Managing with Figaf: from a fresh install to transport, testing and monitoring

cool and yeah if you haven't seen it please what what was the reason you wanted to to attend today what was the the thing we should be covering put that into the chat um good so in the next 90 minutes we will cover uh why we will install we'll do demo of a lot of the different functionalities and yeah the chat is open so put in all your questions and i'll try to cover those cool why automate um so i wrote just a post also on this today of that whenever we make a change in a system that affects the system of record which the sap system is we somewhat need to document who made the change who approved it why did we do it that is from old PI days I started out and the first thing I was asked was create this documentation that shows you what is going on and why we're building these things like yeah that is pretty complicated and it doesn't serve any purposes so it is about merging all of that together with changes and creating a tool where it becomes easier to transport in a government way than it is to work on manually so it is about automating it's recording all the changes that happens from start to end and we want this full transparency the thing about we we want to see why the change was made on friday afternoon we want to see why daniel made this change did some some work in it because of something that's the interesting part of it And then the transparency, we want to know all of this exploration, what goes on here. And then I guess it's about simplifying because you can also build a lot of this yourself and we see different experiments and stuff like that on it. But you want something that is solid, that can be used easily. You don't want to spend 100, 200 hours on building something. and then having to maintain it over so as you can see you can get started here and running the gaff in a couple of weeks um i would say it is a lot less and hopefully by the end of today you would see that so what's the scope of today it's about the the automation it's about the transparency and then we want to simplify everything let's see how many demos it ends up with um so there is four

different three different ways you can run figure I've recommended approach at the quote so you don't see the screen nice thought I had shouted is it better now cool sorry okay so not too much I guess this was what we wanted to cover the why install and demo and we'll hide this one we want to show the automation the transparency and the simplicity so these are the the three key elements about what goes on here we have yeah three different ways to deploy it our btp and we'll just come more into this this is our recommended approach for it. You can also run it on local, on-premise, in a Docker, or it's a JAR file that you just need to run. You need to have your own database. And then the last one is FIGAF Cloud. It is our cloud installation. It's built for trial. Do not use it for productive, but to gain an understanding without spending too much time about spinning up different things, this is a good way to do it. Cool. So the reason we recommend the B2B model is that the data privacy data is inside of your B2B organization. So you have full access to it. Doesn't this one also share on the screen? Ah, so they will both reset. How does it show? All data is inside of your screen or inside of your organization. and it is the flexibility it's easy to deploy new versions you can scale it as it works and as yeah it's easy to administrate there's a tooling around it we have also created some tools that simplify this operation it costs around 300 euros per month depending on your system You can go a little down, I think, if you have smaller systems. But anyway, I would imagine most of you anyway have B2B credits that you would be able to use for this. What we need is three gigabytes of Cloud Foundry runtime, some Postgres, and this is the smallest package. That's why we can keep it in this. and then the cloud connector if you're doing PI to CPI migration. B2B, I think, is a little more expensive than if you're hosting it in Raw, Azure, or some other place, but the ease of use and not needing to get into different platforms simplifies the journey quite a bit.

So what is it that we want to do? So we want to log in, want to create some landscape, want to do some transports, We want to check our iFlow configurations. We want to do some groovy editing. We want to test, monitor, and git if we have time. Let's see. Maybe we'll get some more time. Maybe you have some good questions along the way. Put them into the chat. So that is what I wanted to cover in this session. So here we have the Figaf tool in the cloud. I've logged in as user. You can see user details, who you have here. So if you want to start this out, you will go to scav.com, sign up for a free trial, fill in the form, and then you'll get an email with saying, hey, here's login access to the system. You can log in, reset the password, and you'll get two logins. One is the license key, which you can use on an on-premise system, on P2P.

