On Fri, 19 Feb 2021 14:36:00 GMT, Lin Zang <lz...@openjdk.org> wrote:
>> 8252842: Extend jmap to support parallel heap dump
>
> Lin Zang has updated the pull request incrementally with one additional
> commit since the last revision:
>
> fix build fail issue on windows
Hi,
I've benchmarked the code on my machine (128GB memory, 56 logical CPUs) with an
example creating a 32 GB heap dump. I only saw a 10 percent reduction in time,
both using uncompressed and compressed dumps. Have you seen better numbers in
your benchmarks?
And it seems to potentially use a lot more temporary memory. In my example I
had a 4 GB array in the heap and the new code allocated 4 GB of additional
memory to write this array. This could happen in more threads in parallel,
increasing the memory consumption even more.
If the above problems could be fixed, I would suggest to just use the parallel
code in all cases.
src/hotspot/share/services/heapDumper.cpp line 383:
> 381: };
> 382:
> 383: // Supports I/O ooperations for a dump
ooperations->operations
src/hotspot/share/services/heapDumper.cpp line 650:
> 648: void enqueue(ParWriterBufferQueueElem* entry) {
> 649: if (_head == NULL) {
> 650: assert(is_empty() && _tail == NULL, "Sanity check");
When I have run the code with assertions enabled, this one failed since _tail
was not NULL.
-------------
PR: https://git.openjdk.java.net/jdk/pull/2261
