keptn Community, 10 Jun 2020

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Keptn Community Meeting - June 10th, 2020

Description

Meeting notes: https://docs.google.com/document/d/1y7a6uaN8fwFJ7IRnvtxSfgz-OGFq6u7bKN6F7NDxKPg/edit#

Meetups:
June 10: Cloud Native + Kubernetes Manchester
June 17: All Things Kubernetes Washington DC
Good first issues: https://goodfirstissue.dev/ or start here
Keptn mailing list / google group: https://groups.google.com/forum/#!forum/keptn
Noteworthy Issues & PRs (Review)
Christian K.:
Travis-CI now runs Angular Unit tests: https://github.com/keptn/keptn/pull/1882
Keptn Integration tests for K3s: https://github.com/keptn/keptn/pull/1897
Install Keptn on K3s Docs: https://github.com/keptn/keptn.github.io/pull/559
Less sanity checks when re-using existing ingress: https://github.com/keptn/keptn/pull/1899
Florian:
Manage remediations via configuration-service:
https://github.com/keptn/keptn/issues/1848
Fixed configure domain command:
https://github.com/keptn/keptn/issues/1752
Unleash-service for Keptn 0.7.0:
https://github.com/keptn/keptn/issues/1816
Ermin:
Provide possibility to notify user in bridge
https://github.com/keptn/keptn/pull/1682
Notify user of new available Keptn Bridge in UI Notify user of new available keptns bridge in UI · Issue #1547 · keptn/keptn
Andreas & Thomas:
Introduce RBACs: https://github.com/keptn/keptn/pull/1894
Send action.started before scaling: https://github.com/keptn/keptn/pull/1889
Johannes:
Add ProblemURL to the Keptn problem events: https://github.com/keptn/keptn/issues/1694
Document how to reset the Keptn API Token:
https://keptn.sh/docs/0.6.0/reference/api_token/
Christian H.:

Roadmap → KEP / SPEC / Issues:
KEP 09 - Spec change of remediation action config
SPEC 31 - Implementation of spec change as proposed in KEP 09
#1846: Testing: Platform/integration test for self-healing use case
#1847: Testing: Integration test for use-case of scaling a pod based on a Prometheus alert
Work in progress:
https://github.com/keptn/keptn/projects/29
Backlog Refinement / Next Sprint:
see https://github.com/issues?q=is%3Aopen+is%3Aissue+user%3Akeptn+user%3Akeptn-contrib+label%3A%22next-sprint%22
Planning Board
https://github.com/keptn/keptn/projects/29
Attention
Next meeting, we will provide a slot for the community to share thoughts and ideas on improvements (Gabriel)
As usual, next meetings on Thursdays

Learn more: https://keptn.sh
Get started with tutorials: https://tutorials.keptn.sh

A

Hi everyone and welcome to this captain community reading today we again have a very packed agenda, so let's jump right into it before I hand over to discuss new issues in PRS I just want to highlight that we have to meetups coming first meetup is already today, so I think you should be able to see my screen. So I can highlight this here. So first meetup is today. Both of them are virtual meetups, so they are. They are happening in real life, but they are online.

A

So the first one is hosted and organized by the different companies Manchester group and the second one is next week. I think right before one day before community meeting or maybe around the same time- and this one is organized by the all things kubernetes Washington DC online group. So for minutes there is still some slots open. So if you want to join click on the link here- and you will find the invite another thing- I want to highlight again- we are listed on good first issue def or we have. We have also a link here.

A

So if you want to contribute, if you want to start with nice and probably more easy first issue with captain, you can join us here and you can contribute to captain here and one last thing. We have opened up capsule mailing list, so you can sign up here for the captain mailing list. It's just one click where you can join and you will get informed about ongoing events, new webinars and these kind of things. So if you want to join us here, you can also post questions in there.

A

But of course we still have the old and like traditional ways of interaction like slack and these kind of things. So if you want to stick with slack, it's totally fine, but if you want to join us here, we are happy to see joining the mailing.

A

So with that I think I'm, already handing over Christian I guess, yeah and I will stop yeah. Thank.

B

A

So my first two issues kind of related to well Travis CI, so we've introduced angular unit tests and I'm running them now using Travis CI. The same is true for k3 SOPs we've added integration, tests for captain quality gates on keys and they're, also running on Travis CI.

