, No, I agree with you.
07:49:45 So I'm not sure how to solve a noodle on it and see if there's a way that we could somehow reflect this in JIRA as some type of a tag that would make it easy to cut and paste per release.
07:49:58 Okay.
07:50:00 And I, because we get we're happy to do that because I think that if if I think about it, The global requirements like Kenny said they're there forever.
07:50:12 The you know it's it's a global requirement, it should be on there but somehow that you should, we should be able to trace it back to whatever the, whatever that is perfect for the release.
07:50:23 Whatever the task is for the release, and that way you could track it.
07:50:27 Right.
07:50:28 And I think then it would probably work just to do it automated and cut and paste.
07:50:33 Okay. Yeah, I think maybe using tags, might be a good approach because otherwise it's, it's hard to, you know, without just manually.
07:50:48 You know, adding each individual JIRA.
07:50:53 It's hard to produce query that will, you know, pull that out without tags.
07:50:59 But yeah let me, let me think about that a little bit. See if I can be great to have for Jakarta I'd really appreciate that.
07:51:06 Okay. All right.
07:51:11 Okay.
07:51:12 Yeah, Go ahead. Catherine.
07:51:28 No, I just wanted to know if there is any question from the TSP for the option that we have in front of us is that we we keep going, except the BTL believe is blocked for the pairwise testing. So I would say, either we provide.
07:51:32 We are pool the pool with the exception that has been listed here by David with the missing Docker, find the missing the remaining, not only the Docker.
07:51:44 So he meaning Hades management items.
07:51:48 All we we wait another week, but I tend to say let's keep moving because that's the purpose of the new release cadence, so.
07:52:10 So, shall the gear, doesn't CSE approve the milestone for with the exception that we keep moving.
07:52:07 Okay, or is there any Jesse members who believe we need to to stop and say consider it later around the question.
07:52:19 So, are we are we saying this with an actual looking at it in terms of exceptions, or just move forward.
07:52:27 But I think we cannot approve it because there are still some task, but we can approve it with exception that these tasks are completed within the next two weeks, whatever, right, that that's really what we do with conditions, rather than exceptions because
07:52:46 that we have a separate exception process so let's not confuse those things so I'd say conditional approval.
07:52:55 That's what I meant. I'm sorry, my no no no no problem. and the end the specific conditions are that's,
07:53:05 that's where I'm trying to get to the specific conditions are the specific conditions are completion of these remaining.
07:53:16 What is it, 22 tasks, plus, plus, all of the tasks that have not been closed that are associated with global requirements.
07:53:26 There's no.
07:53:28 There are tasks associated with global requirements I know the right project tasks. I know they're tied to the release but they're not tied specifically Tim for
07:53:41 the date is on the date was tied that they were supposed to be completed by and three.
07:53:48 And so, David, I really don't know how to David, I totally don't know how to track this. I don't know how to make a recommendation.
07:54:00 So, knowing. I think we need to discover these identified the PDF. Right.
07:54:16 These on the commitment not we can include it with the fact that we need to give you the global requirement as well.
07:54:30 requirements based on the exception, and so on and so on. So, I mean, they're very close my recommendation will be good to continue but we're not tracking, but outside of may tracking.
07:54:37 Sure, I'll be part of the release management, I guess.
07:54:41 OK, so the the TSP at the end for with the condition that we have a clear view of where we are about the global requirements on the 16th.
07:54:59 And then with the condition that the remaining.
07:55:09 And there are 22 remaining tasks. plus.
07:55:16 plus CENTCOM level requirements, plus global requirements. I'm sorry.
07:55:23 The 22 does not include the requirement.
07:55:28 It doesn't, it doesn't include the global requirements but these I'm sorry Catherine the difficulty that I'm having is to saying, is when we say, oh, we're going to move forward with on the condition that 22 items are closed.
07:55:46 But.
07:55:52 Right, right, no different from why we get on am three Kenny, I know but the problem that I had and what we got with him three was specific tasks that we're open because this table changes right but but these these, you these tasks are all links so you
07:56:10 can you can pull all these tasks just by accessing these links.
07:56:18 So as long as I saying the 22 open tasks, plus global requirements. We are countless 22, there's 22 for the projects and then there are two for the requirement.
07:56:36 Oh, it's okay. I can work with you to put together a list, Kenny.
