This talk will deep dive into Gojek's journey of addressing the challenge of solving database reliability with Patroni, which is a template for high availability (HA) PostgreSQL solution written in Python.
As Gojek operates at country scale and the app's reliability is paramount for the business, we were troubled with the reliability of our databases by node failures causing systems to go down. Patroni came in as a savior for solving these problems by bringing in redundancy and maintaining 5 9s of uptime.
We'll discuss about how Patroni works behind the scenes and manages a highly available and distributed PostgreSQL cluster along with discussing how automatic failover works with using a distributed consensus for leader election.
We'l also ponder over the experiences and lessons learnt from running it on production and managing around ~400 clusters.
What format do you have in mind for your talk?
Talk
Table of contents
The general microservices and database architecture of Gojek services.
How running a master-slave Postgres was causing reliability issues
How Patroni solves the database reliability problem
How Patroni works behind the scenes - automatic failover, leader election, cluster management
How is Gojek managing Patroni on Production for ~400 clusters - Experiences and lessons
What domain would you say your talk falls under?
CLI & Tooling
Duration in minutes (including Q&A)
40
Prerequisites
Just a basic idea around distributed systems and databases.
Speaker bio
I'm Abhijeet, I usually dabble around DevOps and Infrastructure Automation, and a avid Ruby and Python enthusiast.
I've working at Gojek for the past 4 years solving problems around creating Developer and Cloud Platforms, making dev life's easier(however they might not agree on this xD).
Apart from work, I am a home gym fanatic and amateur musician usually found fiddling with my guitar in a random corner of Cubbon Park.
The talk/workshop speaker agrees to
[X] Share the slides, code snippets and other material used during the talk
Title of the talk
How Patroni solved database reliability at Gojek
Description
This talk will deep dive into Gojek's journey of addressing the challenge of solving database reliability with Patroni, which is a template for high availability (HA) PostgreSQL solution written in Python.
As Gojek operates at country scale and the app's reliability is paramount for the business, we were troubled with the reliability of our databases by node failures causing systems to go down. Patroni came in as a savior for solving these problems by bringing in redundancy and maintaining 5 9s of uptime.
We'll discuss about how Patroni works behind the scenes and manages a highly available and distributed PostgreSQL cluster along with discussing how automatic failover works with using a distributed consensus for leader election.
We'l also ponder over the experiences and lessons learnt from running it on production and managing around ~400 clusters.
What format do you have in mind for your talk?
Talk
Table of contents
What domain would you say your talk falls under?
CLI & Tooling
Duration in minutes (including Q&A)
40
Prerequisites
Just a basic idea around distributed systems and databases.
Speaker bio
I'm Abhijeet, I usually dabble around DevOps and Infrastructure Automation, and a avid Ruby and Python enthusiast.
I've working at Gojek for the past 4 years solving problems around creating Developer and Cloud Platforms, making dev life's easier(however they might not agree on this xD).
Apart from work, I am a home gym fanatic and amateur musician usually found fiddling with my guitar in a random corner of Cubbon Park.
The talk/workshop speaker agrees to
[X] Share the slides, code snippets and other material used during the talk
[X] If the talk is recorded, you grant the permission to release the video on BangPyper's YouTube channel under CC-BY-4.0 license
[X] Not do any hiring pitches during the talk and follow the Code of Conduct