A

The result is that we broke the master branch for a couple of days and I think it's finally working now, because it seems the unit tests for noches. These are the one for bridge and are working. The ones for go have been working all the time and basically we're now doing tests on mini shift keys, Michael Cuban. It is, and mini cube.

A

I think one of the next task will be to pick a couple versions that we don't need to test on, because kind of some of these tests are kind of redundant because we're testing keys in the same version as micro, Cuban Edison, we'll see if you're interested in the actual code that has been done or in the Travis CI scripts. It has been done for that. Please take a look at these two pull requests.

A

Is there any questions on those? Two.

A

All right, the.

C

Next, maybe one question in to know how much memory are used for a keys installation so.

A

Very lightweight, as far as I can tell, but I don't have a number in yet, but.

C

It's not a problem to execute it on a local machine on.

A

A local machine keys is is very easy to run on Travis. You have like a either six or eight gigabytes of memory limits and I'm able to run quality gates. Fine, so that centers.

C

A

The full installation does not work on keys on Travis because of a memory limit, I think, but it's not easy to debug, so we only have integration tests for quality gates.

A

Okay related to that is also the next one I've added Docs for installing captain on keys. It's this pull request, which still I think needs one more review, please. If anyone of you has time. I have two things to show here, because with that pull request, we also kind of take took a look at the quick start for our current develop branch because it was full with instructions for all different cloud platforms.

A

We reduce them now to Google Cuba need Ascension as that's the one that we offer or that we have a partnership that you can apply for credit here and keys. As that, in our opinion, is the one that's the easiest to install, so that would be the instructions for keys. It's really just a curl command to download keys, specify a version that you want, specify some other parameters and you're good to go. So it's it's actually deadline, and then this line it you should have ready to go kubernetes cluster, that you can install captain on.

B

But still there is the link to the to the other cluster setups yeah.

A

The other options are now hidden here and those are, as always, in the full installation guide which we still can recommend to anyone that wants to do well, yeah to to installed on different platforms that it wants to go to a full tutorial.

C

Kristin, can you please go one step back, please yeah. So what I read before in the keys instructions is that you do not install traffic. How? How does this work so, then you don't have an ingress controller. Can you then choose ingress ingress?

C

A

If you install an ingress controller like we do with nginx ingress, it just works. Fine, the problem of traffic or keys in that matter is that keys only has one public happy that is able to route the traffic into the cluster and traffic would block for would use port 80, and if you install a second ingress that also wants to expose a port 80 yr load balancer, it will not get an IP address and it will be stuck in the pending state.

A

So if you have no deployed traffic at it here, which is also described here and the text, if you have that added the IP and the port is free for the load balancer and you can install any ingress for any load balancing tool that you want, so SEO will work. Nginx inglis will work, I guess, link ID would also work for keys, obviously not for captain but yeah. Okay,.

C

Thank you for the explanations you're.

A

Welcome any more questions regarding keys.

D

C

Regarding keys, but I really give you and review on Friday Thanks.

A

So the last one is also kind of related to ingress and stuff. So in the Installer we have like an if-else walk where we check this in English, install option for reuse or not reuse. The idea was that we do some sanity checks, if is to English Gateway, is to pilot Citadel and sidecar injector are available in East your system namespace, but that kind of cost some problems with is to version 1.5 in newer, because Easter pilot was no longer deployment and they switch to STD.

A

So my thoughts on that was to reduce the check to what is actually required for captain to work, and that would be the issue ingress gateway in the issue system, namespace, because that's what we have as a reference in our virtual services and I also took the time to take a look at nginx and for nginx. The check was already reduced to a very bare minimum of checking if the nginx ingress control is available in the Ingres nginx namespace.

A

So if those changes actually that that changes, only some printf but we've to change up here, captain should work with, is 21.5 and potentially also is T. One point: six: if you have it pre-installed and you use the ingress, install option, reuse.

B

Just a quick question, because I right before the meeting I got an email from a customer that they have a problem with exactly this part. They are having Antos deployed and anthos is a service mesh on G key and below antle's. There is Sto, and this means Easter is pre-installed and they are using the reuse option. But the installer fails, because the installer assumes that Citadel is your Citadel. Is there, but it's not available, and is there a way that we can bypass the check.

