 Hau hyn, yna'r ddweud yn y bwysig. Mae'n rhan o'r deli, ac yn ddod o'r ddweud. Mae'r ddweud yn Ni'n Mwydoogol, ym Eichor Cocks. Mae'n ddweud yn Sousa. Mae'n ddweud o Stratos. Mae'n ddweud o'r ddweud o'r ddweud o'r ddweud. Yn y ddweud o'r ddweud o'r ddweud. Ac mae'n ddweud o'r ddweud o'r ddweud o'r ddweud. Yn ddweud, mae'n ddweud o Stratos. sins yma. Dyna'r ddweud o nations wedi'i carpenter. Mae'n ddweud o'r ddweud o'r ddweud o'r ddweud o'r ddweud o'r ddweud, mae'n byw o'r ddweud o'r ddweud o'r ddweud o'r ddweud. Mae'n ydych yn ei gwasanaeth sydd angen wir. Mae'n ddweud o'r ddweud o'r gyngor o'r cwerthau â'r ddweud. Mae'n tynnu yn y bydd ymhywgol yn ddweud o'r ddweud o'r ddweud o'r ddweud. Dyna'r cyfrannu cyfnodau cyfrannu i'r unidol iawn i ddwylo'r unidol. Mae'r ddwylo wedi cyfrannu eich cyfrannu Cfyrsch, mae'n ddulltyniad o'r Cyfrannu, mae'n ddulltyniad o'r Cfyrsch, mae'n ddwylo i'r ddaeth oherwydd o'r cyfrannu cwmparu o'r ddwylo. Mae'n ddwylo i'n ddwylo i'r ddwylo i'r ddwylo i'r ddwylo i'r ddwylo. Maelinewch chi, wrth gwrs ymweld hyd yn ymweld, arwad iddo gwrs, chyfnod i gyrwm yma, neu sgwp yn ddim yn perwyddo a bod yn ddim yn ddim yn digwydd. Rwy'n ddim yn gweld darthwyd. Rwy'n ddim yn ddim yn dwi gyrwm bod sydd fyddwn iawn, ond dyna fyddwn ni'n meddwl. Mae yna'r wathabill gyda wedi'u cyfrifiad, gallwch yn ymarfer y cerdd, efallai maen nhw'n gyflym ar y llai, ond gallwch yn ychydig i'r chancell, Features that you would like or even you know code itself then let's have them Okay, here we go. So we're gonna I'm gonna whistle stop back about a year ago and take you through where we've been over that year And then I'm gonna give Richard as much time as possible to give you a demo of how it is today And then I'll wrap up with where we're going next Okay, so this time last year Strattles kind of new I was here in buzzle. I did a very short lightning talk on the last day There was less people than this It looked a bit like this. It was in the Suza github repo There was some talk that's not to max mentioned about maybe up streaming it And at that time, it's only really Suza that was using it roll-on sort of January this year It'd been a proposed and acceptors an extension project. We've moved into the github repo We've got a slack channel starting to see more people use it Suza itself released a 1.0 version We decided to drop the UI suffix so you may see that still from time to time But we we affectionately just call it stratos now and then we started working on version two, which is what we'll show you today In April I'm with through this. So we were at the US summit. There was an on stage demo We started to switch to things like Travis so that we're developing more in the open so people can see CICD pipelines monthly meeting and then somewhere along the line. We managed to pick up an official logo I don't quite know that happened, but that's it there Everlong to July. This is the first to a release This is a big refactor I'll talk about on the second that was the focus of work for quite some time But we did also make a number of improvements, which Richard will show So one of those big refactors was switching from angular JS or version one to version two As I said, this was a big piece of work We also did sort of technology refresh using you know stuff You'd expect to see type script, you know RxJS and that kind of stuff But this was was a month multi month effort So for a while we kind of backed off features and we spent more time just to read in the technology platform We also switched the front end to using material design If you're familiar with that and leverage more sort of standard widget libraries and less of our own So hopefully it'll be more familiar to people. So it looks a bit more like this now and less like what I showed before More visual cues. So we did spend some time trying to improve UX as we went along so more UX Visual cues more card layouts trying to really surface some of the things you might care about The 2.1.0 release which we've just done last month One of the main things was support for SSO login So if you had seen it before you'll notice you could only log in via stratos a zone username password But obviously many people have UAA doing multi-factor or Various custom authentication mechanisms So now you can log in to stratos via SSO and that's what our friends at IBM demonstrated this morning There's been various restructuring in the code to make it easier to develop with Um, let's see performance improvements when you push it and a bunch of other features which I'll let Rich talk about I'll mention last month. We also did a community survey So we're keen to get as much input from people as possible find out how you're using it what you'd like to see some of the Key takeaways 17 responses. Maybe doesn't sound many, but for us that was quite good from 14 different Organisations lots of people are actually using it to manage multiple cloud foundries Which was good to hear back because that's something where suzer kind of asserted was important So it's good to hear that back from other people and Maybe a bit of surprise People are actually it was at least managing thousands of apps across hundreds orgs and spaces So the scale that we were seeing from our users was probably bigger than we'd we'd appreciated And we're starting to see some good uptake on v2 These slides will be out