400% performance improvement after tuning these PostgreSQL parameters
Avinash Vallarapu (Avi) is currently working for Percona as a PostgreSQL Support Tech Lead. Before joining Percona, he worked for OpenSCG for 2.2 years as a Database Architect and for Dell as a Tech Lead for 9.8 Years. He has a vast experience in technologies like Oracle, PostgreSQL, MySQL and MongoDB. He is an avid Python and Golang developer. He has co-authored a book on PostgreSQL : Beginning PostgreSQL on Cloud and another book on PostgreSQL is in progress.
He spoke at various PostgreSQL Conferences in the past including -
- PGCONF India - 2017/2019
- PGCON Ottawa 2018
- Percona Live Frankfurt 2018
- Postgres Conference New York 2019
- Percona Live Texas 2019
- Postgres Open Florida 2019
- Postgres Conference Silicon Valley 2019
- Percona Live Amsterdam 2019
- Upcoming PGCONF India 2020, New York, Percona Live and some more conferences.
His areas of expertise are PostgreSQL Training, Consulting and Migrations.
No video of the event yet, sorry!
*Abstract : *
In the proposal, I would like to share how we have optimized a few database parameters that did outburst the performance by several times. It was a PostgreSQL Server of size 325 GB and more than 1500 TPS. After our tuning, we were able to get the database process upto 4900 TPS without compromising on ACID or availability.
** The talk includes : **
- The environment where we have seen disastrous performance.
- The parameters we concentrated for tuning and why ?
- The way we analyzed the active data set and optimized many more parameters in run time
- Answers to questions like :
- Does my active data set fits in memory ?
- Should I increase work_mem ?
- Is WAL compression working to my advantage ?
- How far should I have checkpoints spread ?
- By the end of this talk, you should know what parameters you should tune in PostgreSQL and when.
- 2019 March 22 14:00 EDT
- 50 min
- Riverside Suite
- Postgres Conference
- Use Cases