A

Well, maybe not bypass, but obviously we one could create fake deployments in the name. Space is your system that have exactly those names, it's not very clean, but I guess it would work.

A

Alternative would be to just use the set up bash script that we have here and build. A new version of the installer therefore cos.

B

D

A

Which might be something we should consider, because I think some people have run into that issue now with the reuse option, thanks.

A

Right if there are no more questions, these were my pull requests and I'll hand over to Florian.

D

Well, okay, today, I have three full requests to show you the first one once again, first through the configuration service. So let's first look for that, so we got the new requirement for the configuration service to be able to manage open remediation. Now the steps are taken during an remediation work flow, and thus we needed new endpoint to first create remediation and also to be able to retrieve and eventually reach them. If a remediation has been finished, and this has been implemented in that pull request.

D

So here we have the new endpoints again with some example: payloads the actually look quite similar to open approvals. If you recall that from last communities meeting so I'm not going to give too much details in terms of chasing payload, but in essence yep.

D

If you want, you can look at those examples and of course, if you retrieve a service by get service or as a part of a complete project, in addition to the ultimate purpose, you will now also receive the open remediations, and this will be used by the new remediation service that I'm currently working on. So this will most.

E

Likely be presented.

D

Next week, you alright are there any questions about that extension to configuration service, I.

A

Would have a question, in which case do we want to delete the remediation? Is it it's the the things you showed? It is basically the entry in the database yeah and when do we want to delete this, or do we.

D

What they needed, or is just like in Revelations in case of complete remediation workflow, because this object is only needed during the actual remediation workflow. But we will not lose data because we still have the events stored in the data store and you will be able to get the.

A

Trace for that workflow, okay, it's basically the stack that we have to work through to finally, do the Rena we're eating.

D

It because really wants to keep the payload of the materialist view response as small as possible. Okay, thank you, yep all right. The second issue I, want to show you is related to the captain configure the main come on so here's the issue that was created by Christian.

D

So basically, what happened is when he had a captain installation with certain domain, and you expose the bridge, then the bridge obviously is reachable, but if you afterwards, I want to change the domain. For example, in that case, Christian switched from sip I knew I, oh yes, so we entered that command and afterwards the bridge wasn't reachable. Under this new new I/o I mean the reason for that was quite simple was because the existing retro service or ingress has not been updated. So the post rule of that ingress of virtual.

E

Service still contains.

D

The old sitting idle at the main- and in that case the workaround is, are quite simple. So what you have to do in that case just have to first execute configure bridge with the action lockdown and the existing virtual service or ingress will be removed, and then you can execute the configure bridge.

D

Action expose command again and then the bridge will be in reachable as it was before, and hence we made it that way. In this pull request, we simply append those instructions to the output of the configure, the main command, and, of course this is not 100% perfect, because we still would prefer to have all these steps done automatically.

D

But I suggested to make this part of another issue, because that is the Bertrand was quite easy, and if you really wanted to solve this in a clean way, I proposed to move the configuration of the virtual service etc to the API service, similar.

D

Similarly to the wave we're doing the configure bridge command right now, because right now in the configure two main commands the CLI. That's all the updates of the humanities resources, and thus we will soon.

E

Have a follow-up issue.

A

Yeah, so the follow-up issue is still it's been there, it's been written, but it still has to be decided on which which action we want to take, but yeah.

D

On trips are this, so this one is also a part of the reworking of the remediation we're working on. So the requirement here was to extract the feature: total functionality that was previously done within the remediation service and expect that you dedicated unleash service that take care of future clock tolling by a by an unleash instance. So, basically, we have removed the function that is referenced here in the issue. So basically, this function just is not available anymore. The first bit, first master.

D

Basically what the function it is retrieved unleash, credentials from humanity, secret and then used unleashed API to tutorial certain features that were defined in the remediation channel, and this is now done by this new service, and this one reacts to the new action.

E

But friggin event that.

D

We're going to introduce in Captain o70 and here is the request for a new service and, of course, if you are interested, you can browse for the code at captain country, unhitch, service, alright, and this service will be released together with captain cr7.

