New Logging: Loki

Modern logging

Logging is crucial - and daunting to maintain, if done the old way: A big ElasticSearch Cluster, which is sadly quite heavy on every aspect - CPU, RAM, disk I/O and maintenance. Read: It’s quite expensive! There must be a better way, and as we found out, the folks from Grafana where already on it. Loki is promising us to be:

[…] a horizontally scalable, highly available, multi-tenant log aggregation system inspired by Prometheus. It is designed to be very cost effective and easy to operate. It does not index the contents of the logs, but rather a set of labels for each log stream.

Sure, developers as ops needs to get used to it. Manfred Ackermann has done a great job to learn and teach the “new way of logging”. Have a look at e.g. the ClusterEvents log pushed by an exporter to our Loki: Loki Logs: Cluster Events