There is an infamous saying: One cannot understand thing until they made it. In this series, Tim shared how to approach writing a debugger. This part is about breakpoints but make sure to check other parts too.

There are many options that is convenience but it isn’t secure enough. Such as disk encryption is not enable, lock screen isn’t auto activate. This is a guide to enhance your security, especially useful when connecting to public wife for example.

Disk IOPS (Input/Output Operations Per Second) is a key metric for measuring the performance of any disk system. It represents the number of read and write operations that can be performed per second. For PostgreSQL, which relies heavily on disk access, understanding and optimizing disk IOPS is crucial for achieving optimal performance.

Dolap accidently remove their comment service infra. They click delete on “Cloudformation” stack, which is a process that cannot be cancel. Panic. The key thing help them is “deletion protection” on RDS. While normal stateless component will be removed, the core place where data remains is safe. So do yourself a favor and enable “deletion protection” on all of your stateful resource today.

It’s common for modern hosts to cache static assets in a flexible, but not most optimal way. Let’s explore why that is and what we can do to push cache performance (for some assets) even further.

How to be kind and provide constructive feedback instead of being a brilliant jerk?

AWS RDS has a physical storage size limit of 128TiB for each RDS cluster. Postman has a db of 95TB and growing at 2TB per month. So they plan to prune the old data. However, just drop a table won’t reclaim the space. They needs to rebuild the table, but it’s a costly process. This is real world engineering, read to see how they become creative and overcome the challenge.

How software fails is interesting. But real-world errors can be infrequent to manifest. Fault injection is a formal-sounding term that just means: trying to explicitly trigger errors in the hopes of discovering bad logic, typically during automated tests.

This article is a Rails solution by 37signals fork themselve. But many principal can be applied to anyone. Every engineer would need to debug issue, and have a need at poke into some kind of console app to access to the sensitive data. How do we log those access, get approval to run those. how they glue these parts together

A very practicle advice you can do today to handle front-end upgrade. It leads us through a process of setting up github action, run and take screenshot of your app. Compare them visually. of course, we will need some tips to hide flaky element along the line. Such as when first load page there might be no pop-up, but waiting 5 second and a pop-up may appear. Or having animation that appear based on time.