D

B

A question but I have a very interesting side note on the unleash service, because today I heard that gate lab is using an unleash as default feature flagging mechanism and I think there might be really cool use cases around that, and it's it's great- that we have this service available. Yeah.

F

That's one thing: I want to test in the the next couple of weeks because we are using it leavened.

B

D

G

Thank you, my frien I'm, going to present a new feature in the bridge.

G

Actually, first I will present you the pull request, provide possibility to notify user in bridge and second I will present that functionality with combination of the issue, notify user of new available captain bridge in UI. So the idea that the pull request is about showing an information message in the bridge when so, actually this is just the possibility to provide an information in the bridge.

G

This can be either an error warning success or just an information type, and this functionality will be used for the feature which is still in progress to notify the user of a new available captain bridge in the UI.

G

For this we stored the latest bridge version in a version in a JSON file in our github repository and bridge, retrieve stat, when you enable that functionality and shows you the information on top of the bridge and, of course, I will present a short live demo. This is a bridge that is deployed on my demo cluster, since this functionality is still work in progress, but now you have on the right top side.

G

So in this header section, a user icon when you click on that, you see the captain version and the bridge version that you're currently working with and is installed. So you can see. I have yeah, captain version deployed that is yeah, not an official release but and the bridge version is developed. So my development version and initially the version check, is disabled, so you have to enable it.

G

Of course we provide also details about data that are stored here and collected back when you enable that version check. It will retrieve the latest version information from our github repository and provide you with information that, for example, there is now a new captain's rich in version zero six two available, and we will of course provide also a link within two instructions: how to install that bridge or captain version.

B

Yeah that looks really cool, aiming on just one one thing: when you now open the user menu yeah.

G

B

And there should be the text saying and disabled version check and yeah.

G

There is one thing that you forgot actually also in the version check is enabled. We still have to provide reference to this details. Okay, yeah and I. Don't know we can switch that we can say by disabling this feature, yeah we will not collect any for the data, but still that the reference is very important. I. Think, okay,.

B

G

A new feature is enabled so.

B

And did you see? Did you see my note on the issue that, for the documentation, I have created a pull, not a pool, look as a branch that already contains the documentation no further down. Please see the last up a little bit up. Oh.

G

B

When I go up a little bit here, please no last comment in the branch, and that shows that shows the documentation of this feature. Okay,.

G

We can open it, of course, or oh yeah. It will be available on that on the homepage. This is just granny.

B

G

B

Commit, of course,.

G

Yeah this will be available. The link might change so I have here yeah the link as it is in the in the issue provided, but yet this link might change and, of course, as I said, we will provide the link to instructions on how to download or how to install the latest version.

G

B

G

Still working progress since yeah, of course, the state of this checkbox needs to be stored on the client side as a cookie, and there is yeah still some functionality that I'm going to implement and that's why yeah I'm not presenting the pull request yet yeah.

C

But Yemen I would have been to loading issues. Can you please go back to the kevandrach, so the verb enable I? Guess it's not necessary, because the talking part already shows me, but it's enabled this.

G

C

I would call it just version. Tech like it is in the in the issue described and second point is not pretty related, but since I really did we should streamline exactly this message by enabling this feature blah blah blah. This also has to be included in the seal I, and we need this information basically, so this is still missing. So we need an issue for this.

C

Okay, but looks, looks really great.

G

Thank you very much and that's from my site for today so I'm handing over to Andreas.

C

This week, I was working on defining rudest access rules for our kept services, so here I only have a draft could request yet so.

C

But the goal here is to restrict and roles which the the services have to the necessary rights and, as you can see here also is that that we will formulate issues in old cabin contraband, all captain sandbox repositories, how to define or how to set this service account so I'm, probably in the next week on Friday, you will get issues in old cabin contraband, captain sandbox repositories how to set this access routes.

C

So this was the first issue which I'm still working on, and the second issue is action that started so before executing an action and remediation action and the service which which execute the action should send out an action started event which informs the captain Corps that the action has started.

C

So, in my case, I implemented, there is for the health service already and if the action matches the scaling action which is implemented in the home service, I send all the action started event, and but this is just inform captain core and that now the action will be executed.

C

Okay, that's from my side and next is Johannes.

