<p>about your learning experience to help us improve.</p>

You've completed

cleanFiles/bookmap_Getting_Started_with_Terraform_on_SAP_BTP_en-US_learning-sap-com_lsc-prod.zip (MMCP-DROPZONE) - 2025-02-18T09:44:29.000Z

Setting Up a Cloud Foundry Environment via Terraform

Lesson Content

Getting Started with Terraform on SAP BTP

SAP_MOOC_terra1_unit4_lesson4_CLOUD_Video.mp4

Hello and welcome to Unit 3 Lesson 4, where we will be setting up a Cloud Foundry environment via Terraform. So this is actually also the last piece in this unit of our journey to enhance our Terraform configuration. We will set up Cloud Foundry within the overall flow. So meaning that we will set up a Cloud Foundry environment in our script and use the Terraform outputs to expose information which is created during the provisioning of a Cloud Foundry environment. And talking about that, how can we properly provide information from a resource, Christian? Well, that's a good question. So, first of all, when it comes to the setup of an environment, we already know how this is working. It's the same approach that we applied in the previous lesson when adding additional resources. An environment is just another one. Having said that Cloud Foundry is a kind of special with respect to that, we want to have information from the Cloud Foundry provisioning, namely the API endpoint to access the environment, for example, via the Cloud Foundry CLI. To achieve that, without having to dig into the state information, Terraform offers so-called outputs or output values to surface specific values of the configuration. Those outputs are the analogy to returning values in a programming language. In Terraform, we define them via an output block as you can see on the slide. The output block has a name. In this example, as you can see, it is subaccount ID. That's how we would address it. We also must assign a value to this output, in this case the subaccount ID of the subaccount resource that we provided. This way we explicitly hand back the subaccount ID, so the technical ID, of the provision subaccount to the caller. Now, let's go in action and see how this works in the hands-on. Yeah, so in the hands-on, we will enhance our Terraform configuration and will create a Cloud Foundry environment with it. And we will explore the Cloud Foundry API endpoints as an output of the configuration that can also then be used at a later point in time. So let's get our hands dirty. So we are back in our GitHub repository for Unit 3, Lesson 4. Let's have a look what we need to do now. So we want to add the Cloud Foundry environment. And as just said before, the way we do it is actually the same way like we did as well for the other resources. There is actually one specific thing about the Cloud Foundry environment where we will have to have the initial variable called CF landscape label. That's something that we need. We set the default here to empty and then we'll use this landscape label to actually create the Cloud Foundry instance. So let's do this now maybe step by step, Christian. So we... Yeah, so let's start with the environment. I mean, again, as we said, you find the information in the Terraform registry, in the documentation. Now, with environment instances, you really have to know how this environment instance works in order to create it. So we have several examples, but you really also have to take a look at the Cloud Foundry or kima documentation in order to know what to fill in these parameters, because they might be quite generic. Yeah, and as we just said right for the Cloud Foundry environment instance, we need to have this Cloud Foundry landscape label as additional variables, so let's add this to our variables.tf file first. So we save this, and now next we will actually take the necessary resource and let's copy that over for the sub account for the environment instance in our main tf file and we see here that there is now in line 62 and 68 something red yeah so so we we also have to give it a name of course and this is something usually when you give a name for resources you put that in the locals block in the locals block because you probably will have naming conventions on that. And well, we do that by defining the subaccount CFO, local, via the subdomain. So we just keep it easy here. Maybe we save it again. We save it again, yeah. Terraform FMT will do the job for us. So meaning that our subaccount CFORG is now the same like the local subaccount subdomain, right? So we could, of course, also directly set the subaccount subdomain, but maybe at a later point in time we want to extract things or you learn more. So let's do it also. It's a bit over -engineered at this point in time. Well, and the rest is really then Cloud Foundry specific. So we have the environment type, the service name, the plan name. We are on trial, so we have to have the trial name. And, yeah, that's part of the story now. landscape label. Let's add some information there because maybe you don't know the landscape label. Maybe the caller doesn't care about the landscape label. So we want to fetch that landscape label from the platform if it's not provided by the caller and dynamically add some default value to that. For that, we again take one data source that we have, right, for environments. Yep. So that's what we add. So we read data in order to get the environment information. So we add this as well again to our main TF file. We add the data block here for environments. Yep. And based on that information we now have to extract, first of all, the information. If there is no landscape label, we take this data source and the first matching environment from the data source that we get, which is a Cloud Foundry environment, let's take that landscape label and use it as input for the resource. Now, if we would do that via a local block, then this block would be evaluated every time you execute the script, which might then result in different values. So we use one additional resource, the so-called Terraform Data resource, to store this label. And it will become also part of the Terraform state. Exactly. This way we push it into the Terraform state. So once it's provided, it won't change, as long as you don't change something. So that's important to do. So let's switch to VS code again. We add this again, Terraform data resource for the CF landscape label. And then you see a well-known friend. the ternary expression. So we say if the length of the landscape label is greater than zero, then take the variable of the landscape label. If not, then, well, iterate through the environment's result, so the data source through the values, and then check if there is a service name with Cloud Foundry and an environment type of Cloud Foundry. And then take the first one. And then take the first entry and take the landscape label from there. So, as you can see, there you really need to have a lot of knowledge around Cloud Foundry, how things work on BTP in order to have a, I would say, user-friendly provisioning of the resource. But, I mean, here you have it. You can use it at home. And with that, I would say the setup looks quite good for the resource. Now, you said we also