Common ways to avoid and fix Heisenbugs
Table of Contents
Working as software developers, it’s our daily bread and butter to work on different sorts of software bugs and fix them. The complexity of a bug is directly proportional to its reproducibility. Some bugs are cosmetic in nature and are easy to reproduce while others are reproducible only in a certain environment or with a specific set of data. Then there are Heisenbugs, which are inconsistent in nature and hardly reproducible when being studied
What is a Heisenbug?
A Heisenbug is a bug that seems to disappear or change its behavior when you try to debug it. Heisenbugs have a reputation for acting in highly surprising ways. Any attempt to reproduce or debug a Heisenbug may result in its disappearance.
It is a nightmare for a developer to work on these kinds of issues since they know that this is going to be incredibly difficult to study the behavior and make sure that their fix covers all the cases.
The name of this bug is derived from the uncertainty principle conceived by Werner Heisenberg
Example of a Heisenbug
We as developers, often encounter bugs that are difficult to reproduce and not consistent in nature. But I would like to share the popular story of a bug in the “Crash Bandicoot” video game. Almost all of the time the video game would work fine but every once in a while the read-or-write operation would time out and ends up corrupting the memory card.
Have a look at this article for more details on this: https://www.gamedeveloper.com/programming/my-hardest-bug-ever. This will be an interesting read
Causes of Heisenbugs
It is more common to run into issues that are rarely reproducible or inconsistent in nature as software grows in size, complexity, and asynchronicity. The asynchronicity effect, which makes it challenging to reproduce issues in the Debug environment, is the most obvious.
The likelihood of running into problems that identify themselves as Heisenbugs increases with the number of interconnected components (Distributed systems, microservices, etc.). There are more layers that can cover up minor defects and “upgrade” them into severe Heisenbugs as complexity and encapsulation increase.
How to avoid and fix Heisenbugs
A Software Developer needs to have some substantial abilities and focus in order to fix a Heisenbug. While there is no set of universal techniques, there are a few common techniques and best practices that might be beneficial for addressing the problems.
- Debugging may not always help: Since the software that we usually work on is asynchronous in nature so there is a high chance that debug environment may impact the execution speed of Threads and we won’t be able to replicate the issues caused due to synchronization issue or race-condition.
In such cases, it is better to add some extra logs or print statements and then run the app/software in the environment that mimics the prod environment. - Knowing your software: Heisenbugs thrive in asynchronous distributed systems. Knowing your software’s most intricate, encapsulated, asynchronous, and distributed components is therefore always advised because you’ll uncover Heisenbugs there more easily.
- Install static analysis software: Static analysis software can assist developers avoid obvious issues that might go unnoticed during QA verification, such as utilizing the wrong schedulers or neglecting to dispose of the Rx Observables.
- Never believe a Heisenbug: Usually, it’s difficult to tell for sure if you’ve cracked a Heisenbug. It is advisable to err on the side of caution and prepare your debugging equipment.
So this brings me to the end of this article. I hope you find this interesting. If you do, Do not forget to add a comment below and share this article
Want to connect?
My LinkedIn profile.