Nasza strona używa cookies. Dowiedz się więcej o celu ich używania i zmianie ustawień w przeglądarce. Korzystając ze strony, wyrażasz zgodę na używanie cookies, zgodnie z aktualnymi ustawieniami przeglądarki. Rozumiem

The warning signs of impending IT disaster

A set of questions for every project - to pacify the problems that might lead to a real IT disaster.
23 08 2018

With more than ten years in web development, we’ve experienced all kinds of IT problems which sometimes led to a real disaster. To avoid these dreadful mistakes, we came with a checklist of questions that we use for every IT project. If you can’t answer most of the following questions, or do not understand their importance, your IT operation might be on the verge of a costly catastrophe.

Who, Where, How

Who is involved in a project? Where are we with the project? What is its current state? How many tasks are supposed to be deployed? If you can’t answer these questions, you probably don't use any task tracking system. We recommend adopting one as soon as possible.

The tickets

Have we estimated all tickets (project tasks) we’re currently working on? Do we have actionable steps to execute for each of them? Having ignored these questions, your tickets might stay in the “In Progress” bucket much longer than they need to. Prepare all estimates, plans of execution for complex project tasks, and/or split them into smaller more-manageable pieces.

Version control tools needed

Do we create pull requests? How often do your developers commit code? How is your branch structure organized? If you can’t answer these, you might not use any version control tools yet - start using one as soon as possible!

Updates, entropy & regression

Do you know how much entropy and regression every update causes? If your tickets return from “To Be Tested” back to your “To Do” list, consider getting some external QA help so your developers are not responsible for testing their own work. Run some
basic tests using tools like Ghost Inspector that will at least automate this approach, saving time.

Dear PM

Tired of asking your Project Manager to check updates on staging? If your tickets take too long to get to staging in order to be tested as a batch of features, you need to configure Continuous Integration tools that will automate a process of rolling out your changes to the staging instance.

Bugs, bugs everywhere

Tired of seeing all that hate from users in your inbox? If production releases of your app receive numerous bug reports from users complaining about data loss or buggy functionality, try to do partial deployments and perform a load balancing to
present new features to a limited amount of users and not scare them all at once.