An Overview of Java Garbage Collectors

Transcript

Hi friends! Today, we are talking about Garbage Collectors in Java.

Choosing a right GC implementation may help you boost the app performance. HotSpot JVM offers seven collectors for various performance requirements. So, let’s look at all of them!

A couple of words about Java Memory Model

Young generation, old generation

Serial GC 

Serial GC is the simplest and the oldest collector in Java. It works in one thread and freezes all app threads while performing collection. It is good for single-threaded client-side applications that don’t need extra small pauses. But there’s a catch: if you limit your application to 2 GB RAM and less than 2 processors, SerialGC will be used automatically, even if you explicitly specify another collector! So be careful with that.


Parallel GC

Parallel GC is also called a throughput collector. It freezes all application threads, but works in multiple threads during the collection. By default, it works in multiple threads when it collects the young generation and the old generation. You can set the number of worker threads, pause time, and the time application spends on the collection.

Concurrent Mark Sweep GC

Concurrent Mark Sweep GC is designed for shorter pauses and uses multiple threads to perform the collection. But it is suitable for applications that can afford to share processor resources with the collector while the application is running. This collector was deprecated in Java 9 and removed in Java 14. 

G1 GC

G1 GC was designed to replace CMS GC. It is now a default garbage collector in Java. G1 GC tries to provide good balance between latency and throughput and is a very good fit for applications running in multiprocessor environments with a large heap of 10 gigabytes and more.

G1 means Garbage First. Just like other collectors, it works with young and old generations, but it splits the heap into multiple equal-sized regions. First, it performs global marking concurrently with the application and determines which objects are reachable. Then, it reclaims space in regions that are mostly filled with garbage, and so frees more memory, this is why it is called garbage first. G1 GC reclaims space using evacuation: it copies and compacts live objects from one or several regions into new regions.

Z GC

Z GC was first introduced in Java 11. It is a scalable low-latency collector, and it is great for latency-sensitive applications and applications with a very large heap of terabytes.

Z GC performs the expensive work concurrently and doesn’t stop the application threads for more than 10 ms. It requires minimal configuration, in most cases, just set the right heap size and you are good to go.

In Java 21, it became generational by default. It means that it maintains separate generations for young and old objects and collects young objects more frequently, which improves heap memory overhead and garbage collection CPU overhead.

Shenandoah GC

Shenandoah GC is also a low-latency collector introduced in Java 12. It performs most of its work concurrently with the app, including the concurrent compaction, so the pause times are not directly proportional to the heap size. Like Z GC, it is suitable for latency-sensitive applications and applications with a very large heap of terabytes. Oracle Java doesn’t ship Shenandoah, but this GC comes with major OpenJDK distributions including Liberica JDK.

Epsilon GC

Epsilon GC is a very interesting garbage collector. It doesn’t perform any garbage collection and only allocates memory. Why would you use a garbage collector that doesn’t collect any garbage? Well, it has a purpose. First, you can use it for performance testing to see how fast the program would run without any garbage collection and if there are performance bottlenecks not induced by the collector.

Also, you can use it with applications that almost don’t produce garbage. Or if the app is short-lived and doesn’t have time to leave garbage. Then the program may run faster without garbage collection cycles.

Another interesting feature of Epsilon GC is that it will remain an experimental feature forever so that you don’t accidentally use it in production. So, to use Epsilon GC, you need to explicitly enable experimental features.

Summary

Java provides multiple garbage collectors (GCs) tailored to different performance needs. Serial GC is ideal for single-threaded apps but pauses all threads, while Parallel GC uses multiple threads to prioritize throughput. G1 GC, the default collector, balances latency and throughput, making it suitable for large heaps, while Z GC and Shenandoah GC cater to low-latency and massive-heap applications by performing most tasks concurrently. Lastly, Epsilon GC skips garbage collection entirely, useful for testing or short-lived apps, but requires explicit enabling as an experimental feature.

About Catherine

Java developer passionate about Spring Boot. Writer. Developer Advocate at BellSoft

Social Media

Videos
card image
Feb 6, 2026
Backend Developer Roadmap 2026: What You Need to Know

Backend complexity keeps growing, and frameworks can't keep up. In 2026, knowing React or Django isn't enough. You need fundamentals that hold up when systems break, traffic spikes, or your architecture gets rewritten for the third time.I've been building production systems for 15 years. This roadmap covers three areas that separate people who know frameworks from people who can actually architect backend systems: data, architecture, and infrastructure. This is about how to think, not what tools to install.

Videos
card image
Jan 29, 2026
JDBC Connection Pools in Microservices. Why They Break Down (and What to Do Instead)

In this livestream, Catherine is joined by Rogerio Robetti, the founder of Open J Proxy, to discuss why traditional JDBC connection pools break down when teams migrate to microservices, and what is a more efficient and reliable approach to organizing database access with microservice architecture.

Further watching

Videos
card image
Feb 27, 2026
Spring Developer Roadmap 2026: What You Need to Know

Spring Boot is powerful. But knowing the framework isn’t the same as understanding backend engineering. In this video, I walk through the roadmap I believe matters for a Spring developer in 2026. We start with data. That means real SQL — CTEs, window functions, normalization trade-offs — and understanding what ACID and BASE actually imply for system guarantees. Spring Data JPA is useful, but you still need to know what happens underneath. Then architecture: microservices vs modular monolith, serverless, CQRS, and when HTTP, gRPC, Kafka, or WebSockets make sense. Not as buzzwords — but as design choices with trade-offs. Security and infrastructure follow: OWASP Top 10, AuthN vs AuthZ, encryption in transit and at rest, Docker, Kubernetes, Infrastructure as Code, and observability with Micrometer, OpenTelemetry, and Grafana. This roadmap isn’t about mastering every tool. It’s about knowing what affects reliability in production.

Videos
card image
Feb 18, 2026
Build Typed AI Agents in Java with Embabel

Most Java AI demos stop at prompt loops. That doesn't scale in production. In this video, we integrate Embabel into an existing Spring Boot application and build a multi-step, goal-driven agent for incident triage. Instead of manually orchestrating prompt → tool → prompt cycles, we define typed actions and let the agent plan across deterministic and LLM-powered steps. We parse structured input with Ollama, query MongoDB deterministically, classify risk using explicit thresholds, rank affected implants, generate a constrained root cause hypothesis, and produce a bounded containment plan. LLM handles reasoning. Java enforces rules. This is about controlled AI workflows on the JVM — not prompt glue code.

Videos
card image
Feb 12, 2026
Spring Data MongoDB: From Repositories to Aggregations

Spring Data MongoDB breaks down fast once CRUD meets production—real queries, actual data volumes, analytics. What looks simple at first quickly turns into unreadable repository methods, overfetching, and slow queries. In this video, I walk through building a production-style Spring Boot application using Spring Data MongoDB — starting with basic setup and repositories, then moving into indexing, projections, custom queries, and aggregation pipelines. You'll see how MongoDB's document model changes data design compared to SQL, when embedding helps, and when it becomes a liability. We cover where repository method naming stops scaling, how to use @Query safely, when to switch to MongoTemplate, and how to reduce payload size with projections and DTOs. Finally, we implement real MongoDB aggregations to calculate analytics directly in the database and test everything against a real MongoDB instance using Testcontainers. This is not another MongoDB overview. It's a practical guide to actually using Spring Data MongoDB in production without fighting the database.