VMware Tanzu Community Edition Community Meetings, 20 Apr 2022

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Tanzu Community Edition Weekly Community Meeting - April 20, 2022

Description

Tanzu Community Edition Weekly Community Meeting - April 20, 2022
We meet every Wednesday at 11am PDT. We'd love for you to join us live!

This week we covered the release candidate v0.12.0-rc.2 and John McBride did a demo on additional-repos. Check out the full agenda here: https://hackmd.io/CiuO4V0AT6WL_TgA47MXBA?view#April-20-2022-Agenda

A

Hi everyone welcome to this week week's edition of the tanzu community edition weekly community meeting. If you're watching this from home, we encourage you to come, join us live. We have two different options for you to choose from. We have this one which we meet every wednesday at 11 a.m: pacific time, and then we have a monthly community meeting in which we meet every second thursday of the month at 9. 00 a.m.

A

India standard time it's a way for you to come and meet the rest of the community, meet the maintainers, bring up any questions you might have about. Tanzu community edition, provide feedback or any feature requests. Anything like that. We would love to meet you and have you come join us live if you aren't able to make it to any of the meetings. There are other ways to reach us.

A

You can find us in the tanzu community edition channel and the kubernetes slack workspace. You can find us in github discussions. You can add your request or question or suggestion to our issue list. You can also email us and you can find us on twitter at vmware tce.

A

If you do have any item you wish to discuss with the team and anyone that's on this call right now, if something's come up, please add that to the agenda down at the discussion topic section and we will get to it towards the end of the meeting.

A

If at any point you are interacting with any members of the community, the maintainers, we ask that you please read and abide by our code of conduct- and I did not say the date earlier, but today's date is april. 20Th 2022.

A

uh If you are using timely community edition, we would love to get some more details about how you are using it. So we created this pinned issue and it just asks for you to provide a few different details regarding your your usage, mainly about your company. um What's your scenarios for using tons of community edition and there's an option for you to add your logo and we'll add that to the adopters adopters file, but it's not required to participate in this.

A

It's just a way for us to get abundant better understanding of different use cases out there for using tons of community edition, as well as like helping other members of the community, understand how others are using it and in turn it helps them know better ways of using it as well. So we highly encourage you to fill this out. If you have any questions about it, please let us know and everyone who's on the call today make sure you put your name and any organization that you're representing in the attendees list here.

A

Moving on to some announcements, uh we wanted to acknowledge some some feedback that we have been receiving regarding our communications and messaging, around tonsi community edition, whether that's on our website or in our social channels, and we really really appreciate all the feedback feedback that we have been getting and we're looking forward to working towards bettering our communications and messaging.

A

And if you have any feedback, our dms are open. Mind in particular, I'm I'm very receptive to hearing what you have, because we we want to know. We want to make sure that we are building this project in the best way possible for our community and making a very comfortable space there.

A

So you can find me on twitter at microwavables. If you want a backstory about that, handle I'm happy to provide it, and uh you can find me in kubernetes at nancy and make sure you put the I in there and so yeah. Very, uh very open um would love to know your feedback on that and we hope to start making improvements soon. Regarding this any questions, while I have other folks on here, if they have any feedback, they want to share.

A

Thank you, okay, on to some announcements, so we did um have the the release candidate 2 for 0.12 it's now available.

A

We appreciate any and all earlier testing on this release and we are, you know, planning on releasing 0.12 from may 10th. So please give that a try. Please provide any feedback and we would really love to know how you know if there's any areas of improvement for this before may 10th. So thank you for that.

A

Moving on to status updates regarding this release, john, you want to take it away.

B

Yeah, I'd love to so, as nancy said, rc2 is out. Please download it. Try it install it uh yeah targeting may 10th for the actual ga of zero 12 0., so very exciting, a lot of great features for unmanaged cluster. uh This is going to be on tanzania framework version, 0, 11 4 and we've maintained that compatibility with tmc so yeah. This is going to be a really really exciting release. um I'm not going to call out too much within the actual status stuff.

