Rootconf 2015

DevOps and scaling infrastructure

Logstash, statsd and graphite combo for PostgreSQL log monitoring

Submitted by Himanchali (@himanchali) on Tuesday, 31 March 2015

videocam
Preview video

Technical level

Beginner

Section

Full talk

Status

Submitted

Vote on this proposal

Login to vote

Total votes:  +17

Objective

The session (full length talk of 40 mins) will aim to explain the use case of log monitoring done for Postgres DB server with the help of logstash, statsd and graphite. The same combination can be used for other logs as well.

Description

Mindmap Link for the session: http://i.imgur.com/cYti1E1.jpg

When you have a system where the DB server is generating 1GB log per 10 minutes, monitoring those logs for different kind of performance and security analysis is a big challenge.
The session will cover the monitoring setup for db server logs using the 3 opensource tools and how these graphs can help in analysis improvement.
The key points covered:

  • Introduction of the used tools

  • The basic setup of these tools

  • The basic conf setup

  • Proper aggregation of metrics

  • Issues faced during setup and their solution

  • Final performance

  • Kind of metrics generated and their usage

Speaker bio

I am currently working as Tech lead Production & Infra Engg @ Inmobi. Other than being part of DBE team, has been involved in evaluating and implementing open source tools for production environment.

Links

Preview video

https://youtu.be/OnRcwQfun-4

Comments

Login with Twitter or Google to leave a comment