On Memory Utilization of Prometheus Exporters

Prometheus is a popular metric reporting framework. It utilizes a pull-based collection system where exporters expose endpoints and a centralized collection agent periodically requests metric values. This scheme has been validated by prolonged testing and scaled deployments. However, while there are certainly “best practices” in defining metrics, the exporter API does nothing to safeguard you from inefficiencies. In our applications, this means extremely high memory pressure on metric exporters using the Prometheus golang client....

2022-07-07 · 7 min · Daniel Rammer

Using a Pull-Through Image Registry with k3d

In my daily role I am a software engineer at Union focusing on backend development. We are the core maintainers, and the most frequent contributors, to the open-source Flyte project. Per our documentation it is described as “The Workflow Automation Platform for Complex, Mission-Critical Data and Machine Learning Processes at Scale”. Basically, it is a framework to abstract execution of complex data workflows and the coinciding cloud infrastructure management. This enables teams to efficiently and effectively scale data processing....

2022-04-15 · 6 min · Daniel Rammer

Installing Debian 11 with btrfs Subvolumes on LUKS

While btrfs may be relatively “new” in regards to file systems, it is receiving widespread adoption, being included as the default FS in many distros. They is because it provides many advantages when compared to the traditional luks on lvm encrypted linux install: Subvolume Snapshots: btrfs is a copy-on-write FS (CoW) meaning changes to an individual file result in a new file being written. This means that each subvolume can be snapshotted, where a complete version of every file is captured at a moment in time....

2022-04-14 · 5 min · Daniel Rammer

Password Management with Stash

Password management is becoming increasingly important. Best security practices suggest using difficult to remember passwords consisting of either long string of unconnected words or randomized alphanumeric strings. These per-password difficulties are compounded by the increasing number of services brought about by the digitization of everything. While there are certainly efforts the improve this space, they all have specific limitations and the non-emergence of a widely adopted solution tells just how significant these limitations are....

2022-04-12 · 3 min · Daniel Rammer

You Should Have Something To Hide

Issues of privacy are coming to the forefront of our society. Most notably, the continued evaluation of Big Tech by the US Government. In casual conversation I frequency see privacy views as highly polarized, trending towards either “nobody should know anything about me” or “I have nothing to hide”. I am a stout privacy advocate many for reasons. As such, I tend to disagree with the latter argument - very few people are so pure that they would be OK with everybody they ever meet seeing everything they’ve typed into a search engine....

2022-03-04 · 2 min · Daniel Rammer