07:56:43 But I don't know that we want to do it in real time.
07:56:46 Given that we're, we're getting short on time, but I don't know I'll put together a list and I'll send it to you.
07:57:05 me.
07:57:05 Okay.
07:57:05 So I tried to put the tree Bible condition. We need to have a deal status by nine next week. And then the project as a new skills I would like to take a time books because some of them have a dependency on the fly studying and then doing this thing so
07:57:23 so we should complete except the documentation task.
07:57:29 The importance by the by the end of this month I think that's what I would suggest to have a time box, because otherwise we will need to catch up a lot at the last minute.
07:57:39 If somebody has an account.
07:57:46 So we keep moving by granting.
07:57:58 I don't know if you can hear me now I'm just to says, I think, at least one of the that just can closer.
07:58:18 Okay. Yep.
07:58:24 All right,
07:58:28 somewhere we we ever understand each other so are we okay to vote, or do we need to put all the bits and bytes, on an email and make an offline case evil.
07:58:41 I'm going to take what you have in the chat window and that's what I'm putting forward as a vote.
07:58:47 Okay.
07:58:51 Thanks.
07:58:52 We will do it as well, maybe just not know but that's great feedback.
07:59:07 and Chicago to stop the bit to get ready because I think it would be nice to hear your proposition for the mark Paul formation, while the DSP is voting on the milestone so I'm trying to save some time so the no issue.
07:59:35 the presentation right I'm just trying to set up the logistic for you. Because if you give it to me Catherine that'd be great presentation pickle jump in with the technical knowledge, Catherine, Catherine Do we still have to cover the portal issue that
07:59:46 was under release or is also related.
07:59:49 And that's just that, that's what I'm My fingers are crossed that it's a really simple question.
07:59:55 Okay, portal is an ongoing project is it, or the containers, being or the portal containers. part of this release, or would you pull.
08:00:06 I mean what was any testing done, is there are there dependencies on portal for assemble.
08:00:14 But no horse that we have decided to apply as part of the and maintain project proposition which was opposed by the TSP is that we continue to use and maintain as we continue the latest version of Docker off a component that has been put in under maintain,
08:00:36 if there is a need, looking at the impact for component poker was not impacted, so it's up to the community to tell us if they need to have dialogue not.
08:00:47 We know a lot of things happening via CDs now and you you why. So maybe we can add the PCL on Monday.
08:00:56 If they really need or in the use case owner, if they need to book them. And we need to integrate the latest version of Docker all, if we don't need to.
08:01:09 Okay, and we do we need.
08:01:12 Do we need with for this city. That's also a question,
08:01:18 which is not in fact is not being done defined by any news grace.
08:01:28 Neither for these two components doctor and we're not in non define as an impact for the city.
08:01:32 So can we take an action items to confirm that with a use case owner of sending a note.
08:01:38 And then also to have these items, discuss on the bigger cool.
08:01:43 Is it fair, Amy.
08:01:46 That's a great idea. Thank you, Catherine.
08:01:48 You are welcome.
08:01:51 I'm Kenny, take your time to put the boat, and then I will give the floor to q&a, I think you know the presentation. right.
08:02:08 Okay. Hopefully you can see that.
08:02:11 Yes we can hear her.
08:02:27 I also probably need to vote this but I think, consider that I'm for moving on the boat so if it becomes a tiebreaker I can drop them tech out in
08:02:27 This presentation is, I guess the culmination of a body of work that factor and myself and some others have been looking at within the de map context, it's effectively a presentation on some thoughts and approaches we would like to take and terms of how
08:02:44 we evolve the Dema architecture forward.
08:02:48 We presented it to the architecture subcommittee think it was two weeks ago now at this stage or week and a half ago.
08:02:55 And we got pretty positive feedback there but I think it's something that we're going to need to share.
08:03:01 I guess more broadly, and more continuously in order to try and get an understanding of what we're proposing on some buy into the concepts.
08:03:08 And I think it's at least in part to call to arms to help us participate to drive this forward.
08:03:16 What I'll skim through here in this presentation is a description of what we see is the problem.
08:03:21 And then we're going to go into a couple of aspects of the solution to the problem or the problems.
