From mboxrd@z Thu Jan 1 00:00:00 1970 From: Roberto Spadim Subject: Re: high throughput storage server? Date: Thu, 17 Feb 2011 11:36:49 -0200 Message-ID: References: <4D5D017B.50109@anonymous.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: <4D5D017B.50109@anonymous.org.uk> Sender: linux-raid-owner@vger.kernel.org To: John Robinson Cc: Matt Garman , Mdadm List-Id: linux-raid.ids with more network cards =3D more network gbps with better (faster) rams =3D more disks reads with more raid0/4/5/6 =3D more speed on disks reads with more raid1 mirrors =3D more security with more sas/sata/raid controllers =3D more GB/TB on storage with more anything ~=3D more money just know what numbers you want and make it work 2011/2/17 John Robinson : > On 14/02/2011 23:59, Matt Garman wrote: > [...] >> >> The requirement is basically this: around 40 to 50 compute machines >> act as basically an ad-hoc scientific compute/simulation/analysis >> cluster. =A0These machines all need access to a shared 20 TB pool of >> storage. =A0Each compute machine has a gigabit network connection, a= nd >> it's possible that nearly every machine could simultaneously try to >> access a large (100 to 1000 MB) file in the storage pool. =A0In othe= r >> words, a 20 TB file store with bandwidth upwards of 50 Gbps. > > I'd recommend you analyse that requirement more closely. Yes, you hav= e 50 > compute machines with GigE connections so it's possible they could al= l > demand data from the file store at once, but in actual use, would the= y? > > For example, if these machines were each to demand a 100MB file, how = long > would they spend computing their results from it? If it's only 1 seco= nd, > then you would indeed need an aggregate bandwidth of 50Gbps[1]. If it= 's 20 > seconds processing, your filer only needs an aggregate bandwidth of 2= =2E5Gbps. > > So I'd recommend you work out first how much data the compute machine= s can > actually chew through and work up from there, rather than what their = network > connections could stream through and work down. > > Cheers, > > John. > > [1] I'm assuming the compute nodes are fetching the data for the next > compute cycle while they're working on this one; if they're not you'r= e > likely making unnecessary demands on your filer while leaving your co= mpute > nodes idle. > > -- > To unsubscribe from this list: send the line "unsubscribe linux-raid"= in > the body of a message to majordomo@vger.kernel.org > More majordomo info at =A0http://vger.kernel.org/majordomo-info.html > --=20 Roberto Spadim Spadim Technology / SPAEmpresarial -- To unsubscribe from this list: send the line "unsubscribe linux-raid" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html