Start of main content

Transactions and locks in high-load microservice applications


The talk is devoted to the use of transactions and locks, including distributed ones, when processing database data in the conditions of multi-pod architecture and asynchronous calls of high-loaded application. 

There will be some livecoding: the speaker will show possible mistakes and nuances of using certain technologies. Stack: Java, Spring Data JPA, Kafka. 

Sometimes developers do not think when a transaction is needed, and when not, when to open a new one, and when to continue the current one. What locks on the database will occur during the execution of the transaction and how not to overload the database with row lock queue, where to use timeouts of requests. And how to avoid database locks. 

Target audience: backend developers, since most services in some way or another work with dynamically updated data.

  • #jpa
  • #spring-data
  • #kafka


Invited experts