there so you can still go in and add to the survey If you want to see more detail and results the links in those A growing user base, so I've only mentioned four companies These are the people that are happy for me to stick their logo up. I know there's a lot more than this And these are also happened to be the companies and organisations that have actually contributed code to the project So we're very grateful for that and for their support And now I'm going to hand over to Richard who will show you 210 or 212 I think it is and some of the stuff we've added more recently So I'm going to take you through some of the new v2 features some of the new new v2 features and Some of the features that we've brought across and it's not going to be exhausted because I don't have time to show you everything One thing I want to note as we come into this this this front end has been configured to use sso As was kind of demonstrated this morning. I can click on login We get taken through to the well-known cloud foundry ua sso login so I will Do that now I'm sure not to typo So once you've logged in you get taken to the application wall Here you can see all of the applications in your cloud any cloud foundry that you've connected So that can be one that could be many And you can filter by organizations in space and do other lists Faced things you can see here that each application has some kind of color color status to it So again, you've been drawn into places where you might need to do something So the billing server the staging has failed. So as a user of the system you might need to sort that out What I'm going to do now is deploy an application Which is going to go perfectly So I'm going to select the organization in space that I wish to to deploy in And then I'm going to deploy from a public public it have repo Get the name right There we go finally Finally we're there. So in version one this is where basically your your deploy experience would almost end You would hit the deploy ban and then you get taken through to A stream of your app being deployed in version two We can now select the commit which is kind of important. So we're not just deploying from head And we can add some overrides So previously if you try to deploy the same manifest twice you would obviously in the second deploy Just overwrite the first deploy. So now we can do things like supply a custom name You can tell it not to start Change the stack memory, disquote it, number of instances Things to do with the route. We're going to create a random route And now we're going to hit deploy and here we should start seeing the stream Okay, it's a rather interesting because it was working 10 minutes before but as it's a live demo Obviously it's not working now. So in classic Here's one I've created earlier So this is a this is the go environment Once you finish deploy you will get taken to this screen where you can see various things about the application We've got summary screen All very self-evident we've kind of brought up the concept of instances to the summary screen so you can see if If an instance is down We've also brought out beforehand I think instances and routes were pegged on to the bottom of the screen and often they were up the scroll bar So you wouldn't actually see them. So we brought those up into tabs So they're more obvious and we've expanded on the instance at screen You can see some various stats You can scale an instance So I wasn't planning to use this one interestingly. Yeah, we've exceeded the memory limit on this one But I can scale down Which is less impressive But you can manage you can manage instances through this screen We've also brought out the routes to a top level And there's no roots in this one because I wasn't planning to demo it But if I were to create a route I could do that here and I could bind to it here I could also bind to an existing route and I could also see all of those or unbind here or delete those routes Okay, I'm going to do the old refresh Okay, it could be that my back end has gone down So I'll discover that in a bit Also things you could do on the application is look at the services The same kind of thing is route you can create those bind to bind them to your application You could unbind them We also have a variables tab which is kind of all my self-evident and the list of events Uh also what we used to show on this screen is there's a lot more to do with the services We actually had the service marketplace hidden in this screen now we put it out to the top level Um so here we can only see we can see we've got four services in the cloud foundry that's connected What I'm going to do now is connect a new cloud foundry Back to the marketplace, um we're going to refresh that list And we're going to see lots of new interesting services that we could create instances of Oh, yes, sorry. I probably yeah, probably the wrong one. Sorry. That's a good point. Um So yes, so here we can see all the new services in the cloud foundry. We've just connected Um if we want to Play some sequel we could If we wanted to see some service instances that we previously created Um this is the wall. This is the place we could do it again. We brought up to the top level We can filter we can filter by cloud Cloud foundry or by space Um I'm going to briefly go through the cloud foundry section because a lot of that is just