Ask HN: What metrics stack are you using for side projects
After having worked with a nice metrics stack in my worklife. I feel kind of blind in my side projects. What are all y'all using for metrics and visualization for your side projects? I'd prefer a SaaS solution, and I'd love to be able to (easily): - Publish (time-series) metrics from my code - Create dashboards visualizing these metrics - (Optional) setup alerting on these metrics
The stacks I've been able to find are either very tied to Prometheus, or tracing. I would like to emit more metrics semantically more close to the application. Examples: - Backlog size - Concurrent items being processed - Number of items in a specific state
So what are you running in your projects, where it would be overkill to run your own prometheus/grafana/influx/whatevs stack? I've been using Grafana Cloud (free tier) for over a year now. Still very happy with it. I did a comparison with Influx Cloud: https://pqvst.com/2021/06/22/hosted-monitoring-evaluating-in... I also recently started using Logtail for one of my projects. They also have a good free-tier and it comes with a Grafana instance embedded, so that you can visualize metrics computed from your logs. Seems quite useful but haven't had a chance to really test it fully yet. The kube-prometheus-stack (helm chart) is pretty cheap to maintain. I have that plus selfhosting Loki.
If you are looking for a similar setup, there’s Grafana Cloud with a generous free tier, but I haven’t tried that.