And the other one is where you can use the tool here in the FIGAF cloud for demo purposes to make sure everything works smoothly. Questions? Okay, cool. So let's get started. The first thing we want to do is we want to connect to an integration suite system. We have an integration suite system here. I'm just connecting to a trial because that is easier. And the first thing I need to do is just copy the URL here. So we'll say create a CPI agent, the system ID, local trial. It figures out its integration suite and it needs to propagate runtime. I think that is the standard right now that you need to do. The next is we need to get an API key for connectivity. and the easiest way is just to have a split screen here so we go under the services instance instances descriptions they create here and here we need to create a processing runtime we need to create it as API and we'll call it API their roles that we need here is all access policies auth group administrator business expert and integration developer there's also some under no it's not here but i think that should be enough we say create and this will then take a few seconds for it to create in the meantime because i know it we also need to create the other one so we need to create one that enables us to send messages into the integration i would probably prefix these with figaf messaging if you have your own systems and it gets these roles as default so now the api is created and we can go and create the service keys and we can view the credentials it's not there it normally is as fast as you wanted oh to just go up here and say copy we can then go and say copy we can pass and then we'll figure out all the details on it then we need to have the message flow and here we also say create service keys questions put them into the chat see this one here we got the details and we say pass so now we have connected two different ways we have connected api to fetch all content from the integration suite and we have created connected with messaging where we can send messages to the integration suite system okay next is we unfortunately still need to use the web ui from integration suite and we have two primary way we recommend one is the custom idp here figaf is working as a custom idp would enable you to connect to the system it is just enabling itself to log in and so that simplifies a lot of the process the other is we have sap passport and here we have a sap passport we can select um obviously you should create your own a passport um with your own user and give it if you're trying here on our trial you can just add this uh this user to the system so go and here on the security users create a new user it is then just we have here so we create this as a user and give it access to it um and here we need to have AI there is I think we also need if we are working with the B2B we need this that's file content read what is this this is integration so there is different B2B roles that we can also add if we had separate roles on business access so what's policies access policies we should also give it here because it is a web

user that needs all these access policies so we will use this we will say login or next this will then log in and see what do we have on the system if everything so it says here do you want to create sample landscape we say no because we are going to create that ourselves we can test here so we do not need this one anymore so we can test the configuration to see is everything running i hope it is it is and we can see it also fits mix max content if you do not want that you can go in here and disable b2b i think we just want to do that because it's not a part of the the demo and just gives us more challenges but that is basically it you need to do so now we have connected and we can then go into here our trial and we can synchronize synchronize will then download all the content from this integration suite into figaf and that means we got access to message mappings iFlows all this relevant content and can see what is going on why we're dealing with the content the way we are and depending on the system it may take a little while one of the things you can do if you are synchronizing large systems is just to like select the cpi and then select a package that way it does not so when you're not selecting this the cpi content or if you're selecting all it would fetch all the partner directory and b2b content and that is in some cases quite large data set so obviously you need to synchronize but you may not need it every time you do work on on on something like this good so that was what required to connect to a tenant um i think in the agent there is uh it's there is a little fewer extra steps uh if you want to use this custom idp for it where you need to download this file you need to upload as what it's called a service key a certificate into the b2b and then share a link but it's two extra steps uh so it's pretty simple um you should obviously ensure that the certificate of this this is still supported um and yeah yeah so yeah i guess that's it that's all you need to do

to be able to get a content here and the cool thing is then if we go into our iFlow custom orders here i don't recall here this so let's say we just want to make a small change in this so we go in here and say hey we have a header here it has we want to change it this is a bit old 26 09 so we save it as a version and then we have a new version one way we can simplify this journey is if we go here to the plugins and enable the figaf plugin and here we will just put our url for figaf if you have figaf in b2b you would obviously put that in there and then just say hey close and that means i think i maybe need to refresh but the idea is here i can go and say synchronize which will then call our integration suite system and we will get information and we then be able to fetch the content of a guff will fetch the content of the iFlow and we will be able to see okay so what happens what was the difference here so we have here two revisions of this iFlow oh that's a different one oh yeah we have scenario documentation this gives you an idea about what scenario we are working on so let's see if i get my excel open so here this is the documentation that it creates for us give us the bpmn model it give us the number the version that's live the connections the test cases we have flow descriptions configurations and resources so i guess configuration requires we have a landscape but we'll create that but the other thing is we have here on the versions we can go and compare versions in different ways so we have bpmn model viewer this is really nice if you want to see okay so what was the difference between these two bpmn models and we can see hey there it was this value that was changed um this obviously gives us a lot of uh makes it a lot easier to see larger changes in in content so we also have the option to do it in div mode where we can see what are the content what are the BPMN changes. But as you can see here, the changes in a BPMN model becomes sometimes a bit random. And that means it is difficult to check it out in this format. Yes, we'll try to get the video published afterwards, but I guess it is going to miss the first bit of it. Unfortunately, you will just see me talking, which is always nice. Yes, cool. Okay, so that's a little on this. So the other thing we can see here is we have the comparison. We can see what are the scripts, what are the different artifacts that's a part of this. So we have the where used scenarios. If you have a partner directory content, stuff like that, you may be able to see it inside of here and see, okay so what is actually happening across this landscape here good cool so the idea is we will