08:03:28 One of which is around the introduction of a change to the deployment mechanism for the Kafka message bus restaurant, the use of operators. On the second one is related to a change in the event bridge that we use for converting messages from Kafka to
08:03:44 rest and back and forth, or proposing a specific open source component to bring into to replace some or all of the capabilities and the message right or functionality.
08:03:56 Will do. I guess a whistle stop tour of the alternatives considered on there is a conference page which gives you more detail on those.
08:04:04 And I guess the last part of it that is how we, if this is an agreed approach to how we go about evolving from what we currently have, and deprecating out that and evolving towards this.
08:04:17 So on that basis, I'll get right on into it.
08:04:22 I guess favorite took on the T the PCL role in December last year so it's been about nine months in the role.
08:04:29 It's a fairly central component in the order of architecture.
08:04:33 It's a fundamental part of the backbone of the communications infrastructure and on up.
08:04:38 And the calculations that we've done over the last number of weeks, a month to identify it's around device was more than 60 topic interactions span across more than seven projects and all that.
08:04:54 was getting impact in Europe code base, and that's only the ones that we've been able to identify so far.
08:05:00 Three team members to commanders.
08:05:04 Primarily, this is an Ericsson endeavor. Looking at the foundation metrics tourism the broad amount of community engagement and the and the Dema Congress.
08:05:15 It's probably 90 to 95% committed coming from the three main developers are.
08:05:22 It's quite an old code base, it's quite complex there's quite a lot of features in there which we don't fully understand and we don't believe that a lot of people are using.
08:05:32 When we say we don't believe it that indicates the lack of, I guess users are that code to ask about how they're using it and also the knowledge and the team in terms of being able to maintain it.
08:05:44 There's also a number of unpublished dependencies, the binaries are available in Nexus but not all of them have source code code available.
08:05:51 That's a link to a subset of those component repositories and the Maven repository, and I think you'll also notice the earlier discussion around the Java.
08:05:59 Java migration.
08:06:15 It was one of these external dependencies which needs updating the job I love and in order to be able to, to make that transition for the message Roger components so that's the reason for the request for an exception on message Roger for that transition.
08:06:17 So thinking during the adoption Java 11 as a mission.
08:06:21 And then the, the general code base for software components and licensing those around 164 violations in the, I guess governance open that we can see.
08:06:33 So that's a significant effort required to, to address all of those.
08:06:38 There's a tight coupling to zookeeper and zookeepers used to share API keys to projects as one example. And so that's something that we're looking at to see whether or not that's potentially a security issue.
08:06:51 And when we look at what own up does, we look at what message writing is message writing isn't what we would perceive as being a definitive differentiating capability or not but it's an enabler for the broader network automation capabilities that don't
08:07:05 have provides it's not in and of itself a differentiating differentiating capability on our a bunch of other open source alternatives out there.
08:07:12 So what we would prefer to do is to rely on and work with that broader community, to build, so that we can build specific features of interest, rather than investing time and effort in maintaining a messaging infrastructure that isn't receiving a lot
08:07:31 of TLC at the moment.
08:07:34 Okay, that's the problem.
08:07:36 Now let's see what the solution could be just to talk about the architectural context on the left hand side here is the sort of the system context you that we do for all of the components and we're just highlighting demand and professors that were proposing
08:07:49 to change in that context.
08:07:53 As you can see a semester drivers source and consumer interfaces so that the interface is used to send and receive messages is one aspect that we're proposing to influence which therefore has a cross cutting impact on quite a number of projects, and these
08:08:06 are only examples of the projects are not the sum total of all of the projects involved.
08:08:11 On the right hand side here I have a picture which shows a kind of a very simplistic view of the, the overall architecture. So, we have the message Roger which is the bulk of the Java code, which provides an obstruction on top of the, the, the Kafka cluster
08:08:26 which does the heavy lifting of the message sending and receiving and provides the broker capabilities and the topic management and so on and so forth.
08:08:34 The message writer supports the transparent.
08:08:39 Question topics to the producers and probably getting messages out to consumers.
08:08:44 And the message itself uses zookeeper Kafka zookeeper for cluster control and so on so forth.
08:08:50 So we're proposing to changes in this overview, the first change is related to the deployment Afghan zookeeper itself.
08:08:59 That's just simplify the way that the helm charts that we use and deploying Kafka simplifying them, and using a set of open source operators that can can make that much easier.