kind of we've ported over Um here you can see some basic stats at the number of applications in your cloud foundry the all users Remember you see there's a list of recently updated apps Uh go and visit the top there because we've recently updated it Um you can see a list of organisations Uh users and their roles we do do rural management. I'm not going to go into that now Um because I don't really have time We can see the fire we should be able to see the fire Obviously something's not quite right uh feature flags build packs stacks and security groups Um so one of the big things we've working on recently is metrics Um so I'm going to demo one part of that today, which is sell metrics Uh to do so I'm going to connect Uh I'll start us metrics endpoint Which is thankfully successfully connected and we can see If we go to an application We have a new tab Where you can see uh cpu usage memory usage And disk usage overtime And we can do things like uh have a look in the past five minutes Or we can set a custom time window One of the other things we can do is we can go into the instances tab we can see We can see uh which cell your instance is on at the moment this cloud foundry has one cell So they're all going to be uh cell i t zero We can drill down into that cell We can see some more information that um in terms of current usage the memory out of the total available Disk and containers and if it's been healthy over time Um We also see some kind of similar metrics for this Uh the amount of containers remaining So you can see here kind of it It changed because we uh decreased our go in vat by one instance It's gone to 242 to 243 The amount of memory remaining disk remaining We can also see what app instances is on this cell Uh Okay so that's the end of the upstream part What I'm going to go through now Oh there's one part I missed sorry Uh we ported this feature over I think it's quite snazzy Uh you can ssh ssh into an instance or you can't Um you've got a web socket problem That your web socket is uh Yeah it's not working So I will skip over that Um what a uh so so uh this is upstream we have a downstream product Um Which is our Susan Stratos cloud um console You can see here we've customized it it's got a new um login screen This is using the old login method of just using a password rather than SQL sign-on A lot of things like the copyright there So I'm going to log into this And here we are at the app wall again You can see it's a little bit visually different We've changed the or with uh with Stratos you can customize things like the accent colours You can also add things like the um the copyright there at the bottom again And if we go to the about page We can see the lovely geeker there in the Susan logo Uh you can add things like the euler Um and one of the things also we're working on is plugins So one of the plugins we've been working on is Kubernetes integration with the Stratos console I'm going to briefly touch on this So we've gone to the Kubernetes and we can see lists of names, namespaces, pods, applications And also we can drill down into things like nodes Um and team metrics for that node Um and that's the end of the demo So I'll hand back over to Neil Awesome so the point of showing that last stuff I'll come on to the second is Is that one of the things we're working on is to make it easy for people to take Stratos And customize it for their own company's needs So I'll mention it in a second but the kind of things we should show there I'll talk about as extension points we're trying to enable So hopefully that whetted your appetite If you've not seen it before give you an idea of what's in there now I wanted to talk about a few other things before I talk about roadmap Um kind of other stuff that we've been working on that we kind of felt was important Um is around qualities of test and test automation This is stuff that you maybe you don't see Stuff as a team what we're doing but there's a kind of no visible kind of aspect to it We're trying to work hard to spend a lot of time in the last couple of months So I'm particularly working on this side of things So you know we're trying to get to releasing more frequently I'll come on to that but you know we're trying to get to a monthly release schedule at the moment Maybe we'll go to every two weeks soon but a monthly release to start with And so we want to make sure that you know we have automated tests And automated test deployment that's working reliably So we've always had you know some aspects of this but we really want to ramp up our efforts As I mentioned at the start there are many ways of deploying Stratos So you can push it to Cloud Foundry You can push it to Cloud Foundry with single sign on enabled You can push it to Cloud Foundry with backed by a MySQL database Or a Postgres database So we want to make sure all those different deployment That make tricks deployment approaches is all tested So we you know we're leveraging Travis for our CI pipeline You can if you submit a PR that will run against it And you can you see it yourself it's running the E2E tests You know we now have those that are capturing videos of the screens As those tests run so if there are problems we can diagnose where the problems are We see a lot of issues with reliability If things run a little bit slower in Travis Some of the test might start failing So