Friendship ended with the garbage collector

Re:Friendship ended with the garbage collector by secondaryacct on Friday August 27, 2021 @06:14AM (#28326026)

That s what we do in low latency java in big banks, 3 GC per day, between 5 and 6 am, anything more and 10 people are profiling.

[ Parent | on HN ]

Re:Friendship ended with the garbage collector by kaba0 on Friday August 27, 2021 @02:06PM (#28331101)

Is that truly a good idea in case of the JVM? Of course there will be problems where this is the correct approach, but in certain cases simply letting the very advanced GC do its job is actually faster.

[ Parent | on HN ]

Re:Friendship ended with the garbage collector by bobbylarrybobby on Friday August 27, 2021 @06:00PM (#28333424)

It’s not about speed (throughout) but about latency. If the GC takes (say) several minutes during a period of downtime so that it doesn’t have to run at all during uptime, then it’s worth it.

[ Parent | on HN ]

Re:Friendship ended with the garbage collector by kaba0 on Friday August 27, 2021 @07:18PM (#28334052)

What about ZGC then? If throughput is not the metric to optimize for, ZGC promises less than 1 ms pauses for up to TBs of heap size.

[ Parent | on HN ]

Re:Friendship ended with the garbage collector by pjmlp on Saturday August 28, 2021 @09:37AM (#28337932)

Java suffers from Python 2 like syndrome with Java 8.

Those shops aren't going to use ZGC.

Nor they will profit from the several performance improvements, including the free JIT cache across runs with PGO data on Hotspot and J9, that used to be a commercial feature on third party JDKs.

[ Parent | on HN ]

Re:Friendship ended with the garbage collector by nextaccountic on Saturday August 28, 2021 @07:51AM (#28337346)

What about a concurrent and parallel GC that doesn't stop the world?

[ Parent | on HN ]