Improve Write-Master and Read Replica scale, safely without application changes
Roland Lee's background is in application networking. He has been a developer and product manager in startups and large companies. All through his career, his industry focus has been to increase the performance of Internet user data. He is currenly Head of Products for Heimdall Data, an off-the-shelf data access layer for applications accessing SQL databases.
No video of the event yet, sorry!
Horizontally scaling your SQL database involves separating the write-master from read-only servers. However, writing to one node and reading from another can result in inconsistent data due to synchronization delays.
Database proxies intelligently route SQL queries to the best data source for even load distribution, while guaranteeing consistency.
In this session, we will provide a hands-on demo of how read/write split works alongside query caching. We will show how this solution met the dynamic traffic demands of our AWS customer, CyberAuctions.com during high peak time load.
- 20 min
- Postgres Conference 2020
- Ops and Administration