B

Here uh I think, as we noted last week, the only at-risk thing is uh using andrea as the default c9. um That's that's in work with the entrance team. It's just going to be a bit behind on the 012 target, so we're hoping to get that in to uh zero 13., uh but for now uh calico will remain the default cni for unmanaged clusters, uh yeah not much else going on within there. I would love to, though, get the screen and uh fire off uh our demo here.

B

uh So this is kind of continuing on our theme from last week's demo. If anybody was in, let me pop up: uh where is my mouse whatever? uh This is kind of continuing our theme from uh last week's demo with unmanaged cluster. We're gonna talk about uh three features. This is the big cup of coffee edition bear with me, as I am very tired, so we're gonna talk about additional repos, we're going to talk about cluster statuses and these start stop life cycle hooks forward clusters.

B

So these are new features coming in 0 12 for unmanaged clusters and first we're going to talk about additional repos, and what this is is the flag that you can add for bootstrapping your unmanaged cluster to automatically install a package repository during that cluster bootstrapping right now. What happens? Is the user manage package repository automatically gets installed, and then that gives you a bunch of packages, but we wanted to provide a way for users to provide their own package repository url.

B

So that say, you were working on something and you wanted that to be automatically available or even package devs needing that to kind of get that extra step of having that package repository there. uh What's also really cool about this, is you can specify it multiple times, so you could install multiple package on bootstrapping of your unmanaged cluster.

B

So without further ado, let's do a quick little demo. Let me get over to my terminal. So let's look at what unmanaged cluster version we have. This is 0 12, 0 rc2. You can go, try this right now, but what we want to do is, uh let's kind of create, I guess a scenario for ourselves. Where say I wanted to try the alpha release of some package repositories. So what I'm going to do is I'm going to look at the projects.registry.vmware.com.

B

Tc main and surprise, surprise, we actually do have some of these alpha releases of our packet repository so down here you can see there's a 0 12 0 alpha numbers numbers automatic things. So what that is is I believe, nick made this last week, nixon pto shout out to nick, but this is a most up-to-date version of our package repository. So let's say I really wanted to try this out. I could actually automatically install this on my cluster using this here.

B

So that's the tag inside of the user package repository lots of words, lots of things. So let's actually use that in our cluster bootstrapping. uh Let's do tanzu unmanaged cluster ls just to make sure I don't have anything going on no clusters.

B

So let's bootstrap one and we are going to create we're just going to call this test cluster and I'm going to give it the additional repo flag and then slap in that repo register. Url, slash, tc, slash main and then paste that in there and then we are off to the races.

B

uh So I'm actually going to stop this just so that you know it takes a hot second, but I do want to show that right here we can see that it selected that additional package repository, so we would get all those packages automatically on here. uh Let's see that for real, but what I want to do you know kind of as continuity from last week is. I want to specify a install package flag now, if you remember from last week, we were still calling this profiles. uh Thank you, everybody for the awesome feedback.

B

This is now install package, uh so what I can do with install package is, give it the prefix to a package name, let's do fluent bit and then it will reach in and automatically install that package. Given that prefix this could be. You know a full name, the project's, not registry, blah blah blah, and we can, you know again give it a tag. This could be latest. We could give it a values.yaml file, but this suffice. This will work. We can just do fluent bit and that will automatically install fluid.

B

But first I need to get rid of that. Last cluster delete test cluster. That's going to get rid of that. Thank you very much unmanaged cluster. Let's try again, okay! So now it is off bootstrapping and we can see again here. It's selected that alpha repository here, it's automatically going to install that and then right now what it's doing is bootstrapping our kind cluster.

B

We can see that here this will take just a minute and once it's done doing that it's going to install cap controller, which obviously manages the packages manages the package repositories all that there it is doing cap controller and then, after that, we'll be able to see it actually getting and reaching into those package repositories installing those and then the actual package itself fluid bit.

B

So a lot of really nice things kind of behind the scenes happening, uh feel free to throw questions in chat. I have that open as well or just interrupt me. That is that's fine as well, so this will take just a second.