08:09:13 On the second aspect that is related to impacts on the message Roger interface itself and that again involves leveraging a, an open source component in this space which provides better delivery guarantees and aligns much closer to the semantics of messaging
08:09:28 that is supported by Kafka rather than the semantics of the existing message right, guys.
08:09:35 I'll pause for a couple of seconds to see if there's any questions or comments to check the people can actually hear me.
08:09:42 Yes we can hear you.
08:09:44 Good. Okay.
08:09:48 So then, so then, Karen, what is one of the implications that the projects that actually use the messaging would have to change their API's or the calls that they make.
08:10:00 So that is impact. Yeah, that is one of the implications and that's where we need a kind of a discussion on how if we should take that on how we should take that on what the implications of that would be per project, I think.
08:10:13 I think they could correct me if I'm wrong here but I think the impact is higher on the consumers that on the producers.
08:10:19 And I think we've also identified that there's a large number of it rather esoteric libraries out there in various different projects but we're not sure of the Continental, and we probably need to help them and sort of identifying what changes are required
08:10:42 there is quite a lot of just quite a lot of custom clients that different projects have felt themselves in house. And that not not using the client deliverable, which is the higher impact on consumers, which will kind of touch on later in the presentation.
08:10:54 shall move on.
08:10:56 The first aspect that is related to that bottom part of that picture on the right so talking about the, the change and deployment mechanism.
08:11:02 And what we're proposing is to use this open source stream as a project, which introduces these Kafka operators which are provided is native constructs that can be used to effectively manage the Kafka cluster as a series of Kubernetes resources, which
08:11:20 greatly simplifies the interaction with the cluster management capabilities.
08:11:40 It's the Zf sandbox project is licensed by Apache so it should meet the needs of the community from a governance perspective and and alignment with the open source needs
08:11:39 reasons why we would want to make this transition.
08:11:42 There's a lot of bells and whistles and knobs to turn on a calf the configuration, and the minutes trying to maintain that mapping and and a home file as you add more and more configuration parameters becomes quite tedious and quite time consuming.
08:11:58 The up the cut the operators themselves are using Kubernetes, negative construct so they allow you to do a declarative scale out of the bluster, they allow you to do declarative Kubernetes net of user and topic management.
08:12:14 It also brings in more or less for free, other aspects of the cabinet pocket ecosystem like the Catholic like to learn maker capabilities for more more advanced use cases.
08:12:25 And we can start to take advantage of other cool features like authentication different security models etc etc.
08:12:31 And it also means that we can sort of direct expose calculator and talk about the REST API is that coming
08:12:40 as well.
08:12:42 So how we do this, we effectively use a home chart to rubber on the cuff co operators to Kafka operators deploy Kafka clusters, using customer resources, and then we can use a cluster operator to deploy topics and users per cluster.
08:13:04 That's.
08:13:04 Yep. Good. I just wanted to ask a clarifying question in this case folks aren't familiar with operators and is it safe to say that the bulk of the code base replacement here is Kafka itself, and this, the streams the stuff is sort of the management components
08:13:20 on how you spin up and manage that Kafka cluster so it's probably, you know, relatively speaking, you know, smaller none of the code base.
08:13:29 Well we're effectively replacing our home charts, which we use to deploy Kafka zookeeper with the operators that are used to deploy and manage Capcom zookeeper.
08:13:38 Yeah, so it's more of a swapping out the deployment mechanism that we use for different deployment mechanism.
08:13:43 But Kafka itself is the same Kafka that we've been using all along so from that perspective, it's going to be the same semantics the message buses the same man the the message propagation mechanisms will be the same, it will just greatly simplify the,
08:13:57 the management of the Kafka itself.
08:14:03 And do we have any other cases where we're using operators in an app is anybody know, kind of, presented this to the, the OMT maybe for you want to say.
08:14:27 Maybe it's too long or Morgan or on the, on the call I'm not sure if there's any if there's any operators for custom custom operators or resource definitions and I think there's one CRD in om at the moment for me.
08:14:27 I'm not sure if we're using any software.
08:14:35 So today I saw an operator is roughly, as a gives a new definition in a component is crystal so that's what a track record the CRM system, which was deficient.
08:14:42 And also, also mechanism to under this new definition.
08:14:48 So we have today once you're the insider for set.
