5 software development skills to learn for rapid development

The world is changing and technology with it.

Main because since software the product deployments are becoming more frequent and Software house and Service-oriented companies are pushing hard on T2M (Time to Market) selling factor to keep them visible on the market.

Although all upcoming projects are prerequisites of non-functional requirements still the same (mainly),

  • the infrastructure design leveraging from Service Oriented Architecture
  • the solution must be scalable and automated to provision
  • the solution capable to be hosted in the Cloud as well as on hybrid network infrastructure
  • the solution is ISO 9126 compliant
  • first release completion time of 6 months

, amount of functional requirements needed for the first release keeps growing and in most cases do not help to achieve delivery in a given time.

And that is pretty bad.

Therefore “smart” selection of the frameworks and tools to use for building whatever investor wants to build a solution is an absolute must.

But, you won’t be able to succeed without the technical knowledge and experience of the production team! (the place where things are getting serious)

To to get familiar with what skill sets to seek out while building the team capable to produce the business value early from the beginning of project initiation, I have created a list of the suggested frameworks and platforms to use.

Hope it helps you to battle this constant competition market and investors pressure and elevate the progress in the initial phase of solution development as much as possible.

  1. Outsystems PaaS

    You are maybe already familiar with the term “low code“. The word on which many conceptual developers are rolling their eyes up. But hold on – if all that investors want is to get the product out of the door as soon as possible and for the cost related to head x time spent on the project (which would probably be somewhere around 50% less as opposed to traditional way of coding in this case), just give it to them!

    Every solid developer must be familiar with this PaaS these days if not with Outsystems then with some other alternative such as PowerApps, for example.

  2. LoopBack

    Heyou – all Node.js Devs are lifting their eyebrow. Yes, very powerful framework, indeed. Usage of LoopBack CLI cannot be easier thanks to documentation built by many contributors from this OpenSource project.

    Simply put, this framework allows you to build your complete backend infrastructure with speed which elevates your project progress exponentially. You can choose from REST, SOAP, GraphQL, RPC servers/services and manage all of these nodes with PM2 process management systems.

  3. Mocha

    Don’t forget about a testing framework. This option will work well with the ones mentioned above and you cannot go wrong with learning this framework right away. Javascript is rocking all over the globe right now and would be silly not to pay any attention to this programming language intentionally.

    And so why not leverage the JavaScript syntax in every SDLC phase? Sounds logical, hm?

  4. AplifyCLI

    This CLI utility from AWS is becoming more and more popular among developers from generation Y. Nobody likes to deal with building the infrastructure on DevOps level, unless it’s ABSOLUTELY necessary. And to be fair a lot of the service provisioning commands can be easily automated.

    Therefore, a utility that scaffolds everything you need for hosting your system is a necessary skill these days.

  5. Terraform

    If not going to use any of these “low code” platforms mentioned above for building your solution, solid provisioning automation system (“engine”) and paradigm for not only infrastructure automation provisioning but also for keeping track of infrastructure changes in source code is the must. You cannot go wrong with Azure DevOps/ARM Templates or Terraform. Both offer you a lot of capabilities and automation to follow IoC (Infrastructure as a code) paradigm.

    To me, Terraform is better options for those thinking to incorporate platforms of different technologies into the solution.

This is all for today, hope you enjoyed this reading today and leave me your thoughts down below in the comments!

Cheers…

PS: The technology cannot set the project for success if architecture, design, and test automation are getting compromised. Not having the right team, implemented processes, following the best practices and a need of keeping good progress momentum on the project, your entire ship can turn in the opposite direction and end up with catastrophic failure.

How to provision Azure Function in Azure by using Terraform CLI

Choosing the right way how to keep an infrastructure versioned and well maintained in source code is becoming a quite big issue in these days. There are several options to choose from on the market currently, and it’s easy to get trapped in a never-ending research cycle. For those working with Azure services only, ARM Templates is more than an obvious answer to this but what if want to have more flexibility in going over beyond the Azure boundaries?

You may be wondering why I should use anything else but ARM templates?

The answer is simple. The ARM templates may be a bottleneck for the IT solutions using different cloud providers (multicloud solutions). In this case, managing infrastructure as a code may become a quite tricky (and ugly) thing to do over time. The thing is that every tool used for infrastructure management has it’s “own ways” of how to work with it and that comes with necessary knowledge base every production team must have beforehand. And as an implication of this, choosing the right tool for your infrastructure management (including deployment) is very important.

Terraform would be a great way how to face this challenge. Just as a proof of its simplicity, what I would like to show you here is a short demonstration of how easy it is to provision Azure function in the Azure cloud by using HashiCorp Configuration Language (HCL) and TF (Terraform) CLI utility in PowerShell.

You might be wondering, what features does TF has over the ARM templates? Well, these infrastructure management tools have “the same” set of functionality but TF has other perks on the top of that, which makes it more secure and convenient tool for DevOps (besides multicloud cloud support).

Key features are:

  1. HCL (HashiCorp Configuration Language) – high-level configuration syntax, well structured and intuitive language (TF also supports configuration using JSON for these JS geeks)
  2. Execution Plans – shows you exactly what is going to happen with infrastructure before the change is getting executed
  3. Resource Graph – the visual understanding of the infrastructure, and in my opinion, Terrafarom has done a very good job on this feature (don’t forget that it’s OpenSource!)
  4. Change Automation – yes, every change needed on infrastructure can be automated -> that means less human interaction -> and less room for human errors, YAY!

If you’re new in Terraform and want to get a feel of what Terraform is, have a look at this introduction video footage with the Co-Founder and CTO Armon Dadgar.

