Prev Next

Java / Java 17 Garbage Collection Interview Questions

Explain the internal working of G1 GC's concurrent marking cycle?

G1's concurrent marking cycle identifies which old regions hold the most garbage so later mixed collections know what to reclaim. It runs in several distinct phases, most of them alongside the running application:

flowchart LR A[Initial Mark - STW, piggybacked on a young GC] --> B[Root Region Scan - concurrent] B --> C[Concurrent Mark - concurrent, SATB-tracked] C --> D[Remark - STW, finalize marking + reference processing] D --> E[Cleanup - mostly concurrent, reclaim fully-garbage regions] E --> F[Mixed Collections - evacuate high-garbage old regions]

Initial Mark marks the direct roots and is piggybacked onto an already-scheduled young collection to avoid a separate pause. Root Region Scanning then concurrently scans survivor regions (as of initial mark) for references into the old generation. Concurrent Marking traces the full object graph from those roots while the application keeps running; to stay correct despite concurrent mutation, G1 uses a Snapshot-At-The-Beginning (SATB) write barrier that logs any reference a mutator thread is about to overwrite, ensuring no object that was reachable at the start of marking is missed.

Remark is a short stop-the-world pause that drains the last SATB logs and finishes reference processing (for soft/weak/phantom references). Cleanup then tallies live data per region and immediately frees any region found to be 100% garbage, without needing to evacuate anything from it - this feeds directly into the mixed-collection phase that follows.

The SATB write barrier used during G1's concurrent mark phase exists to:
Regions found to be 100% garbage during the Cleanup phase are:

More Related questions...

What is garbage collection in Java? What is the purpose of a garbage collector in the JVM? What are the types of garbage collectors available in Java 17? What is the default garbage collector in Java 17? What is the young generation in the JVM heap? What is the old (tenured) generation in the JVM heap? What is Metaspace in Java 17? What is a minor GC? What is a major GC? What is a full GC? Define a stop-the-world pause? What is the Eden space? What are Survivor spaces used for? How do you enable GC logging in Java 17? What is the purpose of the -Xmx and -Xms flags? Why is G1 GC the default collector since Java 9? How does G1 GC divide the heap into regions? What is a humongous object in G1 GC? How does G1 GC decide which regions to collect first? What is a remembered set in G1 GC? What is a card table and how does it help garbage collection? How do you configure G1 GC's pause time goal? What is the difference between G1 GC and Parallel GC? When should you choose ZGC over G1 GC? When should you choose Shenandoah over G1 GC? What happens during a G1 GC mixed collection? Why doesn't Java 17 include the CMS garbage collector? How is Parallel GC different from Serial GC? What is the difference between throughput and low-latency garbage collectors? How do you troubleshoot frequent full GCs in a Java 17 application? What is string deduplication in G1 GC? How does the JVM decide when to trigger a GC cycle? What is promotion in the context of generational garbage collection? Why should you avoid explicit calls to System.gc()? What is the difference between soft, weak, and phantom references? Explain the internal working of G1 GC's concurrent marking cycle? Explain the execution flow of a ZGC collection cycle? How does ZGC achieve sub-millisecond pause times using colored pointers? Explain the lifecycle of an object through the JVM heap under G1 GC? How does Shenandoah's Brooks pointer (forwarding pointer) enable concurrent compaction? What is an evacuation failure in G1 GC and how do you handle it? How does garbage collection behave differently in a containerized environment with cgroup memory limits? How can you optimize G1 GC for large heaps (32GB+)? Explain the tricolor marking algorithm used in concurrent garbage collectors? What are write barriers and load barriers, and how do collectors use them? How do you analyze a GC log to identify a memory leak? What is the difference between G1 GC's young-only phase and space-reclamation phase? How does adaptive sizing (ergonomics) work in the JVM's garbage collectors? Explain how class unloading interacts with garbage collection and Metaspace? Which is better for a low-latency trading application - ZGC or Shenandoah - and why?
Show more question and Answers...


Comments & Discussions