[159667] in North American Network Operators' Group
Re: Netflow Nfsen Server Hardware
daemon@ATHENA.MIT.EDU (PC)
Thu Jan 17 11:07:14 2013
In-Reply-To: <CAL9jLaZjUTurxBo9KUf=h=suNWsKuEndCrUD-XDwJWXmjeCPgg@mail.gmail.com>
Date: Thu, 17 Jan 2013 09:04:33 -0700
From: PC <paul4004@gmail.com>
To: Christopher Morrow <morrowc.lists@gmail.com>
Cc: "nanog@nanog.org" <nanog@nanog.org>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org
I agree here with Christopher; A SSD to handle the high IOPS requirements
of real time data logging; combined with a scheduled transfer which can
"move" the stored data in a linear large block copy operation to ordinary
spindles, would be a cost effective hybrid solution.
This of course is assuming the application can handle this separation of
data; and I know nothing about Nfsen
On Thu, Jan 17, 2013 at 9:01 AM, Christopher Morrow <morrowc.lists@gmail.com
> wrote:
> On Thu, Jan 17, 2013 at 9:05 AM, Joe Loiacono <jloiacon@csc.com> wrote:
> > Tim Calvin <tcalvin@tlsn.net> wrote on 01/16/2013 05:51:11 PM:
> >
> >> PowerEdge R610 -
> >>
> >> 2x Intel E5540, 2.53GHz Quad Core Processor
> >>
> >> 32GB RAM
> >>
> >> 2x 300gb 10k 2.5" SAS HDD
> >
> > Since netflow processing is generally I/O bound, you may want to invest
> in
> > 15K drives.
>
> I had suggested off-list that perhaps primary storage as SSD was a
> better path, is there a reason to not do that? (with some larger
> storage on spinning-media for historical storage/query).
>
>
