Press "Enter" to skip to content

Category: Testing

Messy Code and Reasonable Expectations

Published 2020-09-04 by Kevin Feasel

Rachel by the Bay has a doozy of a story:

One day not so long ago, I was in a meeting listening to a team explain why their service had gone down and taken out a big chunk of a business. They were one of those things that has to exist and work in order for the actual “thing that makes money” to go. Think of delivering pizzas, connecting dog walkers with dogs who need to be walked, that kind of thing.

It turned out they had been crashing every time a request came through for a certain part of the country. That is, not all pizzas, dog walkers, or whatever it was were handled identically, so they had their own city or region configurations. Think of differences in pricing, taxes, features, or whatever. Trying to process a request for this one particular region had caused the entire process to die when it hit a new config that was “bad” somehow.

Read on for the story. This sounds like a boundary issue. Boundaries are messy and need thorough examination to handle as many possible points of failure as is reasonable. Taking seriously the point that it makes the code messy, the answer is not “Don’t do the checks,” but rather “Put the checks in a place where their messiness has a minimal impact on the rest of my beautiful code but still does the important work we need them to do.” Failing that, live with the mess and have a working process.

Comments closed

Testing Columnstore Data Loads on Eight-Socket Servers

Published 2020-09-01 by Kevin Feasel

Joe Obbish puts on the lab coat and safety goggles:

I elected to use a high concurrency CCI insert workload to compare performance between a four socket VM and an eight socket VM. Quite conveniently, I already had a test columnstore workload that I knew pushed the SQL Server scalability limits in terms of memory management. To perform the threading I used the SQL Server Multi Thread open source framework. I wanted all sessions to go to their own schedulers. That could have been tough to manage with tests up to 200 threads but the threading framework handles that automatically.

For those following along at home, testing was done with SQL Server 2019 with LPIM and TF 876 enabled. Guest VMs were built with VMware with Windows Server 2019 installed. The four and eight socket VMs were created on the same physical host with about 5.5 TB of RAM available to the guest OS in both configurations.

Read on to see how an eight-socket server fared in comparison to a four-socket server in this task.

Comments closed

Mutation Testing in Action

Published 2020-08-28 by Kevin Feasel

Nathan Thompson walks us through a mutation testing experiment:

Since our hypothesis was that the implementation differences between Jasmine and Jest could affect the Mutation Score of our legacy and new test suites, we began by cataloging every bit of Jasmine-specific syntax in our legacy suite. We then compiled a list of roughly forty test files that we would target for Mutation Testing in order to cover the full syntax catalog. For each file we generated a Mutation Score for its legacy state, converted it to run in our new Jest setup, and generated a Mutation Score again. Our hope was that the new Jest framework would have a Mutation Score as good as or better than our legacy framework.

By limiting the scope of our test to just a few dozen files, we were able to run all mutations Stryker had to offer within a reasonable timeframe. However, the sheer size of our codebase and the sprawling dependency trees in any given feature presented other challenges to this work. As I mentioned before, Stryker copies the source code to be mutated into separate sandbox directories. By default, it copies the entire project into each sandbox, but that was too much for Node.js to handle in our repository:

In my undergrad days, I loved mutation testing mostly because of the terminology. I’m happy to see a proper implementation of mutation testing and I’m even happier to see that they have a .NET version.

Comments closed

Calculating Test Coverage of Azure Data Factory Pipelines

Published 2020-06-22 by Kevin Feasel

Richard Swinbank wraps up a series on testing in Azure Data Factory:

To determine which activities have been executed by a test suite, I need to collect and aggregate activity run data from every pipeline execution triggered from any test fixture. In the previous post I developed components to retrieve and cache activities for a pipeline run – I’ll use those components here to collect data systematically.

I’m going to create a new helper class to contain functions specific to coverage measurement. It’s a subclass of the database helper because I want to exploit functionality from classes further up the hierarchy:

Read on for the code and process for measurement.

Comments closed

Unit Testing Azure Data Factory Pipelines

Published 2020-06-05 by Kevin Feasel

Richard Swinbank walks us through what it takes to run a unit test against an Azure Data Factory pipeline:

In part three of this series I looked at functional tests for ADF pipelines: verifying, in isolation, that pipelines are “doing things right”. In this post I’ll be testing isolated pipelines to check that they’re “doing the right things” – this is one description of a unit test. In a general-purpose programming language, unit tests might be used to verify that an individual line of code is executed, or that it has a particular effect. In Azure Data Factory, the smallest unit of development – a “line of code” – is a pipeline activity. I will be writing tests to verify that specific activities are executed (or not) and to inspect their results.