we've been fixing those holes to make them more resilient to timing issues We've extended the E2E test suite So we've got roughly four or five times as many tests as we had in V1 So we're testing much more of Stratos from an end-to-end perspective And we're trying to leverage more open source tools So you'll see us use things like code, climate, Gover Report Recently we've integrated browser stacks so that we can start to run our end-to-end tests Against Firefox, Safari, more browsers and more devices basically So all this is just saying you know we are working hard to kind of make sure that We have automated systems in place to kind of make sure the quality is still there Stuff doesn't break and we can have good confidence when we release As I said going forward we do want to release monthly So we took a bit of a hit this year as I mentioned with the tour release Where that was a major refactor We're at the stage now that we want to release monthly So there was a release in September There will be a release in October this month, next week or the week after Whenever October ends And we're very community driven So we rely on our users Obviously Suza has things it would like to see in the UI But we rely on you guys to help us to prioritise The features you'd like to see in Stratos Bugs, improvements, and tiny features Please submit issues into GitHub Get us on Slack Help us understand your needs And where we see the community wanting Where there's stuff that clearly many people want We can prioritise that higher up the backlog And across releases this is maybe basic stuff But we try and blend a mix of working on things Like improving test coverage And test automation Paying down technical debt As well as bug fixes and improvements to new features So across any particular release you'll see a mix of those So if some releases may not seem to have many new features in And that will be because we're working on some of this other stuff Stuff is coming up Not a whole lot of detail here at the moment But October extensions I'll mention this in a sec Is maybe the big thing we're putting in You saw Richard demonstrate the kind of stuff that we're adding in From Suza's ability We want to make sure that other people can do the same thing So we're going to have some extensions points in there The metrics which are talked about will be in this release I think coming up in November the plan is to support GitLab as an application deployment mechanism And that's really a forerunner to kind of refactoring that a little bit So it's easier to add any source code repository in there Scalability improvements So again from the survey it was clear people are Using quite a large scale in some places Somebody was using 20,000 apps for the service they were running And they wondered whether they could use Stratos And the kind of answer there was no But there are things we can do to make Stratos Not degrade but work well when you have a large amount of scale And some of this is due to the way we use the V2 API And limits of having multiple cloud boundaries Jason's scheme of support, so our friends in Orange Have done some work and submitted a PR That we're trying to get through And get into the system So there's a lot of stuff around the open service broker That we want to track Being able to use the Jason scheme And have a UI that presents itself A form based on the scheme for that service Is something that we're going to add And then December, December we're kind of Christmas month I guess So we all do a bit of a blitz So whether those bugs minor issues we'll get those out of the way We're going to update to Angular 7 Which should be out by then Which should give us some performance improvements And we're going to look at the private source repository support So briefly on extensions which is coming this month So the current approach is You take Stratos, you fork it And you can apply your customisations In a way that means That when we make changes upstream You can just rebase those And hopefully have no conflicts That's the intent It's not currently anything around dynamic plugins We may go down the line The intent currently is to allow people to take Stratos Customise it for their organisation And these are the kind of things you can do You already go beyond what you can do in v1 We're particularly on the front end trying to do it In a way that makes it very easy for people To develop extensions so if you're familiar with Angular There's the notion of components So we want, this is how we actually do it If you have an Angular component So if you're using the Angular CLI You can do ng generate component To give you a simple component You can then simply apply an annotation Or a decorator around that component That says I want this to be a tab I want it to be an application tab Here's my label and here's the link I want you to use in the route And so then if you can cast back To Richard's demo where he went into an application You'll suddenly get a new tab up here When you click on that your component will fill that page And so we can use the same thing For adding to the side nav We can use the same thing for adding actions In the top bar and we'll be using the same approach To let you replace tiles in many of the card views So very quickly roadmap So some of these I mentioned Are coming releases but other stuff That we've kind