08:14:58 Self Service the scene that is delivering TMP v2 certificates, just he just.
08:15:04 It can be seen as an operator and but it's not really an operator, as we can see for stream z so screens he would get streaming D will be the first but as I traction in a incremental just world around us.
08:15:34 So I think it's kind of the way we should go.
08:15:27 Just some things that I think it's was to notice that it's important to understand that screams de, de, I think the next next part of the presentation.
08:15:39 It's not just Africa.
08:15:44 It's also a CMC bridge that we will see maybe just have to wait for the gap gap out.
08:15:49 I think it would be important to be able to connect to any character that is available, because maybe, and in your production and your manager already have a great gift guy and that's when you would like to use but if you don't have one, then Kinsey Casco
08:16:06 Viejo would be the apt to deploy your wonderful app.
08:16:09 And that's perfectly fine.
08:16:15 Thank you.
08:16:18 Okay, so that's the deployment aspects that's really talking about this lower part of this picture and I think the the bridge this jersey bridge is where we start talking about the offer letter off the picture.
08:16:30 So, that I have a follow up question about what was just mentioned about the existing Kafka cluster So will it be possible to only deploy the topics and users as part of on, for example, if I, if the end user already has deployed Kafka cluster.
08:16:53 That's a very good question. I don't have an answer. I'm not fully sure either em in terms of the two custom resources if we look at the what what the struggles at home chart offers, and I'm pretty sure they're, they're looking at their customers, as
08:17:23 I said, customers first is defined within distributed home charts so I'm not sure if they would.
08:17:21 I'm not sure they would they would they would run, or they would play within. So, that's the question I can't answer, actually, sorry, we can come back with an answer to that though it's a good question to see whether or not that's something that is capable
08:17:31 You can come back with an answer to that though it's a good question to see whether or not that's something that is capable of doing. If you just take a look, I think I've heard it's more than one, like, people asking if we can we reduce the footprint
08:17:31 by existing by using existing resources like a message bus or craft cut, specifically, so that'll be interesting to see if we can do something like that and reduce the footprint.
08:17:53 Yeah, I think for for a stretch goal I think that's a good stretch goal, to be clear though it's not the main purpose of this activity would be to effectively simplify the deployment.
08:18:05 If we can also provide opportunities to reduce the footprint and that would be great.
08:18:12 Any other questions on this.
08:18:17 Okay.
08:18:19 The second part is around upstream Zee Bridge.
08:18:23 What it is, is a rest of the best for public access which could potentially be a replacement for the direct message Roger resting.
08:18:31 Why would we want to do that. It's quite a simple code base is a community driven code base with an app developer base, which therefore means that the owner community has a reduced
08:18:43 amount of development activity required to maintain what is effectively a lower level infrastructure component.
08:18:52 And also improves over the existing message writer because it provides improved confidence of our preserves the Kafka semantics around, at least once delivery guarantees and no data loss, which is not something you get from message Roger today.
08:19:07 And it also manages connectivity security towards Caprica declarative Lee again using the custom resources.
08:19:15 So how do we do that.
08:19:16 It has its own operator. We just tell the students the operator to spin it up and it leverages an API gateway or service mesh for security.
08:19:27 So the implications on this would be, I guess initially we could run it as an alternative REST API alongside message writer, which then could lead to the eventual deprecation and removal of messages from the on up code base as is.
08:19:41 And for topic provisioning message driver today does sort of automated order creation of topics as a producer attempts to send a message on a topic the topic will be auto created by message driver This is seen as not particularly good practice from a
08:19:59 dimensional characteristics and security perspective. So, it probably makes more sense to move it into the communities operators anyway.
08:20:07 Any questions on that before we move into the discussion around how to do it, to do that in a bit more detail.
08:20:19 Okay.
08:20:21 So sorry it's terms of our alternative so there is a page in conference.
08:20:26 We put up some of the analysis we did have the different options so there's a rest proxy from Confluence Kathy Pixie calf capex is a different architecture.
08:20:34 There are pros and cons.
08:20:40 There isn't. I mean, they're all sort of good solid solutions but unbalanced we believe strings itself as a better component because of management the deployment administration for operators as well as providing the REST interface.
08:20:48 So it provides us with a one stop shop for integrating and exposing Kafka
08:20:58 that onto the evolution strategy conquer deployment and management.