B

We should make this even faster, so demos have zero downtime, uh so their cap controllers up uh installing package repository this one is the core package repository with things like the cni and then next we will see the actual user manage package repository coming up.

B

And there it is so we can see projects our registry- and this is the v 012 alpha one that nick made last week. That is reconciling and then here we can see installing package fluent bit. It's just selecting this package via the prefix and using the latest version with no configuration file. So if we go into canines, we can look at our package repositories here. This is what I had up earlier and we can see that it selected this one.

B

If we go into actual packages, we can see a bunch of these, so these are all the alpha packages out there. So I hope that demonstrates kind of how additional repo actually can uh you can select that and install that automatically.

B

uh Let's, let's quit out here, as with all things in unmanaged cluster, we can actually do this via a configuration file. So I forget what this file is called vim mycluster.yaml.

B

This is a configuration file that I had bootstrapped earlier, but here in this field, the additional package repos. We can see that I had added uh that alpha repository and again you can specify this multiple times. So, let's you know we could do you know this three times this would be three times you know. Obviously we can't install this multiple times, but uh this could be something like uh my test registry dot.

B

You know mydomain.com and then that would automatically reach into that registry url and install that alongside this one, uh this alpha v012 one as well. So a lot of power with this for sure, uh let's jump back over to the demo they're, not the demo. The slides uh next step inside of here is status, start and stop. So this was uh some feedback we got from the community. uh Thank you so much for this awesome feedback.

B

This led us to a really fun feature here where we can actually reveal the status of a cluster running in both the kind provider and the minicube provider. The statuses you will see are running, stopped and unknown, and then that also gave us the capabilities to provide start and stop, stop hooks so say: you're running a clustering kind via unmanaged cluster.

B

You can actually provide danzo and manage cluster stop and it will stop that within the docker provider uh kind of underneath the hood. uh So better, to show than tell let's take a quick look at that uh scott had a question: how is the name decided for the package? Repo? That's a good question! That's a really good question. uh Let me get out of here: oops, not questions. uh Let's see if I can quick uh exit out. If we look up here, we can see that this was the name.

B

It had kind of created uh the it's essentially making a valid sort of domain name for the kubernetes object, so it can't have like slashes. It can't have uh you know things that would appear in your registry url. So essentially it takes that replaces the invalid characters and I think it just puts uh it just puts these dashes in so it should be almost the same as the registry url itself, but with some of those illegal characters replaced um if that makes sense, but yeah very, very close to the actual registry url.

B

So for our uh stop and start, let's do tanzania unmanaged cluster uh ls and within the ls command the list command. I guess you can actually see this new status field here and we can see that this cluster that I had with the kind provider is running now. If I look at tanzu unmanaged cluster, uh let's do help. We can see that we have these two new experimental commands start an unmanaged cluster and stop on a managed cluster.

B

So since this cluster is already running, why don't we do tanzu unmanaged cluster stop and this was called test cluster and it will reach into the kind provider. uh You know for more details underneath that into docker and actually stop that. So, if I do talk docker ps, I can see. I have nothing running that that cluster is completely stopped.

B

So if I go the opposite direction, let's do unmanaged cluster start test cluster.

B

It's attempting to start the cluster and I can go back to docker ps and I see that that is back up and running same works for minicube minicube provides those apis to us and it can start and stop in that manner as well. So yeah this was a bunch of feedback. We got from the community really really love this uh yeah really short and sweet kind of on that feature side, but uh any, I guess any any questions that was. That was it. That was the demo.

B

Steve loves it thanks steve.

A

uh Scott has a question: I'm not sure if you got to that, how is the name decided for the package, repo.

B

Yeah yeah, so that that was kind of that bit on you know the illegal characters inside of the url getting replaced with with dashes. So it should look almost the same and that's let's see if we scroll back into the logs. We can see right here where it's reaching into this, and it's almost the same as that actual registry url, just with some of those legal characters like the slashes replaced so um yeah, you should see almost the same thing.

