📖
Grafana Loki Deep Dive
  • Grafana Loki Deep Dive
  • What is Grafana Loki?
  • Ingestion Process
    • Posting logs to Distributors
    • Routing logs to Ingesters
    • Buffering log chunks on memory in Ingesters
    • Flushing memory chunks from Ingesters
    • Write-Ahead Log
    • Unordered Write
  • Query Process
    • Query splitting
    • Query scheduling
    • Query for all of ingesters
    • Select the chunks by inverted index
    • Filter the matched chunks
    • Aggregate the results from queriers
  • What is BoltDB Shipper?
  • Cache Strategy
  • Design for failure
  • Performance Improvement
  • Monitoring
  • Clustering
  • Tips
    • Failure stories
    • How to retain data in each component
    • Stateless or Stateful?
  • Contribution
Powered by GitBook
On this page

What is this book?

Grafana Loki is an awesome log monitoring platform provided as OSS.

However, it is hard to run in production because it is constructed with many microservices and we can't know how they work easily.

I aim to complement the official document and make the readers plan capacity, select parameters, and run Loki stably by themselves to provide this book.

This book shows you the following knowledge.

You can read only the part which you want to know.

  • What's Loki?

  • Dive into ingestion process

  • Dive into query process

  • Cache Strategy

  • Design for failure

  • Improve query performance

  • Monitoring for Loki

  • Some other tips

I hope that everyone enjoys "Loki Life"!

Who should read this book?

I expect the readers as follows.

  • Now considering introducing Loki

  • Want more detailed information than the official document

  • Be curious about distributed system

Expected set up in this book

  • Loki version: 2.4.1

  • Chunk Storage: AWS S3

  • Index Storage: BoltDB Shipper + AWS S3

  • Deployment mode: Microservices

Last updated 3 years ago

  • What is this book?
  • Who should read this book?
  • Expected set up in this book
NextWhat is Grafana Loki?

Grafana Loki Deep Dive