Prerequisites (before we start)

  • Terraform utility downloaded and configured on the local environment (guide of how to do it … here)
    for Win10 users, in case of having an issue with WSL2, I recommend following this article to get over this issue
  • Azure CLI installed and ready to roll (guide of how to do it … here)

Steps to follow

For these going exactly step by step as described in this guide make sure that any resource name starting with ‘ms‘ needs to be unique. I recommend using some other characters as prefix just to be sure that this exercise on your side will go smoothly. You won’t go far with copy&paste technique here – oops!

  1. Log in to Azure by using Azure CLI (Azure Command Prompt) or PowerShell
az login

2. If you have multiple subscriptions, skip this step otherwise. List them all out by running this command and choose the one wanted to be used (subscription_id )

az account list
Subscription details after login

3. Find out what is the latest supported AzureRM provider here (at this time of writing this post 2.29.0). This step is not mandatory but I would highly recommend to do it this way as AzureRM API might change in future so better to have a version of the CLI referenced to the code batch file.

4. Create a folder and the file within main.tf (mine is located at c:/Temp/terraform-test/)

5. Add this snipped code at the beginning of the file. This will configure Azure CLI authentication in Terraform

provider "azurerm" {
  version = "=2.29.0"
  subscription_id = "<your Azure subscription id from the step 1 or 2>"
  features {}
}

6. Append the file with the rest of the script from below. For this exercise, the data centre in Australia Central is going to be used (but change it if you like), new Azure function is going to be using consumption service plan as well as running on Windows OS (this is the default option anyway – change it to the Linux if you wish)

resource "azurerm_resource_group" "example" {
  name     = "azure-functions-cptest-rg"
  location = "australiacentral"
}

resource "azurerm_storage_account" "example" {
  name                     = "msfunctionsapptestsa"
  resource_group_name      = azurerm_resource_group.example.name
  location                 = azurerm_resource_group.example.location
  account_tier             = "Standard"
  account_replication_type = "LRS"
}

resource "azurerm_app_service_plan" "example" {
  name                = "azure-functions-test-service-plan"
  location            = azurerm_resource_group.example.location
  resource_group_name = azurerm_resource_group.example.name
  kind                = "FunctionApp"

  sku {
    tier = "Dynamic"
    size = "Y1"
  }
}

resource "azurerm_function_app" "example" {
  name                       = "mstest-azure-functions"
  location                   = azurerm_resource_group.example.location
  resource_group_name        = azurerm_resource_group.example.name
  app_service_plan_id        = azurerm_app_service_plan.example.id
  storage_account_name       = azurerm_storage_account.example.name
  storage_account_access_key = azurerm_storage_account.example.primary_access_key
}

7. Navigate to the folder with main.tf file created, open Command promp or PowerShell and type this command below

terraform init

This action will the create selections.json file at .terraform\plugins\ and download the AzureRMplugin into the .terraform\plugins\registry.terraform.io\hashicorp\azurerm\2.29.0\windows_amd64 directory. The CLI utility starts its live time with the batch files from now on – perfect isolation approach from the running environment (although CLI utility itself may be quite hungry for disk space!).

The selections.json file content

8. You can skip this step if in hurry but continue reading if you want to know more about how to generate the infrastructure change plan … Open the Command prompt or PowerShell and run this command from below to see the infrastructure plan before the change execution. I can strongly recommend using some advanced IDE like MS Code for working with TF (because the text editor and the command console are all integrated into one app) as opposed to switching from the text editor back to command console – this can be annoying…

terraform plan

The plan should look similar like in the screenshot below. For those using MS Code, I would recommend downloading HashiCorp Terraform extension to accelerate your further IaC development – I found it very useful in time efficiency!

Terraform infrastructure change plan

9. Let’s get ready for D-day. Type this command to apply and execute the changes to Azure

terraform apply

This command is going to generate the infrastructure change plan and prompts the confirmation message to the user – I am happy with the planning changes, so typing yes.

The Terraform confirmation message

10. ..and if everything has finished successfully, you should be able to see this message in the end

Resources successfully created
Resource group with all resources created in Azure portal

Entire main.tf file content

provider "azurerm" {  
  version = "=2.29.0"
  subscription_id = "834b29c3-9626-408d-88e0-12e92793d1f5"
  features {}
}

# Azure functions using a Consumption service plan on Windows OS (default option)
resource "azurerm_resource_group" "example" {
  name     = "azure-functions-cptest-rg"
  location = "australiacentral"
}

resource "azurerm_storage_account" "example" {
  name                     = "msfunctionsapptestsa"
  resource_group_name      = azurerm_resource_group.example.name
  location                 = azurerm_resource_group.example.location
  account_tier             = "Standard"
  account_replication_type = "LRS"
}

resource "azurerm_app_service_plan" "example" {
  name                = "azure-functions-test-service-plan"
  location            = azurerm_resource_group.example.location
  resource_group_name = azurerm_resource_group.example.name
  kind                = "FunctionApp"

  sku {
    tier = "Dynamic"
    size = "Y1"
  }
}

resource "azurerm_function_app" "example" {
  name                       = "mstest-azure-functions"
  location                   = azurerm_resource_group.example.location
  resource_group_name        = azurerm_resource_group.example.name
  app_service_plan_id        = azurerm_app_service_plan.example.id
  storage_account_name       = azurerm_storage_account.example.name
  storage_account_access_key = azurerm_storage_account.example.primary_access_key
}

Also available on GitHub https://github.com/stenly311/Terraform-AzureFunction-InAzure

Overall Terraform CLI rating

  • Cloud provider portability
  • Fewer lines needed to achieve the same infrastructure configuration need compering to Azure ARM Templates
  • Intuitive and fast to learn
  • OpenSource with a wide collection of “get-started” production like examples
5/5 Rambo rating