go so right now we have connected a system we want to create a landscape and we select composite that's the default way to do it and we we have two versions of transport the one that's called latest version and latest version is the way where you go and do patches into your test system so you create a lot of patches into test once you have tested all of that you go live and say create a new transport based on all those fixes and move it into production that has that was our well original way of dealing with this because that was how it made most sense because you don't want to have move 500 deltas into your productive systems if you only need the 300 that is actually live you that was the reason say yes we are ready to do a live go live which select the 10i flows that's a part of it transport those into production and move on good but yeah and then we added the other one which is locked version which is a new one which works precisely like as a piece transport system where you go and take a transport from dev to queue and then once that's been imported you can then go and take the transport from dev to production and it handles the setup so let's do that so we have here uh two systems we will add the uh the two trial systems as as default and then we have an other system we've connected we will use that as our prod system don't mind the name so we would Then use dev, QA, prod as the description for this. This means we can go and transport all the content between the landscapes in this way. Great. I guess we have a 24-hour login time. Good. Yes. So we get to do it again. Trial, trial, which is our dev system. here and demo test and you will call dev qa i don't like to call this one a test because that means we will get artifacts that is called prefixed with this and test just makes it difficult to see

what goes on we have selected latest version and we say next so now we can configure the transition from dev to test and from test QA to production. We can go and say, what are the tickets? So whenever we do a change, we need to have a ticket. We call it as default issue. And the other one is transport. So you call it QTransport. And then we have the other one that's transport. we can go and say we want to have the FIGAF technical recommendation do we want to deploy this after the transport do we want to have naming validation do we have approvers set on this and we can set up multiple approvers in this to do different things and we can also specify who can approve the transport I really enjoy that I have approvers in my dev system that would go and say, hey, from dev to QA, that will check, do they follow the guidelines? And especially if you have a lot of externals, this is the place where you'll give them a review of what they've created, ensure that everything is correct, and then deliver it. And there's also other things here about B2B stuff, a lot of things on B2B. We have here name mapping. So how do we translate names? And in this case, we are adding prefixes to a lot of them. We can go in and say, hey, we do not want to use postfixes or add postfix, stuff like that. So here, postfix, stuff like that. So there's different ways we can go and identify this. And it needs to be something fairly simple because the idea is you go in and create access policies around these things also. We have parameter replacement rules. We'll get into this later. Something about some tagging rules, other things about email notifications, webhooks. So there's a lot of things inside of this. and here for our production transport we do not have any approvals so so you can go in for each different transition to set up the scenarios we'll create create it and it's a save save and now we have created a landscape way and so that's good so that means we can go in on our landscape here we can see the normal

transition path and we can see okay yes we have these collections and that's we can see here that should be a lot more in our product system but we can see that there's nothing here yet and that means let's just take this content here do a transport so we want to transport it into our qa system we hear there's a title here what is the name of this we can add in a zero or whatever task definition tool you have here it's just a link you add you can add the link to it and you can add there about what you're doing why you're doing this and we then request licenses for iFlows when we do transports on them and we can see the content here and we can see this transport we can see is do we have any test cases we haven't created them yet let's do that later we can see if there's alerts there's transport reports that we need to do but let's start the transport and now we have created the the transport we can see it's a package there's a script collection i don't think that there's a link to the script collection so if you want to see you can go into the script collection and you can see okay is this used in some iFlow in this case it's not so that's not that important for for this setup to go we can off the bat here see the design guidelines checks what is going on here and we can see they should probably not be promoted or we need to go in and do some checks and say okay yeah let's um in here the design guidelines go and say analysis uh this is single payload script use none uh do i need to be in edit mode to go and check the design guidelines analysis ah so here i can go and say yes and grab them but up these uh use of non-recommended libraries we obviously need to unsupported java file in this script here so we yeah it pops up and says hey fix this it's not even used so is it no okay anyway we'll say close here and let's just see if it's able to pick up this design guidelines here we we ignored so check the design guidelines so here it can see hey, we have skipped like this, that it's not a problem. It is not a blocking issue. We do not see it should be that because these are sometimes not really required. I think in general, you should probably have gone through all of these, but I think it is more the fact that you go in and check, is this okay to skip this requirement? Because, yeah, how does that follow your structures? Oh, there's a question. Yeah, we got that. Yes, I'm sorry. It is, yeah, hopefully you'll get the recording. Good. Okay. So this is the preparation.

