Skip to content

Hello, I am interviewing people…

Hello, I am Kris (Article about a Twitter meme in english). I was a Database Scalability Engineer at

When interviewing people for my replacement, I have been looking for a “sense of cost” in architectural decision making. I have been giving people rough problem descriptions of what others or I am currently working on, and asked them what they were thinking about the problem.

Since I did not have a solution to the problem yet, my goal was more observing the idea-discovery process and the reasoning about consequences of various approaches in terms of load, toil, network communication and so on, than an actual solution for the problem.

Oh, and I have been a MySQL consultant and database person since 2005, and still google ALTER TABLE syntax.

Many years ago, when I was still at university, I was earning money for my studies by teaching stuff to long term unemployed people at Akademie für Berufliche Bildung in Kiel.

Classes at afbb ran for one year, and I was supposed to teach Unix Sysadmin and Networking to people. We wrote four large tests a year.

The first one was without notes.

That was the easy one, because a solution was deemed to be correct if the idea and general outline of commands was correct. The goal was to verify that the candidates understood certain basics that every Unix person is to know by heart (permissions, for example, or some basic commands and everyday procedures).

The three tests later in the year had to be done with Book, Notes and Manpages available.

The main idea was: If it would be available to you being alone on a site, you could use it in the test – with certain adjustments to equality requirements for all candidates. These tests were much harder, because if you can validate and look up things, they better be correct in your writeup.

Published inWork


  1. Hartmut Holzgraefe

    When studying electric engineering I could bring as many books and devices as I wanted to almost any written test.

    The only exception that I remember was the main analysis test after the 2nd semester of math lectures where calculators were banned.

    Reason for that was that the first calculators were being able to do symbolic differentiation and integration, so the options for the tests were either:

    a) we’ll make the test hard enough so that calculators will not be able to solve it, and so most students will not able to solve it either

    b) we’ll ban calculators, and we’ll make sure that all numeric results are simple integers, so that you’ll know that you have made a mistake if you end up with a result with decimals

    For obvious reasons they chose b)

  2. Alex

    I have been interviewing people over the last 4 years, from a couple per year to a couple per month. Mostly product management and technical positions. When it came to “tests”, we gave them a case and – the internet, if they had their laptop with them. What I tried to understand was their ability to grasp a problem and be able to express the challenge (and then propose a solution). That’s one aspect.

    The other aspect is, that I expect someone who has studied economics, to know the basics of microeconomics. The concepts are so simple, that you do not need a book to explain to me the principle of decreasing marginal revenue and apply it to a case (admittedly, I interview in german and had just to use google to be sure about the expression in english). Similarily, I do expect a network engineer to be able to simply explain what ARP does and how a switch works. I would not expect him to write IOS commands on the whiteboard, but the understanding must be shown.

    As an interviewer, I need to make that difference – things you need to know by heart and things you can safely look up.

Leave a Reply

Your email address will not be published. Required fields are marked *