WeakOrderQueue finalizer needed?

[Nat]

I wonder why Recycler$WeakOrderQueue has a finalize() method. Is it possible to make it non-finalizable object (probably by using PhantomReference)? The reason I asked because I was trying to debug a long GC pause and just noticed that there are a lot of WeakOrderQueue in the finalizer queue. (Note that so far I do not have a good evidence that this causes the problem yet. I just noticed that we have relatively high "Ref Proc" time even though ParallelRefProcEnabled is already turned on.

[Norman Maurer]

Can you open an issue in our issue tracker for this ? I will need to investigate a bit and want to ensure it not get forgotten before I find the time

Thanks

norman

On 25. Oct 2017, at 00:30, Nat <nat....@gmail.com> wrote:

I wonder why Recycler$WeakOrderQueue has a finalize() method. Is it possible to make it non-finalizable object (probably by using PhantomReference)? The reason I asked because I was trying to debug a long GC pause and just noticed that there are a lot of WeakOrderQueue in the finalizer queue. (Note that so far I do not have a good evidence that this causes the problem yet. I just noticed that we have relatively high "Ref Proc" time even though ParallelRefProcEnabled is already turned on.

--
You received this message because you are subscribed to the Google Groups "Netty discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to netty+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/netty/c6db400f-9e54-40cc-b040-f5c966035fc7%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Nat]

Done. https://github.com/netty/netty/issues/7343

[Norman Maurer]

Thanks

Am 25.10.2017 um 00:30 schrieb Nat <nat....@gmail.com>:

I wonder why Recycler$WeakOrderQueue has a finalize() method. Is it possible to make it non-finalizable object (probably by using PhantomReference)? The reason I asked because I was trying to debug a long GC pause and just noticed that there are a lot of WeakOrderQueue in the finalizer queue. (Note that so far I do not have a good evidence that this causes the problem yet. I just noticed that we have relatively high "Ref Proc" time even though ParallelRefProcEnabled is already turned on.

--