As a developer, I can go in and check these contents. So I can go in here and see what are the externalized properties that we have for this. and we can see that there should be an odata password we probably need to fill that in and we need system translation so we go in here and this is one of the cool things so we can go in here in our landscape definition say hit substring if we see hana-dev replace it with hana-qa so that means that every time we see hana-dev it's been replaced and we can do the same thing over here in the QA to production detail I used the latest version if you had locked version it would be dev to production so that means here we need to have it QA to prod and this is one of the things that really nice with SAP transport names because there is a lot of similar names across the landscape that you need to remember every time you do a transport and that means if i go here and we read it it can say hey in this system it is going to be called hana qa that's nice there somehow needs to be a old data password so one of the places we can go in here is in the figaf tool we have security artifact we can create a new security artifact, actually I don't think it's used for anything, but that means we can add it also here in the externalized file, password this is the externalized number, so we'll add it

save as a version, now we do not need to save it as a version let's just save as a version, so we now we need to make a change here, so we have a new version of it We will inform Figaf that there's a new version. So Figaf needs to get all of these changes, and we do not want to pull data, so it's more on a user's flow to create all of these changes. But the thing is now in our landscape, in our ticket, so we can see our ticket here, and the ticket is in production or is in transport, but we cannot make modifications to it in the objects that we have so to do that we go and switch it back to development we can select the objects here say update to latest version and we now see that there is two versions of this guest customer and customer configuration and this enables you to have multiple different versions in the system so there's a raised hand Do you want to talk or just put in a chat? Oh, there's a question here. So could you lock the CPI production via FIGAF? So FIGAF does not enable you to lock down the system, But obviously, the idea is you give users limited access to production because most of the transactions that they need to do can be done as a part of the preparation. So that means no one actually needs to go in to the system and do any things in it. I hope that answers. Otherwise, put more context on it.

So we now have one that's canceled and we have one that's created here. we can see it is revision 2 that we're trying to do and we can see it also hopefully still had the hana there we can see it has the password that's being transposed so as a developer and you have this clone with latest version that goes and checks that if there is a newer version of the iview but here we'll go and send it to approval obviously you should not approve your own transport, but there is a setting here that says transport approver cannot transport his own request. And obviously you need that into the system. Okay. Here. So here we have transport. We say, if you have multiple review on the text time, we have the start review. So we can see, hey, Daniel already started this. So you don't spend time reviewing other people's stuff. Done. We say approve. And if we said, yeah, reject, it goes back to the developer to fix it. We can say, hey, now we're ready to import. You can also schedule the import and say, hey, we want to run it at a certain time. Questions, comments? Raised hands? I've got plenty of time. But I do hope you see there's a lot of functionality and we'll see how much we get into. Maybe we can even get into some of the AI cool stuff we have added around this setup. Good. And the reason it does take a little while is because we are creating snapshots of this Eiffel before and after it's created. So we do have an option to do rollbacks, which is really nice. for a lot of scenarios um if you want to do that and we can see here the iFlow has been added where did it go with with a qa as a prefix and that means it is pretty easy to go in into the tool and set up um rules saying hey qa cannot be be accessed i don't want to do it because then we need to just check also the authentication and login users for it. Cool. And if something fails, we have some traces here that gives you some more details about what goes on in the system.

