~/talk

Comparing Java startup performance options with Quarkus

Talkat JCON EUROPE 2025 · 2025-05-12..05-15· [java, web-performance]
links:none on file
slides:none on file

Compares concrete options: HotSpot vs OpenJ9, AppCDS and JDK24 AOT class linking, OpenJ9 shared classes cache, CRaC checkpoint/restore, remote JIT (JIT server/Seamru), and GraalVM native images. JVM cold start itself is ~70 ms; Quarkus reports ~2.507 s; container startup adds ~500–600 ms (up to 1.2 s). OpenJ9 cuts time-to-first-response ~30% and uses ~half the steady‑state memory of HotSpot but can spike at JIT time; JIT server removes those spikes; native images start fastest but add multi‑minute builds and occasional runtime class discovery failures; CRaC restores in a few hundred ms and matches native startup without rebuilding code.

AI-written summary · about

  1. 0:00Introduction and speaker intro
  2. 1:25What fast Java startup means
  3. 18:52Java startup tech options overview
  4. 22:42Pure Java JDK options
  5. 25:09Hotspot class sharing
  6. 27:56Hotspot AOT class linking
  7. 30:25OpenJ9 shared classes
  8. 34:19Checkpoint restore with OpenJ9
  9. 38:30Remote JIT compilation
  10. 40:56Native image with Quarkus
  11. 45:12Choosing the right option
  12. 47:43Future improvements and Q&A
~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.