apex-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Thomas Weise <tho...@datatorrent.com>
Subject Re: Window Commits
Date Tue, 25 Aug 2015 15:51:04 GMT
BTW You don't need to delay anything because of migration. Any changes that
get committed for 3.2.0-SNAPSHOT will all be pulled over automatically when
the repository access is in place.

On Tue, Aug 25, 2015 at 8:24 AM, Atri Sharma <atri.jiit@gmail.com> wrote:

> I will fork the branch and try, thanks.
> On 25 Aug 2015 10:39, "Vlad Rozov" <v.rozov@datatorrent.com> wrote:
>
> > Please take a look at my SPOI-5823 branch. I am almost done with the
> > enhancement and just finishing up some synchronization code. Note that
> the
> > branch is for review only and I am waiting for code migration to Apache
> to
> > commit my changes.
> >
> > Vlad
> >
> > On 8/24/15 18:25, Atri Sharma wrote:
> >
> >> Sound pretty cool.
> >>
> >> I have half baked patch for the thing I talked about. I think I can help
> >> with your patch and/or extend it to something of what I have in mind.
> >>
> >> Could you please point me to the right repo?
> >> On 25 Aug 2015 06:33, "Vlad Rozov" <v.rozov@datatorrent.com> wrote:
> >>
> >> To add to Chetan comment, I am working on an enhancement that offloads
> >>> bufferserver spooling to disk to a separate thread. This thread will
> kick
> >>> in when only half (should it be configurable?) of memory blocks are
> left
> >>> available. This should further avoid possibility of I/O spikes.
> >>>
> >>> Vlad
> >>>
> >>> On 8/24/15 14:16, Chetan Narsude wrote:
> >>>
> >>> The bufferserver writes pages to disk *only when* it runs out of memory
> >>>> to
> >>>> hold them.
> >>>>
> >>>> Can you elaborate where you see I/O spikes?
> >>>>
> >>>> --
> >>>> Chetan
> >>>>
> >>>> On Mon, Aug 24, 2015 at 12:39 PM, Atri Sharma <atri.jiit@gmail.com>
> >>>> wrote:
> >>>>
> >>>> Folks,
> >>>>
> >>>>> I was wondering if it makes sense to have a functionality in which
> >>>>> bufferserver writes out data pages to disk in batches defined by
> >>>>> timeslice/application window.
> >>>>>
> >>>>> This will allow flexible workloads and reduce I/O spikes (I
> understand
> >>>>> that
> >>>>> we have non-blocking I/O but it still would incur disk head costs).
> >>>>>
> >>>>> Thoughts?
> >>>>>
> >>>>> --
> >>>>> Regards,
> >>>>>
> >>>>> Atri
> >>>>> *l'apprenant*
> >>>>>
> >>>>>
> >>>>>
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message