want to have some information exposed, right? Yeah, specifically around the, for example, the subscription, we might want to know the URL for that subscription. We also might want to have the API URL of the Cloud Foundry instance and actually the idea of the Cloud Foundry organization that will be created. I think that's something quite useful also, as you said before, to use these kind of information in other CICD pipelines maybe. So that's something that we should introduce now with the outputs, right? So a new concept that we will be introducing now. Exactly. So we will create a new file called outputs.tf, and that we will put in this information. So let's do that. In our directory, we add a new file. Outputs.tf. A little typo. Of course, and then let's put that in. Now, as you saw on the slides, we just assign a value to the output. The output has a name, API, URL, org ID, landscape label. This will be filled with the value, and it makes sense to add some description. And again, as you can see here, it's not super trivial to extract the API endpoint or the org ID, but it's part of the instance information. So in real life, you might want to provision that stuff, then take a look into the state and then extract that information because you don't know that from the top of your head. But I think here the whole concept of infrastructure as code is really something where you immediately see the value, right? So you are putting knowledge about the platform and the way it works into code so that it's described, right? So you don't have to ask somebody how to do this. you put it into code, and from there on, you can use it, right? Yeah, that's right. Now, we stayed quite slim with the output. When you take a look at the ReadMe file, you see that there are some more attributes that you can set for the outputs that we don't use in our setup. But also here, looking into the, let's say, the generic Terraform documentation, and as Terraform is quite well known, there are a lot of web pages where you can find out more on how to use this. But for now I think we're ready to go, right? For our setup, that's all that we need to add and that we want to add. We saved everything. So let's start the Terraform CLI game again. The same procedure. Terraform FMT. Okay, we made some miss formatting. Which is now fixed on the main .tfile. Which is now fixed. It was, I think, in the local spark, right? Yeah, we did that intentionally. Then let's do the validation. Looks good. So we didn't make any error there. And then we have the Terraform plan, one is out, and then we have Unit 3-4. And this will now be interesting what happens now. So again, the game starts with refreshing the state. Fetching all the data from the platform and And as you can see here, we have two resources that will be added. Yes, because we added this Terraform data resource. And our Cloud Foundry environment instance. Now, you might wonder about one thing. We didn't do a Terraform init, right? But we didn't initialize this Terraform resource before. There are some resources that are built into Terraform where you don't need to do a Terraform init. So that's why this worked just in case you are having some question marks there. And as you can see, we have this outputs defined, but of course there are not known. And we see in the CF landscape label in our case that we intentionally did not provide a landscape label because we want Terraform actually to create it. And by default, it's set to empty. So let's see what happens if we now apply the changes to our... global account. So we see that the environment instance is being created. Okay, so now we have a little bit of a mistake there. So we have this, obviously it's the landscape label that causes us some trouble there. So what did we do wrong? Well, let's take a look at the main TF file. We have the landscape label and we did a lot of of work to set it to the Terraform data resource, but we did forget to change it in our resource, right? Yes. Oh mate. So let's do the Terraform data. It's already helping us. So we have Terraform data, then we have the CF landscape label. And the result. Because we did not define it, but we actually wanted to fetch it. So that was the error. That was the error. So let's see what happens now if we try this again. So let's do... Of course we have to redo the planning. Okay. I think there was some issue with the... You type terraform, I think, in the... There is a typo. There is a typo. Yeah, yeah. If you go to the main TF file, you see you missed the A in terraform. The A and terraform is important. Yes, so we try this again. No. And the... So I think the type of head support tricked us here. Yep. Let's... Yeah, let's shortly check what we did. I think we just need to have the... The output. Ah, output. There it is. We could have copied over, but I think it's also good for you to see some common mistakes that can be made, also some errors. So let's try again if it works now. Looks better. Looks more promising, I would say. Okay, it's refreshing the states, the state of the Cloud Foundry environment instance because we tried it before. Okay, now it says only one change because the landscape label is now there. We're fetching it. Yep. And now let's do... Yeah, so that's kind of because we did the planning before, and now we have to do it again. So let's do the Terraform Apply. And keep our fingers crossed. Yep. So, of course, because we created it before, it now needs to be destroyed first. As we saw in the sub account before. Yes, exactly the same thing, right? It's exactly the same thing. Now this might take a bit. Yeah, so let's wait until it's destroyed. And recreate it then. And here we go. It has been destroyed and has been added. Wonderful. Created again. So that's cool. And now we see we have the CF API, we have the landscape label, we have the org ID, let's shortly switch to the sub account, let's refresh it. And we should actually now see, also see the Cloud Foundry environment. That's what we hope to see. And there we are. There we are. As expected with the right org name, as expected with an API endpoint that we also have here in our output. Now, if I do a clear, the output is gone. Do I have access to the output again? Of course. Terraform output. It's Terraform output. So if you unintentionally clear your console and you now struggling, where do I get the output from? Don't worry, Terraform output, and here you go again, because the output is also stored in your state file. And I think we are... With that, I would say we are good to finish the hands-on, right? Yeah. Fantastic. Now we're back from the hands-on part. What did we learn in this lesson? First, we learned that Terraform can explicitly expose information from the configuration and the concept to do so is called outputs. These outputs come in quite handy when it comes to use them in follow-up. activities after the configuration is provisioned. One example would be a CICD pipeline where you might want to kick off some further activities using this information. And yeah, with that, we are at the end of this lesson. Thanks for watching and looking forward to seeing you in the next lesson. Bye.