Grafana Loki The road to higher scales and lower latencies for logs

[@UmeshKumar-jf4yd]

is there a document on resource benchmarking of Loki components to achieve 2TB/h of ingestion and 80GB/s query throughput? I want to ingest 2TB/day, I am not sure how many distributor, ingestor and queriers I need and how much CPU and RAM I must assign to each components.

[@diguacu4131]

Is there have a grafana dashboard template for loki like this video showed at time 09:02

[@nupursingh5021]

How Loki can retain logs? if loki pod gets restarted. i have tested it and it lost all logs

[@UmeshKumar-jf4yd]

You must run ingesters as statefulset. If you are running Loki in microservices mode then you can configure it with a local or remote storage server to store logs and indices. The supported backend stores are s3, azure storage account or GCS bucket.