Thanks, Michael.
> 2021年4月2日 下午10:22，Michael McCandless <[email protected]
> <mailto:[email protected]>> 写道：
>
> Hmm I don't think the original selection of this size was due to CPU L1 cache
> limits. It was more a tradeoff of not wasting too much RAM for tiny indexing
> events (e.g. index one small document and then refresh) while still getting
> benefits of scale for the more common/important high indexing rate use-cases.
>
> Maybe try changing it and see if things get faster/slower for your usage?
>
> Mike McCandless
>
> http://blog.mikemccandless.com <http://blog.mikemccandless.com/>
>
> On Fri, Apr 2, 2021 at 2:05 AM LuXugang <[email protected]
> <mailto:[email protected]>> wrote:
> Hi,
>
> In class ByteBlockPool , per buffer's default maximum size was set 32KB（2 <<
> 15）, is this such choice relevant with CPU L1 cache which cache size is 32KB？
>
> If not so, could anyone give some tips？
>
> <ByteBlockPool.png>