08:21:04 So that's the this is the first part using strategy to deploy and manage Kafka focus have discussions with the BTL and positive feedback so far.
08:21:14 And it's a fairly reasonable approach to use a third party Helm chart for these external dependencies.
08:21:20 There is a park underway.
08:21:23 And if there's an acceptance from an architectural perspective we propose to bring it in. In the journalists are some challenges with the zookeeper dependence and message Roger.
08:21:35 But as we've identified already that's something that we need to possibly addresses as part of a security issue anyway.
08:21:43 But it may mean that in the short to medium term we need to bring in a second instance of zookeeper. There is a and Kafka Caprica project itself, there is a
08:21:55 proposal to remove zookeeper entirely from the dependencies, you will not need to keep her as part of a Kafka deployment. So, that would fall back and quarter one next year to a single zookeeper instance that's what kept the three out of course by bringing
08:22:10 in the streams at the community around this jersey will hopefully drive the evolution to support the new releases of Kafka the new releases of Java and so on as a going concern within that project rather than us managing it as well and be on a project.
08:22:27 Any questions.
08:22:48 Yeah we propose to correct, yes exactly correct. Thank you. Yeah, we're we're working off master at the moment so yeah, it would be chilly Germany's. Yeah.
08:23:04 Okay.
08:23:02 Okay. the REST API, little bit more complex.
08:23:05 We've done some analysis around what plants are there.
08:23:10 We need some input from all the details to get complete, complete picture.
08:23:15 There is a need to change the way topics are created so effectively we need templates to create topics. I don't think it's a bad thing having a having some kind of report the place to to sort of list and understand what topics are being created anyway.
08:23:32 That's pretty good.
08:23:34 Migration publishers fairly straightforward migration subscribers then requires us to understand all these libraries that are out there.
08:23:43 We have a mapping clients that's one potential brooch.
08:23:47 How do we plan this in over I guess multiple releases if we were to go forward, what we're aiming for. If this is an accepted approach is that we eventually deprecating remove the message Roger API, potentially across two releases as an example here,
08:24:03 how does this get done.
08:24:07 If it's if it's an accepted approach.
08:24:13 Just leave it on that slide for them.
08:24:15 That's the presentation.
08:24:17 That's quite a lot to take in so we wanted to iterate a few times over this but I'm quite simple questions
08:24:34 was, as concerned onlooker
08:24:39 me this makes a lot of sense.
08:24:48 Thanks guys. Thanks Kenny.
08:24:52 Karen from a security perspective this is a great, a great path forward because like you said message router really isn't fix their vulnerabilities in it so on and so forth.
08:25:02 And it's using very old versions of things but one of the questions I have, because we include on maintaining components in a release if there's still a dependency.
08:25:17 What is going to be the, what will be the path forward.
08:25:22 At some point, we're going to get hit a breaking point where the unmaintained things. We're going to have it, there will be a cascade effect.
08:25:30 If an unmaintained project is still using the old message router they won't be able to upgrade to the new because there's nobody to do the work. And if their dependencies on that.
08:25:40 How do you remove something from a release that now is causing problems.
08:25:46 That's a really good question.
08:25:47 I don't have an answer.
08:25:51 No, I don't either.
08:25:54 Go ahead. I had one. The I take on when unmaintained project is needed somebody will do something about it.
08:26:05 If I'm in pain, maintain is not needed. It will just die by itself.
08:26:10 Right. So let's see how it goes. Again, we have the question pretty stable v then maybe at one point we just need to move them to Africa. Right, which is the next stage, but I'm convinced.
08:26:23 If one company somebody or university or anybody need an MBA and then project.
08:26:30 He will do something about it. Otherwise he will be blocked.
08:26:37 So are not soon. So, here's, here's my concern Catherine I'm sorry to interrupt you. My concern is that we will maintain, we will have the new string z Kafka implementation which is really good I mean it's a it's a well maintained project I looked it
08:26:54 up well Karen was presenting. But there will be projects who want to keep the old message router and because they have some downstream dependency and they don't, they're not fixing it but they just continue to insist that that stay in the release.
08:27:07 And now we have two messaging infrastructures in place.
08:27:11 But I think we have been in the situation at one point with all those topics where we were in every state in the map need to.