B

Okay, yeah first one yeah last Friday afternoon, I felt the pressure to touch some code and I took the issue that says that there should be a problem ul in a capital problem event. This was brought up by Rob and Adam. Both were working on a remediation use case really needed to have them the problem URL so that they have a backlink back to an address that directly points to the to the problem and that yeah.

B

The proposal was to add this problem, URL also to repay loan of a problem event and I did this by updating the NGO utils and adding this property to the problem event. Data then I had to update the dynaTrace service to also take the problem URL and and take it over to a captain problem.

B

This is very matching happens between the problem event and the internal problem, and the same is true for a premier service yeah. Also, the groom user service has to do the mapping between the incoming and forwarded problem event. I can now show you the pull requests. Let me take that one, the dynaTrace service and in this screen you should now see the payload of a problem event, and here is now the problem. Ul included- and this is provides a nice use case for yeah, going back to dynaTrace and taking a look at the actual problem.

B

E

Great, it's Rob, that's great! It is such and I think just in it just throw this out there. So, in addition to the JIRA service force that we also were using this in the I thought in the notification service or wanting that for like the notification service too, because I think what people were doing was constructing that URL with their own custom logic. So that's.

B

Repair I am told me that he took the domain from order kept the tenant and then added the problem ID, and so he he he created the UL. But actually it's already Darren. Why not using the property.

B

Okay! Yes, thanks for bringing that up, Rob I think it opens cool possibilities to build, create use cases around there, and the second issue was working on was a documentation issue, because we missed the information.

A

You that moment, when somebody cuts out.

B

My my notebook blacked out and I'm back yes, okay, okay, for some reason, I, don't know why! Okay um yeah to this issue, what I did is I provided instructions how to retrieve the API token how to create a new one and also how to reset an API token and especially for creating a token I had to take care of providing instructions for Linux and Mac OS, as well as for for Windows yep.

B

All right, that's from from my side, can I now hand over to Christian to tell us about his new book this week.

F

You don't want to reveal about my actual works this week, so I prefer to present about about the captain stuff. Oh.

G

F

So can you see my screen.

D

F

Okay, so I was working on the law, specify an upstream repository for existing good projects of Captain projects. For this I created to pull requests. The first one was in the captain: go you're totes, where I first added an other function to actually doing foot requests with a configuration service, because there were only four for post function available and then I added. The update, configuration service project function to the project details and in the captain seal I.

F

Adding the update command then added the update project command with all the stuff inside tests unit tests and then I added in the configuration service to the Gibbs, get go file the origin function and for the project handler I, yeah added all this stuff, which is necessary to to update the project and let's take a look because I fixed my mini cube installation. So I have here my empty. A git repository I have created a project for check my job and now I'm trying to update the project. Hopefully it works.

F

It successfully, no, we should have here the files, nice yep, that's it, but from my side really.

B

Cool and many people are waiting for this feature.

A

B

B

Okay, then, it's again my turn well from a road roadmap perspective.

B

The the cap enhancement proposal that drives the roadmap is the implementation of the remediation service, as Florida mentioned, he's working on that and will show us the implementation next week and when it's available, then the missing pieces are testing this guy and testing it in sense of a building integration tests and also building a kind of end-to-end scenario with parameters. India.

B

All right, then, now it's time to take a look at the Kanban board. What I did this week is that I actually moved all issues that were now presented into the done column and but still we have four or three issues that are there and where I need at least a statement to do that. One I know that we are missing the final approval for the documentation right.

B

As Christian still here, no.

A

B

Okay, then I think we are fine with that one then Andres. What's the status on on this issue.

C

So actually- and this was addressed in the toko documentation, so I'm going to say it's completed- okay.

B

Then I can move it over all right and I mean this. One is still open because we can't merge the pull requests because then we have a branch when we do that. Yeah, okay and what's in the progress yeah, Andres and Thomas are still working on the air back topic, then notify user of new available captain's bridge that also that was mentioned by airman in his presentation.

B

Then the yeah I assume that you continue working on the environment screen when notifying the user is done. Yeah.

G

I'm going to finish the notify user feature and we are planning to provide an early access preview version of the bridge when that is finished, and then I'm going to finish the functionality in the environment screen. So the next one, which is which has a dependency on the switch to the public API, correct, yeah,.

B

All right, then, Florian you're working on the remediation service.