B

C

B

Any other questions.

B

Well feel free to yeah, give it a try, uh use it and uh yeah we're always in the kubernetes slack. That's where um we're gonna, be. You know, hanging out and giving uh or seeing your feedback. So thank you guys very much.

A

Great thanks, john: um let's see it looks like there weren't wasn't anything added for discussion topics is.

D

David was raising his hand, I think, or maybe was that just waiting.

A

Did someone have their hand raised.

A

Someone clapped.

D

I think it was me clapping, yeah, okay, cool. I misread it.

A

It's okay, thanks for pointing that out.

A

Okay, uh anybody have any anything they wish to bring up before part ways.

C

uh I'm gonna come up with an unscheduled thing. um So there's uh I don't know how to say this in the correct way, but well there's some chatter about changes that are coming to tc uh both and how we uh handle this as a community project as opposed to a product, so basically work being done in the open and also about changes to the deliverables that we have.

C

So as we learn and the project grows and adapts things become different.

C

So, basically, what I wanted to say is there is a lot of work going on, which is the vaguest thing you can say, but uh uh next week I'd like to present every one of uh you know the the publicly the the approved for public consumption version of what these changes are that there's some chatter about going on already, but no one everyone's kind of confused. So I just wanted to say I know I'm sorry, but next week.

E

Yeah and if I can have one one thing too, it's like uh this is a really important project and one of our big things is just to get feedback and clarity about what people find useful and valuable and where the best user experiences lie.

E

And to I mean that was really the reason that of this, this project is to is to get people involved and and get this great feedback and and amazing, like a one dot, uh pre 1.0 product, I'm I'm astounded at how high the engagement has been and how positive it's been so like for us. This is a we like we're we're in a really good position where we can make some decision based on feedback where, instead of like just guessing like a lot of open source projects, have to do so.

E

um I think what v's really alluding to is that we're we're synthesizing the feedback. We know we've been doing this now for six months and I think that uh that kind of opens the door for us to talk about areas where we can um increase in in our effort and and and expand the things we do, and maybe some areas that are that are getting less traction or or don't seem to pan out in terms of um those use cases like we can, you know, pull back or or do different things there.

E

So it's really just like trying to focus on the things that the community is telling us unambiguously are the most important things and then to continue building this great community around it. So I I just want to double down on that, because I think this is great it's. I love the fact that we're getting the feedback- and it's super helpful.

A

Yeah, thank you for for bringing that up v and thank you for for sharing that jace.

A

um It looks like just off topic. It looks like scott had asked a question when we add an additional repo, it replaces the user manage repo. If I understand correctly and john you say yes, it will not install the default user manage repo. If you specify one, it is an all in configuration right now.

B

Yeah, that's a great point, scott! That's! That is something I missed. um So when you provide additional repo, it will not by default, give you the default one that comes with uh our our tanzania kubernetes release our tkr.

B

So if you wanted to still use that one, you could provide it in your configuration file and then as your additional one, so you were working on. You know some extra packages provide that as an additional flag um yeah. If that's a sharp ux, would definitely love to see that in an issue uh or like your kind of user flow. If that wasn't working for you.

F

Yeah, I think that at least having that mentioned clearly in the docs, because someone who is working with additional package repos that doesn't necessarily I have no problem with that ux, it just seems not intuitive necessarily, and someone may think, hey where's my prometheus package. That should be there.

G

Additional is not really a useful name for it, then, because it's not additional to the to the base one. It's alternative to their alternative to the base. One.

B

Yeah we do call that out in the docks um naming as it always seems to be very hard. So would love to yeah reconsider that if that's something we want to um yeah get a touch point on for sure.

A

Okay, anything else regarding, regarding that.

A

Okay um and going back to what v and j said, I I want to reiterate to please: please share your feedback, don't be shy, we we really want to have the community involved here, it's very important project and we want to make sure everyone is heard in and we can, you know, do and improve upon this project um the best way for the community at large. So thank you and um we look forward to seeing you next week.

A