There are many great use cases of Thanos, we collect them here for others to refer to.
Case add by @RayHuangCN
If your want to build a horizontal scaleable Prometheus for large kubernetes cluster monitoring, you can use Thanos + Kvass.
Kvass is not anther Prometheus compatible TSDB but a Prometheus sharding solution, which uses Sidecar to generate special config file only contains part of targets assigned from Coordinator for every Prometheus shard. Thanos is used for global data view, rules processing and long term storage capabilities. your can see usage detail docs.
Thanos + Kvass is working well to scrape basic metrics of cluster (kubelet, cadvisor, kube-state-metrics and node-exporter…) with following cluster size. Just use one Prometheus config file, without any federation
or hashmod
.