Okay. So that's kind of the idea about how we create transport, how we can manage it. And that means if we go here to our landscape overview, we then have the option here to say, hey, we have this, but it is in deploying state. So one of the things we do have here, if we go and do a synchronization on the system, we need to initialize all these externalized properties of it. This will check that. So here, if you can see here, it says here that there is a mock that has a different version than it's assigned, and that is because we have deployed it in different formats so hopefully this will go away and we can see it works smoothly if we wanted to we can even deploy from here so just say hey deploy and we would probably need that because we want to create some test cases also on this if it works but it's really nice need to have the option here to create deployment of multiple different night flows in one go as we'll see here so it should be status starting so we'll just do the synchronize again and yeah if things fail so this is synchronized forcefully that takes all the content that you have and read so whenever we create changes in models stuff like that it would need to update a lot and here we can see it's starting and hopefully this one would also go away once it's been deployed correctly and stuff like that oh so obviously it's a little problem i should probably have deployed some iFlows because now it would we need to be lucky to see if these would actually work uh efficiently so there's outlook is added that's some different content that's mcps uh that's on integrations so what uh i would have loved to see that and also have the cloud integration up here as an option to choose from um yeah maybe i don't know um why it's missing okay one other thing we need is to create a system we need to go into our discover content here and here there's a figaf figaf connector we have some uh jira service now integration stuff like that that where we have already built a lot of all the integrations that you need but we need this uh connector for our integration suite so now we have copied it and there is some iFlows here we would need uh for for some of the testing so we have a jms and a process direct to to deploy so we can put in here deploy why did it fail and that is the big question because here i do not have this okay so let's see here so hopefully this one gives us more details so it should find a new package and see that it works okay did i just do it for synchronize synchronize all packages let's do that okay let's see maybe there's something on the integration suite system that we did not anticipate um we'll try to do something else then so now we can see that these are at least they are starting and so that's good but this is really a good overview to see what you have and to see is that the same version you have in one system versus the other so we can see there's a difference here we can go in and compare these to with our model viewer and we can see that there's some some differences we added a design block uh error handling stuff like that on it removed it um there's a clear filter up here to remove all the filters you have uh when you're dealing with this okay let's see um so one of the nice places to do this is we have this testing tool here so we can go and create a test case that is customer i could spell it would probably be okay and we can then select one of the iflip we want to use for this. Here we select this. We can add it to your collections, and we can add some xml. I think it is just called... Let me see if I can... No, it's not this one. Oh, I think it's called... Come on. Why doesn't that work? Like that. Okay.

and then we have a route should obviously not be closed here okay so here we have a payload we can say run let's see if if it all works and it says here that this eye flow was not found um in this so we can just check the trace but that is so we have the here the option to say save and we'll just run again see still give us this error on this endpoint we cannot send messages through it so there could be two reasons for this one is this service key we used here on the iFlow map host is incorrect or there is actually some problem with this integration cell here and the reason why so enabled the integration cell a week ago or something like that um so that you have here under it you have run times and it says that this is the default one you cannot disable it um yeah i guess that is what so what's a hair so so this is just called cloud integration but it is yeah i would have loved to see it here um no worries um we will do our test in production instead. Yeah. No, sorry. Good. So we have here our iFlow. So let's check our landscape here. We have our landscape. It moves it into this trial tenant. And here we also just want to say it should deploy after transport. Submit it. and that means we can go and save it. If I go here to my tickets, so I can see my ticket and I can go in here and say, hey, if I look at the landscape overview, what are all the tickets that's involved in this? And we can say, hey, there is these, so we can go and create a transport here, a new transport on this, or we could go in from the transport we already had. So we have here the transport. So we have it here from ticket. So on the ticket, we also have here create next ticket. That will take us first transport to prod. That would enable us to move it into production. So if you're using this latest version, you are creating two tickets, one from dev, and one from production here we do not have um validations on so you can go and just import it but obviously that is what you configure in your setup questions comments so let's see um so we can see here what it does that it does trigger deployment hopefully trigger deployment of this 255 yeah something isn't that easy you cannot run the iso in your dev system um okay um yes so that means if we look at our overview now we can see that these are starting and obviously packages are not deployed good okay so they're starting so let's try here to send the messages i'll just check here on the agent setup so on the agent setup we do have here that we can select it as production system we don't want this in this case because it's still just a test system but obviously if you have selected a system as test you cannot do testing on or as production you cannot do testing on it so let's see here if we have this we can go and edit and we can then say no it's not like that can change it we can then change it this to this and here we can see there's a lot more content into it uh what is the name here faster this one so we can say here run this is really a nice tool if it worked gosh cannot read see if us token did we remove that is that just no it was different we got a different error in dev as i recall it that was that the ifl was not deployed uh so here's the dev here we just get that it is yeah it says that this is not deployed correctly so we can obviously so the thing is you can then go here in the landscape see your production system you can easily navigate to it get customer orders here for other test cases we do have the support of a cff token i'm not sure why we do not have it here in the send message so we'll just let's not even see if token do that is strange oh well that's what you get when you join a live session so what is going on here let's try this one same idea maybe that is then also a configuration issue so let's try to check it i'll just go into the b2b cockpit any questions along the way um then test so some subscriptions demo tests as it be time profile so i'm right now just creating a new runtime profile for this view credential called the json so if we go back here to our gaf system so here on the agent maybe that was this one that was in row so we can go in here pass our json save it it does not test this because that is just impossible so send messages we have our request here edit run that will still not work because this is connecting on the wrong one here we would need to select the run yay it worked okay yeah it's a pleasure of doing our live demos so now we can send a message in and this is like the pi send miss this is really helpful once you're developing i want to check what does