D

It's about 95 percent done it still only changement.

B

Okay create yep, then we have here the helm service that sends send action started and Rios. You showed us that one today is the true.

C

That's true, but improbably I forgot and to move it. Okay,.

B

Then I fear one of my issues: it's your providing instructions to configure remediation provider and to configure remedy order to create a remediation configuration. It's still in progress, but yeah I have already written most of that, and at this point, I will also want to mention that I'm also writing a blog post about feet. The new feature set in regard to remediation and assume that this blog post will be available by next Thursday, and this should also highlight how you can customize the use cases for your needs and the way.

B

Then you can build remediation actions all right, then, let's just quickly scroll over there signed ones yeah. The first one got me Gabriel wanted to implement this one because he also brought that issue up and he he want to give it a try.

A

We have a question.

E

Okay, yeah sorry I didn't want to interrupt your flow but yeah. So maybe this is just a Florian on the new remediation service. What will it initially support in terms of kind of backends? Is there something like ansible that you know or ServiceNow was? Is that gonna be the first implementation and then what is the roadmap for other things like that? Either you guys or others are in the community or building, because I'm interested in specifically AWS system manager down the road.

D

Yes, sir, currently we provide, or we will initially provide the unleash service and also the health service will be able to execute scaling action, but in terms of other services, maybe.

B

I I can't take this one. I can also no no well, we are. We are in touch with miss Gabriel. This person and Gabriel wants to implement an action provider for ansible for for yeah. Okay for this direction, then a partner from Brazil also came to us and asked us whether he can implement an action provider for on deck, and we had a kickoff meeting with him and he is, it will do that.

B

For us, then Abigail from CFA is is taking our election on implementing launched our clay implementation and what we have already in the in our country. Repo is a service. Now integration I mean this one needs to be updated for old at seven, but still we have already a working example. There.

E

That's awesome, yeah, that's great yeah I mean I. Will right, I, think that went something I'm gonna try to share back with both Atlassian because they have another form of automation. So maybe I can even work with Adam to you know, build the service to hook into that because they have these holder, mediating, actions through ops, genie.

E

So okay, cool yeah I mean so any yeah. It so definitely interested in the spec when that comes out, and that was kind of my question like when is o7 coming out.

B

All right, my blog post and all which also contains the spec and the spec changes will be available by next week and the ordered seven release will be available end of June nish. Okay, that's awesome! That's great! Okay, cool! Thank you! You're! Welcome! All right!

B

Then, let's go on well unless I think there happened a mistake and we assigned you that issue, even though you didn't assign yourself or didn't commit to do that one. It's the implementation of the sand. Action started event for the unleash service. Can you do that? One yeah.

C

All right, this is a small one. Actually, this is.

B

Okay, then I will move over there. Allow specify an upstream I think there is still I mean there is no still the the PR open, but the implementation is is done and was shown by Christine. Today.

B

All right, Florian are still okay doing the refactoring.

D

B

Right then, Thomas, what's your status on that one on there on the park.

C

How much is not here.

B

Okay, then I have I have one issue that should explain how to use is steering captain well. I will not not work on that one, because we have discussions going on to actually rip out a still from from the cat Nico system and I. Don't want to yeah to work that as that we will remove in the future.

B

All right, then let go over to the back dog or a thought that there is a there was a question coming in I. Think it's also time to take that one right now.

A

No no question from the chat you already address. Okay, all.

B

Right um then, let's go to the backlog and let's yeah talk about the issues that are going to be out there and ready to get implemented. The first one is the integration test for a manner of approval use case. Christian pulled that one back to the backlog because he prioritized the work on on keys. Is someone taking this one.

B

Okay, then I think it will ask Christian, because he has most experience and and doing integration, testing and so on.

B

All right, then there is a. There are two partners. One bark related to open shift through opens shift installation because it does not allow to reuse sto, but.

C

Yes, this is somehow similar to your documentation issue. If we, if we decided to not install still in the cabinet solar, then this will not be needed anymore. So maybe we should also postpone it until we have this decision. That's.

B

Good I think we talked about that one this weekend. For some reason, I I forgot to pull a out from the backlog, but let's take a look at the next one can expose bridge is broken for installation for quality gates and Gateway node port I.

