AJ's Blog

Fedora 42 upgrade

As of April 2025, Fedora 42 is available. Compared to the previous release, Fedora 41, this is a quite important release because well 42 is the answer to life, the universe and everything. That’s an old reference to scifi literature.

This release comes with updates like GNOME Desktop 48. This introduces v-sync triple buffering to animations in the desktop. There is also a new installer for Fedora that uses a web UI. I believe this will show up if you use a Fedora workstation live image to boot and install.

Grafana Promtail to Alloy migration

Promtail was a project I used as part of the Grafana Loki log collection ecosystem. It was deprecated as Grafana now recommends using the Grafana Alloy project to help collect telemetry data. Promtail made it relatively easy to collect logs from containers and files on a *nix system. I also used the Kubernetes helm chart to collect pod logs in Kubernetes and forward to Grafana Loki.

My last post about Grafana Loki was upgrading from v2 to v3. Now that my Loki server is upgraded, I am migrating my installations of promtail to use Alloy instead for log collection.

Prometheus remote write

Prometheus can be configured to write metrics to a remote prometheus API to achieve replication and/or sharding. If you are looking for information about what to do with Prometheus or what it is, check out a previous post to get an overview. I am using remote write to have one Prometheus server running in Kubernetes which I have configured to remote write metrics to another Prometheus server on my network that has a larger disk and is outside the k8s cluster.

Elasticsearch

Elasticsearch is a distributed, RESTful search and analytics engine designed for scalability and flexibility. It allows you to store, search, and analyze large volumes of data quickly and efficiently. Elasticsearch is part of the Elastic Stack. This is a powerful platform that has many strengths when most of your data is text based and you may need to search for text. It supports horizontal scaling and stores data in an Index. You can have data replicated into shards to achieve High Availability.

Parsing workout data from Hevy on iOS

While I was writing the post about parsing data from the strong app on iOS, I stopped using this app. Now that I have this historical data there are a few approaches but as someone who is busy I ended up using another iOS app that has a native app for the apple watch. It also directly imported all my workout history so stay tuned as I will next be taking a look at any differences in that data.

Prometheus

Organizations rely heavily on metrics to ensure their applications are running smoothly, efficiently, and securely. One of the leading tools in this space is Prometheus , an open-source system designed for event monitoring and alerting. This blog post will delve into what Prometheus is, how it works with metrics, and why it’s become a staple in modern infrastructure management. It may be overkill for a homelab but it is actually a simple system that scales well. You can start by monitoring no more than a single system and/or application and scale to hundreds or more. Metrics are the core of any monitoring system. They provide quantitative data about various aspects of your systems and applications, such as performance, resource usage, error rates, and more. Prometheus collects these metrics from configured targets at specified intervals.