my iFlow actually do does it work as we would expect it to do or is there something else in it so let's just try with the other one does that also work so that one does not work maybe there's something wrong with the tests or the the iFlow so it doesn't deploy okay all right we will use this number two all right so we'll press save we can then run for recording which run from for recording means we would switch the inflow into trace mode so this is if we want to create a more persistent test case for message we can use it in this format and then we can say hey we have a message we can see it in the system and we can go and say yes demo or more relevant what is the business object that we have for this so this is customer orders it knows this is the message id that we want and that way it will go and fetch all the details from this um this is just really a convenient way to create tests on and ensure you have all the details correctly um yes cool questions um so right now we have one incoming message with 32 outgoing messages and we say hey create a test case on this we have two options

regression testing which is figaf is responsible for sending message and baseline which is the original system is responsible for sending messages so this is really good if you can send have a tricenter system that can generate sap orders you can then use the baseline setup for it and you can also set up landscapes if you have created this in your dev system that means we now have a test case and we can see here the content of the test case which is in this iFlow we send in this message like this we have all the different outgoing steps along the way and we can actually run it and the idea is we run it we switch it into trace mode send messages through and then we can actually see what is going on and what happens in this case so it will send the message in then we'll download all the different messages that exists from this um and we can see the how the process works here it's obviously important to see uh is it able to find all the relevant messages in a process and you would see sometimes here in the results that you get something that is unexpected so say you are creating a calling create employee you call that and it will say hey this employee already creates because you just created it and then it will fail give you a different path that is one of the things you say yes that we will need to create it run it in a different way you need to mock it it is pretty easy um hopefully we can get to that one but now we can see that there's some diff in what we expected and we see that there's a updated here that is added and let me just simplify this just

just like this i think that would work and then we can see there's some messages timestamps here that's changed it's okay to ignore those and once we have done that we will go in and process a comparison again or we can so we do not need to run the test again and we can see now everything is successful so it's pretty easy to figure out what is the differences is between content and That means we just go in here and view the result to see what is the message we receive. Or view the diff and we can see, okay, what is actually the difference between these two messages in this scenario. If we had the case where, let's see, we had the case where we wanted to mock an endpoint, we just say here, use mock endpoints. and when we run then the test case here it would mock all the endpoints so right now we could see here in this and this takes a little longer because it needs to deploy this but here we can see it runs on this iFlow which is our original one that we have here in this target system looks much familiar to to the other one and then the new test case here has been then been run on a different system and we can see it runs here on figaf mock system and here we can see it replaces this one with information about what it is step number 50 mocked iflou testing and then all these have then been replaced with the figaf targeted message which means we really get to see what is going on in this this landscape here But there are some differences here now, and this is because there is different headers that is being added when we deal with this. So let's see here if it finds some less. Oh, that's not. I wanted to do a comparison. Click, right. So we want to do the process of comparison or full so we can see some errors disappear. we have here filter and errors we can check here there's probably some other diff that we added