of got on the roadmap That we'd like your help to prioritise More around application deployment So we talked about GitLab and maybe Bitbucket Other providers being able to Talk not only to public GitHub And enterprise GitHub and to be able to use your own tokens To talk to private repositories The scalability we talked about finding ways To make sure we can handle large orgs And spaces and users and apps Kind of a new aspect that's been floated Is to start to bring out The ability to manage users in the UAA And also maybe clients Our friends at cloud.gov Raise the next one which is the ability To kind of invite users into the organisation So they have this call way that As a user you can kind of Or an org manager, you can go into an org And you can click invite, enter emails Uses the UAA API to kind of send them An invite link to create the account But also in the background They can then already give that user permissions For that org so once they've logged in It's set to go Internationalisation is something we're waiting for Angular 7 so as soon as that's out We'll get that in there And then more on the service enhancement So clearly services is an area where A lot of people have a lot of interest It's what we have is a big improvement On what we had in v1 But there's things like service plans And service keys that we want to be able to manage Within the UI as well Bringing all some spaces views to the top level So with our refactor The hierarchy is kind of within the cloud foundry I think there's something we can do from a UX point of view To improve the ability to kind of navigate And browse your way around orgs and spaces And get a kind of quick view Maybe a biggie is v3 API support So we're aware there's lots going on there I think we're keen at this conference To try and hook up with those guys And make our use cases Kind of known to them We're quite different to the CLI in terms of use cases We have much more complex use cases Richard's written a document here That kind of maybe helps you understand How we use the API and some of the challenges we have there Again, particularly working with multiple cloud foundries We have some issues And so we're hoping that we can get some of this address So we can improve the Responsiveness of the UI When it makes API calls And then there's other things that have been floated like You know SNIC integration It'd be great to go to your app and get that in Kind of in-page I don't know if we're necessarily going to do that ourselves But we want to make sure that we have the right extension points That people can do those kind of integrations And we'd love to see other people in the community Starting to use the customization mechanism To kind of build those sorts of things And then feed those upstream into the core product And then lastly another one from the cloud GovGuys is being able to have more visual Representations of orgs and spaces Applications And services and how those relate to each other So Finishing up kind of we need your help So help us improve Stratos please Start it out by saying Help us kind of prioritize and use those features If you've never used it, take it for a spin If you are using it Please let us know because that's It's always great to know we've actually got users out there Your contributions could be anything We're happy to take bug fixes or bug reports Feature requests Just ideas, it'd be great to do this Or the UX for this isn't great If you've got it, can you think of a way of improving it I find it quite hard Or even code Fixes all that kind of stuff And then lastly we have a monthly call So up until now Every month we kind of do a project update We've tried to talk a bit about what we've been doing We've had some success in terms of people Tuning into that What we're going to do with our monthly release process Is we will do our release And then later on in that month We will have this call And we'll use it to go over that release What's in it, hear from our community What they would like to see So hopefully it'll be a more useful call for people So we encourage you to attend that Details of that will be on our Slack channel So that's it from me and from Richard I hope you enjoyed the talk We do have an office hours session later on today So if you have any other questions That we can't answer now Or you'd like to come and talk to us In more depth then we'll be there for half an hour Get our repos there So take a look, fork it Report issues, find us on Slack Thank you I don't know if your mind is that blown But mine is, and I've seen this multiple times These guys are amazing Hopefully you have some more questions You have some questions for them Let's see, no All right, up, okay, perfect You show some app metrics there Do you show your metrics and service instances as well? We do not yet, no So we just have our metrics At the moment we've done the sell metrics as well That's the new stuff we currently have We don't have service stuff at the moment But now that we've got the core framework in place We can start to add more metrics in We want to tie them into the Cloud Foundry Kind of model So that would be a good one to add in I can add that to the roadmap Any other questions? I was wondering if you could share some details Or elaborate on the customization How it's going to take place in terms of the skinning Is it styling