activemq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Michael André Pearce <michael.andre.pea...@me.com>
Subject Re: Paging
Date Fri, 24 Mar 2017 03:24:50 GMT
Going off Francesca's results when implementing memory mapped for journal :
https://issues.apache.org/jira/browse/ARTEMIS-906

Running mapped with no data sync will be of orders of magnitude on the write path. 



Sent from my iPhone

> On 24 Mar 2017, at 01:53, Clebert Suconic <clebert.suconic@gmail.com> wrote:
> 
> On Thu, Mar 23, 2017 at 9:38 PM Michael André Pearce <
> michael.andre.pearce@me.com> wrote:
> 
>> MemoryMapped for paging was actually the one I was more thinking to be
>> beneficial.
>> 
>> Afaik paging is rebuildable of server crash as essentially just memory
>> spill from the heap?
>> 
>> What I was thinking with memory mapped could give improved io at the
>> expense of guaranteed disk persistence in server crash. But for paging I
>> believe (please correct this assumption) this would be ok?
> 
> 
> 
> It's the same as NIO actually in terms of guarantees.
> 
> But it should save CPU on kernel calls.
> 
> 
> 
> It will be awesome when people start usin new devices that are optimized
> around mmap.
> 
>> 
>> 
>> 
>> 
>> Sent from my iPhone
>> 
>> On 24 Mar 2017, at 01:27, Clebert Suconic <clebert.suconic@gmail.com>
>> wrote:
>> 
>>>> It would make sense to implement Journal using Mappedfile....
>>> 
>>> 
>>> I meant Paging of course.
>> 
> -- 
> Clebert Suconic

Mime
  • Unnamed multipart/alternative (inline, 7-Bit, 0 bytes)
View raw message