this very option and we just do the comparison and now we can see everything has been removed and we have a successful test case even the data is not touching the original system but just going directly with vgov that calculate what is being sent to this um yeah if i had transports from this landscape i could also in my ticket add those to the scenario um and one of the things i also wanted to do was let's say we have this iFlow here we have an option here i think this is one of the really nice features um we have raw enable ide here um and that means if we go back to our iFlow here in the figure of tool in demo test we can see here the the different scripts in our main iFlow so we can see here so we have a script here we can open the ide for it and we can see is there so this is we have not added the ai so that makes sense and here that's one of the things we are not calling get customer groovy set hitters 2 i think this one is used we can see do we have test cases for this no set hitters 2 demo groovy let's see is this better so we can say here what are the messages in an output message for this and we can then run it and here we'll create a deployer that enables us to run this iFlow on the the same tenant and get a result from this and here we're getting something about a string length text because on line 15 okay the the other approach is that you go in here in the in the original one so we go in here select the script here set have us to and open in fiaf open here so now we should we have the test case it finds it it finds the payload for it and now we can actually run it and this runs on the same integration suite system which means we can even use a partner directory content lookups

which is just so much easier to build with this if we had added open ai we could also get help on writing and improving this uh this groovy script um questions comments what do you think is most relevant to see is that adding the ai help the monitoring groovy oh sorry git integration um i think those are the three next things we needed to cover testing monitoring maybe we need to take monitoring and then we'll take the git going forward okay so we have here a number of different ways to monitor the system so we have here and a consumer where we can go in and say okay what is the messages that's been processed uh let's add it on the demo test because at least there are some more we have here the fetches and that means we can go in here and see do we have whole alerts poll alerts will then fetch all the messages that's been downloaded or error and we can see there was this error because we ran it in the integration read with something about 15 whatever that was the idea is you would then either you asked if you got but for help about what was going on here all you're going then you can create a rule saying hey uh script uh they'll be ordered in sequence or you can put it in different flows depending on what you want to see we can also mute it so we'll not get notifications every time it happens we'll send a mate immediately we'll add attachments should we send it as an email or who should receive this email So there's different ways you can define this or webhooks. And that means we would build up a list of rules. And that means next time we get this error. So let's see here if we actually run this again. We should get the same error, hopefully. So it now processes the same error. And that means here if I pull messages, it sees that there was two messages. And it can then see, okay, so this was actually this rule we just created. the script processing that created this message which simplifies our processing because then i do not as a developer i do not need or a supporter i do not need to read anything i can just see oh this is the description of how to resolve this and then we're good instead of having to research and do some different things find it in different places they would be marked with the relevant information about what goes on. Good. So that's one way of monitoring. We also have the message monitor. This is really nice. And the idea is we also have a statistical. I don't know if that is actually running by the different messages. Let's see. So if we want to enable this, we go on an agent, say here. support tool enable message statistics and that would enable us to fetch messages in our system it will probably take five minutes before the job runs and create something useful for us and we do have this and this is really nice i think the best way is actually to go under here on the test suite and the test case run so here we have a message that's been run a number of times and what we can do is we can go here and click open figaf browser it'll create a filter four different messages and that means we can actually see a lot more about what is going on in the system we can see there's this figaf mark that has been processed um I would have expected more would be processed, but we have different ways of filtering. We can see message application logs, stuff like that, which all makes it a lot easier to find specific problems.

And you can scroll out here and search for sender receiver, all the custom statuses, all the things that makes it easier. to monitor your SAP integration questions. So that's basically the monitoring. And here in the statistics, we do see the same details as you would have under message monitoring. We keep it for a little longer period of time and you can search for different durations of the messages when they have been processed, stuff like that. I think that's the most important one. we did the groovy development um maybe do the git so if we go to git github so here i have uh yeah okay let's just use this one for now uh so to use the git we will go into our vgov tool and the idea with the git is obviously not to replace everything but just have a separate place where we can have uh integration content so we go in here say enable git we go in it is dgr also sorry demo and so that's the branch we would need a username and to get that we would just go under settings developer settings personal access token fine brained access token we'll generate a new one this that is owner 30 days seven days um only select repository and then we select our repository and we need the permission to content read write generate token cool generate token so cool so now we have generated a token for this i don't think it's just user dgr we can say ignore but that's basically it there's some things also here we can add scenarios we can integrate with git pipelines what are the the details we want to have in here we just want to update these also it is something about how we are building and managing these things ourselves and that means if i go now here and i just test the connection we should hopefully see that it tests the git connectivity and that means we can go and we can check here init git repository which will then download all the content and if we are lucky upload it into a new branch in this repository that is called demo it obviously takes a little longer because we need to do some extra work view all branches demo branch here we can see all the integration content has been added into the eye flow and one of the things that is different is we have package structures which makes it easier to add your uh your integration so why else so here we can see we get the iFlows