08:27:30 I say even the new the map or is ellipses premium is a new the map that's part of the evolution. So it's up to the PTN, and the fly Keon is doing some research to understanding back then to see how we can maintain all we can move forward.
08:27:43 Right.
08:27:44 So, but again.
08:27:47 So, again, it's up to the PTS to have their voice, we know
08:27:53 that we have as I said as Karen mentioned, Victor has been a fair bit of analysis done on kind of active clients, and, and, and again, there's a huge amount of custom clients of custom message or clients that were developed, I guess, historically, for
08:28:10 whatever reason, and so they will all be impacted obviously.
08:28:14 But we kind of had this discussion the exact actual meeting as well in terms of on maintained or, or, you know, dead projects, if you like, how do they, you know, they're still just employed as part of gazing and butter on Monday on maintained, you know,
08:28:31 so if something breaks engaging because of an unmanned Almaden project, how does that, you know, it's a bigger question I guess really, as, as we go forward.
08:28:41 And I think in the context of the Dema project, I think that what we need and own up generally is a solid API evolution strategy so what we're proposing here is to deprecate an old API and replace it with a functionally better new API.
08:29:01 And do that deprecation over a number of releases were proposing to releases here.
08:29:07 Attorney Jerry, if it needs to be three or it needs to be one, then we can do that differently, but I, my view on it is that we should be able to put a hard stop on when a woman to this is available.
08:29:18 London is is like the first half of.
08:29:22 2023 So, so there's plenty of runway here for for any projects to make the necessary changes.
08:29:37 And I think as Amy said, the longer with a message right around I don't think there is the bandwidth and the current team to invest the time and effort into replacing or fixing all of the issues within that code base and evolving into migrating into Java
08:29:53 11 and then whatever the next version of Java is and dealing with all those issues. So, that is just a license to that briefly yeah we hit, hit a brick wall in terms of, you know, trying to upgrade the dependent dependent code.
08:30:08 And I know, I was contacted chef Katrina and others, trying to actually get this upgrade on books, we ran out of steam.
08:30:22 Firstly, so.
08:30:34 it's kind of a rock and a hard place in terms of the, you know, the trailing trailing under the, the older projects and still needed to but sooner or later.
08:30:34 We need to move the future.
08:30:37 But that's the presentation number at the bottom of the hour so I don't know.
08:30:41 Next steps I guess we will continue socializing this, and I need to talk to the chapter about going back to the committee and so on. So, any guidance that anybody wants to give us some great.
08:30:58 Thank you so much, Karen, I think it was really insightful to see that we we assess what the industry is offering, and he had just what we have on the on a platform to move in the right direction so great presentation but not only that great research,
08:31:18 right, because you have to assess a lot of things before being in front of us. So thank you to you and also chakra, and all the team was contributed to this analyzes.
08:31:32 So I think we can say, based on the vote.
08:31:38 That congratulation to the team.
08:31:39 We have met unfold with the condition that we were discussing earlier.
08:31:45 I know we are running a little bit over time but we really need to share with you some input on that line for the companies who are in our own up, team members who are interested to run for TSCC.
08:32:02 Based on the new TC to the to the deadline to submit your nomination is on the 21st of September, which is only a few days left.
08:32:15 The. They were so bottled up to that oh that we did not a great, we have still some homework to do, but at least the section related to the TASC election is no was approved by the SEC.
08:32:33 I will need some support from the TSP members of their proxy, we have two important task to be done by the 23rd.
08:32:44 I want to you and it's up that I will start to interact with you to accomplish that.
08:32:49 One of them is to prepare or 2022 budget. Right.
08:32:55 I think we, we have, we have a wiki page already set up some way so maybe Kenny, it would be good to add the link, and if there is not the wiki that's okay.
08:33:10 I was created later.
08:33:12 And then we have some input that we will not have the time to to look at now that we own to the board.
08:33:23 Again, we need to complete that.
08:33:24 So, I would like to please the task into to cook.
08:33:32 Again we have not designed to discuss how to organize each other. So I will follow up with you via email.
08:33:40 And finally, the owner committee award has been extended to the 17.
08:33:49 So, you are still ONE DAY PLUS to submit your candidates.
08:34:03 Kenny that you want to add. No, I think that's the main thing, And I did not create a page for the planning. I will do that.