There’s a fair bit involved in this sort of test.

Comments closed

Removing and Refilling All Tables in a Database

Published 2020-05-18 by Kevin Feasel

Phil Factor has a couple T-SQL scripts for us to remove and reload a test database:

Copying a database using a BCP dataset from a DOS script or from PowerShell is fairly quick and trivial, but what if the data is already in another copy of the database on the same instance? Imagine you have a database build that is a copy of an existing database on the same instance and you want to run a ‘destructive’ test on the data, and do it over and over again.

Deleting all the data in a database should, you’d have thought, be trivial. Truncation doesn’t work because it doesn’t like to truncate tables with foreign keys, even if they are disabled. DELETE is safer but you will need to then reset any identity columns. Deleting large tables in one chunk isn’t scalable, so you have to do it in smaller chunks. I like to disable triggers while I do this as well. The only disadvantage comes if you have triggers that are supposed to be disabled because the simple code just re-enables them all.

This is, I think, one of the biggest selling points for containers where the database is built into the container image. You spin up a container based off of an image, perform your destructive testing, and destroy the container afterward. The next time you need to run these tests, spin up a new container. And if you need to change the data, modify the container. This introduces new challenges like how SQL Server on Linux has limitations which don’t exist on Windows, but for supported functionality, it’s a nice solution.

Comments closed

Running Postman Collections via Newman and Jenkins

Published 2020-05-06 by Kevin Feasel

Ankur Thakur takes us through configuring Postman automation:

We can write the automation test suite for any service which can be used for performing regression testing. But we can also integrate our automation test scripts using Newman in Jenkins CI/CD pipeline which can be triggered automatically whenever a pull request gets merged.

We need software installed on our system locally:

– Jenkins
– Node
– NPM
– Postman (For writing the collection)

We’ve done this in our environment and I’m reasonably happy with the results. If you have a REST API, it’s a fair sight easier to work with than something like Specflow.

Comments closed

Operations Testing with Pester

Published 2020-04-22 by Kevin Feasel

Sheldon Hull takes us through using Pester to automate operations tasks:

In my example, let’s start small and say you just have PowerShell, and some servers.

What I’ve discovered is that to actual validate DevOps oriented work is completed, you typically go through the equivalent of what a Cucumber test would have. This “checklist” of validations is often manually performed, lacking consistency and the ability to scale or repeat with minimal effort.

Consider an alternative approach to helping solve this issue, and expanding your ability to automate the tedious testing and validation of changes made.

Read on for an example as well as some additional thoughts from Sheldon.

Comments closed

The Basics of tSQLt

Published 2020-04-15 by Kevin Feasel

Jess Pomfret walks us through the basics of tSQLt:

Getting started with tSQLt is really easy- you download a zip file, unzip the contents and then run the tSQLt.class.sql script against your development database.

There are a couple of requirements, including CLR integration must be enabled and your database must be set to trustworthy.  These could open up some security considerations, but for my development environment it’s no issue.

This is where I’d say putting the database in a container would be extremely helpful, as then you can destroy it afterward. I’m not sure if that’d work, as SQL Server on Linux doesn’t support unsafe or external access assemblies (and I’m not sure what tSQLt requires there).

Comments closed

The Importance of Unit Testing Database Code

Published 2020-04-15 by Kevin Feasel

Chris Johnson shares some thoughts on unit testing database code:

This is a topic that is quite close to me heart. I don’t come from a computing background before I started working with SQL Server, so I was quite ignorant when it came to a lot of best practices that other developers who have worked with other languages are aware of. Because of this, I had no idea about unit testing until I attended a talk at a SQL Saturday all about tSQLt. If anyone isn’t aware (as I wasn’t) tSQLt is a free open source unit testing framework for use in SQL Server databases. It is the basis of Redgate’s SQL Test software, and is the most used framework for writing unit tests in SQL Server.

Since then I’ve worked to try and get employers to adopt this as part of a standard development life cycle, with mixed success at best. My current employer is quite keen, but there are two major problems. First, we have a huge amount of legacy code that obviously has no unit tests in place; and second, the way people code is not conducive to unit testing.

Click through for additional thoughts on writing good tests and an example of modularizing code to make it more testable. I’m still in the camp of “test what you can, but you can’t test everything” with databases. There’s just too much state dependency.

Comments closed