C

Can quickly explain it in our config map captain domain we store a complete URL, including the port, and if you use exactly the string of the conflict map for an ingress, the ingress will fail. Can it cannot be applied, so the trick is to filter out the court.

B

D

Would like to would take that one, because I made the change where I edit the board to the content map. Then.

B

You then, you already know where to touch the code, thanks for an.

C

But but correct me if I'm wrong, but this is now or this now has to be implemented in the API, not the seed I.

D

What this part was introduced because of my change.

B

Alright, then, the next one is related to the preach. It's yeah the button that should approve or decline an approval I mean. Do you think you find some time to work on that one or re-added yeah.

G

You can assign it to me, since the switch to the public API is already ready, and there is only slight change to be done for out of sync information, and that button is combined to the out of sync right. Mm-Hm.

D

Information yeah.

B

Okay- and we have one change or one feature for the CLI, it's enforcing the user name and password when exposing the captain's bridge that, let me just go over this- is it's really it's straightforward because for the command of captain configuring bridge action expose, we also now assume that there must be a username and a password.

B

In case the password is missing, then yeah the user should be asked for the password. This is this makes sense, because then we don't store the password in the history of the CLI, but still for automation. We need to provide the possibility to add the flag to the command.

B

All right- and it's not only an CLI issue- also they also creating the secret for the bridge- should be moved from into the API, because or then we don't need the step of manually, creating the secret and last but not least since we are now focusing on air back topics, it's also a definition of time to consider which and reconsider which rights need to be set.

B

Okay, who wanted to want to take this issue.

B

All right, then, we have the MongoDB directory permissions, causing API to fail.

B

E

B

Was was brought up by by Adam and it's kind of Park, who is already a deep dive into the problem itself and unluckily. We don't have a Christian here because he knows most of the details regarding dish regarding this issue.

B

B

C

Issue already documents: what has to be checked so.

B

B

He has some one time to take that one.

B

E

B

I would leave it in the back Locker.

B

Let's leave it there and yeah, then we have one one task regarding the upgrader and this one was brought up by andreas. Maybe can you tell us a little bit about that? One sure.

C

In the upgrade that we are currently upgrading all helm and releases through hanim version 3- and yes, this is the way to go which we decided, but in order to inform the user that we are upgrading all help releases, not only kettle releases, and we have to add this into our documentation. All.

B

Right, therefore,.

C

B

Talk issue: yeah talk is kind of on my shoulder, then I'm happy to take that one.

B

All right and a couple of issues left then this one, this one is for 100% seal I related when he.

A

Was a Christian already mentioned, he's happy to take a coup.

B

Sangalo yeah, it's showing a warning when the get up stream is not set yeah. It's cool.

A

Just wants to promote his own new feature right.

B

That's true, then he could say mate by every.

A

B

In the morning, please.

A

Check when you're reviewing.

B

Alright, then we have one that is related to the bridge is the highlighting of in the heat map Emin. What do you think should believe it where it is yeah.

G

Let's do it in the backlog: I! Don't think that yeah we will tackle that one soon. But when we have finished your current topics, yeah it's it's on the timeline and it's actually not on the timeline so, but it should be implemented soon. So.

G

B

And that one we already moved on down, because it depends on the decision, how we are going to proceed with the Installer cool perfect.

B

That was it for for this week. Regarding the assignment of the new issue at the planning session and yeah I, have you one additional note to mention, as I said at the beginning? No I didn't set. It didn't say that meeting, but next time we will reserve a spot in staff meeting to give the community the stage to to bring in some concerns. Some issues and Gabriel wants to talk a little bit about his findings, his thoughts on using captain.

B

He has some questions and he wants to share it with us and therefore we will reserve about 10 to 15 minutes and for going forward. The next meetings are on Thursdays, like they were originally planned.

A

Just one last note: if anyone wants to join today, 7:00 p.m. Central European summer time there is the cloud native Manchester meetup and we're going to present captain there. So if you want to try feel free to join us, the link is on the top of this of the meeting notes for today and I. Think with this. If there are any more questions, just raise them or post them in the chat or voiced.

A

Otherwise, thanks everyone for joining for today and see you all next week and I.

B

G

Week, bye, bye, I, see Hey thank.

C