into separate packages with the scripts we can see the there's the documentation we saw before that's also been added here. So one hour and 15 minutes. We got time for questions also. If there is some, please do ask. Otherwise, I will go ahead and continue on some of our different ideas here we can see the different transports that's been imported um and but i think here the the nice one is this landscape overview we can see precisely what was changed as a part of this uh this setup um we have the groovy editor uh let's try with the ai so if i go here so i'll just create an open ai key project read only create a secret key so um in here if we want to enable ai we go under here on the ai enable ai and we can see which models it should use as default um i think that this is probably a different one but anyway Let me just see what we're using here. I think as default, it comes up with three fields, a different one. So here's the idea. And that means if we go back to our Groovy script we had just before here, we had the open infigaf browser. And when we do that, we get the AI help also here, uh, help me find the script. We have different options to do partner directory content, look up stuff like that. And obviously it makes sense that you have the testing stuff like that. It came out and gave some, something that looks at least looks a little simpler, moved up some content and we can still then run our test to see does it produce the same result as we got before

compare it's only these two that are different which yeah it's a bit strange but this simplifies a lot of the processing also here in this let me see here on the agents you can go in here now trial agent we have uh after we added it we got the enable monitoring it's under the change tracking tool should be here tracked objects event mesh enable api management access It's always fun that you cannot remember where it is. Is it because it's one of our cloud systems, we disable it because there's a lot of scripts going on, and we do not want to have change tracking here? Yes, so in the other system we do have more options here, enable automatic synchronization, we we don't want large batch jobs running and it doesn't make sense here we have enabled this iFlow documentation here actually it may work so if we go here on an iFlow integration flow scenario documentation that's the same but if we do it here in our demo prod the what you would then be able to see is you you go in here to a iFlow we get this ai generated

description of what this iFlow does and that somewhat saves a lot of the effort that you have in building and maintaining documentation that you can generate a description just AI documentation generate of you together with the scenario documentation it can build these automatically as you progress setting up your scenarios cool I think that was what I wanted to cover there's a question here so there is obviously also a PI to CPI migration journey in this um all all the content in here there is um we have some other videos on that um but the part is when you're doing a phcpi migration journey you need to do a lot of configuration a lot of other stuff um and like transport like giving the customer a better experience about managing this because now whenever i go into my iFlow here if it wasn't because it was not working we get this view back and forth between content so if i have here my scripts so just this overview here where we can see the history here we can see this is the figaf tool it's this transport number that was a part of this that is really useful so we do have this exploring figaf that gives you some that that gives you how do you actually run all of these but here we also get into the phcpi migration scenarios what is it that you can do in those or to simplify the during that was did i put it into the q a that wasn't a problem it should be in the here um yes but yeah the the it's a different part of the tool that handles this migration maybe we would do another session on that if if we get positive reviews on this or feedback on it other questions good ah so now it is up and running i don't know what happened uh so let's see if our test cases here can be sent try we didn't create them other questions did i forget something um it worked so that was really nice and we can save it so we always have this and if you're going to b2b there's obviously also a huge area on this that that covers some of these different scenarios but that is a little more complicated to set up and go through but I hope this gave you a good understanding of what the tool can do how you can deal with these scenarios cool any other questions otherwise I'll say thanks for attending yeah we still have time why is there it's in the message that has that limit yes cool but yes thanks for attending i hope you learned something i hope you want to go take a little closer look in some of these things and see it's not that difficult use the b2b trial tenant use um to to to to do some experiment with the figaf cloud see these transports as you saw we could do transport inside of our b2b trial tenant we can test uh we can create test cases manage run it all these things just to simplify your your effort so yes thanks for for attending