or is it some approach For doing that customization The skinning? Yes I can say a little bit now Maybe if you're interested you can come along to the office hours I can go into more depth So the skinning because we switched over to material design They have quite a nice theming Mechanism that's well documented So we've adopted their theming approach So they have The ability to specify various pallets So you can go down several layers of theming So to start with you can say Here are my colors or my color scheme My primary kind of palette And that will do the core of the UI We've had to introduce a few other pallets So you can change the pallets for like the status colors For example a couple of the nav areas So to kind of do the core color stuff You just have to declare maybe two or three pallets Really you only need to do one And then you can kind of choose So if you do the other ones then we'll pick those up Otherwise you'll get defaults That's kind of the core In terms of logo then there are particular files That you can replace So you have a separate customization folder If certain resources are there They will be used in preference to the defaults Essentially so that gets you most of the Replacement of logos And the colors If you want to do things like swap out the login screen For example like we've done at SUSE Then again there's a way of using a decorator To say use this as my login screen And so you then just have to create Your own login components Obviously with Angular you can extend the default one So you can do that Say this is the one to use And then just change the HTML page basically Font customization as well Fonts Fonts is A way of doing that with material design as well Which I can't remember But it's easy to swap out the fonts Thank you, great presentation Thank you Anything else? So I'll finish up One question What kind of roles does Stratos support I mean you have this model like org roles And you have like space roles But I read there's also like a global role So when you connect your Cloud Foundry As Richard did with IBM When you come back We go out to Cloud Foundry And we make a number of calls to figure out what your permissions are So we're bound by your permissions With Cloud Foundry Against the API we can't do anything As you that you're not allowed to do But we obviously try and read the permissions Of what you're allowed to do In terms of applications obviously That will depend on your roles When you dive into a Cloud Foundry It will obviously depend on your roles So if you have Cloud Controller Read or Write You're an admin, you'll see everything So you'll see the bill packs You'll be able to manage users If you're an org manager Or a space manager You'll only be able to have permissions within the part of the system You're allowed to administer And you'll only see those Thank you Can you touch on a bit more On submitting feedback Or submitting suggestions on that So just to be back on it I'm an IBMer, we use this And our customers really like it Personally as a developer I love it so far So great work But giving more feedback And also especially on the functionality Additions on the UI itself Is it only GitHub or are you planning to have this Integrated within the platform And submit suggestions or feedback? Yeah, that's a great question So we don't have it integrated within the app yet That's something we certainly could do I think somebody, maybe our friends at Orange Had suggested Integrating some analytics as well So we can see where people are using the app Probably some issues around that But it's something that we could also look at doing again We could put the feedback in, that would be great In terms of so that people using it Would be able to say this is not great Click and fill out a box In terms of direct feedback from yourself As somebody that maybe has heard from users Then, you know, ping us on Slack Submit issues, we're happy to have You know, phone calls or virtual meetings with anybody Come along to the monthly meeting We've already set up meetings with Orange and cloud.gov guys For example, to get their input directly So if you've got stuff you want to discuss Then you'd rather do it kind of directly Then just ping us and we'll make time for that I'm telling you a one minute story About Stratos, about myself And using it So last year in Boston they didn't come And I was doing this overview Of all the extensions And then I wanted to show Stratos But the IBM integration wasn't quite ready So of course he's in the UK, Neil And I sent him a stack message Dude, I need any environment Can you send it to me so I can try it And demo it, and then he's like, okay, yeah, yeah And then very quickly he responded Of course I don't know what time it was So it was crazy, so that's testament to And the team And then of course I'm now On my Uber to The conference And I'm using my phone Which he didn't talk about So this thing works also on your phone And I'm essentially connecting To using Stratos Connecting to an environment on IBM Deploying and seeing the deployment On my phone and then getting Into my session and demoing That is Stratos, that is And that is Neil and his team So Thank you Dr Max Thank you everyone, enjoy the conference There's lunch now and then the next session Is up 2.30 And don't forget office hours if you want to come and talk to us And hear more