Paulo Martins

Security Engineer
Infrastructure and Cloud Engineer
Software Developer

Embracing German in Switzerland

This is a work in progress I’ve recently moved to Switzerland, more specifically to the German part of Switzerland. Even though getting an apartment is incredibly complicated, I managed to find something small close to the city center. While living in the city center means that I can get by just fine by speaking only in English, the fact that everything around me is written in German and I don’t understand it leaves me a bit uncomfortable.

Jupyer Lab in Docker

TL;DR Install Docker Open your terminal in the directory of your project Run this in your terminal: docker run -it --rm -p 8888:8888 -v "$(pwd)":/home/jovyan/ jupyter/base-notebook:lab-2.2.5 Open Why If you work as a data scientist, you most likely use Python and Jupyter notebooks to run experiments and explore data. Also, you were already in a situation where your experiment worked “in your machine” but didn’t work for no one else.

AWS Elasticsearch Ultrawarm cost analysis

In 2010, Elasti.co created the Elasticsearch engine. Five years later, in 2015, AWS launched the Elasticsearch Service. Last year, an OSS licensing war between AWS and elastic.co led AWS to create their own “flavor” of Elasticsearch called OpenDistro. At first sight, OpenDistro is very much like the “original product”, but as we work with more advanced use cases, we will start to notice some differences. Most differences relate to the number of fantastic plugins provided by elasti.

Mastering AWS EKS Logging with Elasticsearch

When you start building a Kubernetes cluster in AWS, one of the first things you need to think about is how to store and centralize the logs coming from nodes, containers or processes. A commonly used solution to solve this problem is to send your logs to Elasticsearch. It scales horizontally. You can store up to 3 PB of data in a single cluster (if you can afford it). Because of the Inverted index, you can do a “free search” in all your logs fairly quickly.

From Excel to Jupyter (with boxplots!)

Why Jupyter? In my first job as a risk analyst, Excel was the go-to tool for analyzing data and building charts for technical reports. We would extract files from the central database, copy them into Excel files, do some data processing to build meaningful tables, and then we would use Excel’s predefined charts to create visualizations. Excel is a pretty powerful tool and there are a lot of tasks where Excel is the right tool to use.

Kubeflow on Windows

After following the Kubeflow project for a while, I decided to try it out. Even though I am an OSX/Mac user, I recently bought a new ThinkPad that came with Windows installed and took this chance to explore what is new in Windows and the Windows Linux Subsystem (WSL). To test the limits of WSL, why not try to make Kubeflow run on it? That way, I can check if some must-have features like containers, Kubernetes, port-forwarding, etc… are fully available, and how easy is it to make them work.