<div dir="ltr"><div><div><div><div><div>Hi Scott, <br><br>The 3 - SAN Storages with 240 disks each has its own 3 NAS Headers (NAS Appliances).<br></div>However, even with 240 10K RPM disk and RAID50, it is only providing around 1.2 - 1.4GB/s per NAS Header.<br><br></div><div>There is no clustered file system, and each NAS Header has its own file-system.<br></div>It uses some custom mechanism to present the 3 file systems as single name space.<br></div>But the directories have to be manually spread across for load-balancing.<br></div>As you can guess, this doesn't work most of the time.<br></div><div>Many a times, most of the compute nodes access a single NAS Header, overloading it.<br></div><div><br></div>The customer wants *at least* 9GB/s throughput from a single file-system.<br><div><div><br><div><div>But I think, if we architect the Lustre Storage correctly, with these many disks, we should get at least 18GB/s throughput, if not more.<br></div><div><br></div><div>Regards,<br><br><br></div><div>Indivar Nair<br></div><div><br></div></div></div></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Tue, Jul 21, 2015 at 10:15 PM, Scott Nolin <span dir="ltr"><<a href="mailto:scott.nolin@ssec.wisc.edu" target="_blank">scott.nolin@ssec.wisc.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">An important question is what performance do they have now, and what do they expect if converting it to Lustre. Our more basically, what are they looking for in general in changing?<br>
<br>
The performance requirements may help drive your OSS numbers for example, or interconnect, and all kinds of stuff.<br>
<br>
Also I don't have a lot of experience with NFS/CIFS gateways, but that is perhaps it's own topic and may need some close attention.<br>
<br>
Scott<span class=""><br>
<br>
On 7/21/2015 10:57 AM, Indivar Nair wrote:<br>
</span><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">
Hi ...,<br>
<br>
One of our customers has a 3 x 240 Disk SAN Storage Array and would like<br>
to convert it to Lustre.<br>
<br>
They have around 150 Workstations and around 200 Compute (Render) nodes.<br>
The File Sizes they generally work with are -<br>
1 to 1.5 million files (images) of 10-20MB in size.<br>
And a few thousand files of 500-1000MB in size.<br>
<br>
Almost 50% of the infra is on MS Windows or Apple MACs<br>
<br>
I was thinking of the following configuration -<br>
1 MDS<br>
1 Failover MDS<br>
3 OSS (failover to each other)<br>
3 NFS+CIFS Gateway Servers<br>
FDR Infiniband backend network (to connect the Gateways to Lustre)<br>
Each Gateway Server will have 8 x 10GbE Frontend Network (connecting the<br>
clients)<br>
<br></span>
*Option A*<span class=""><br>
10+10 Disk RAID60 Array with 64KB Chunk Size i.e. 1MB Stripe Width<br>
720 Disks / (10+10) = 36 Arrays.<br>
12 OSTs per OSS<br>
18 OSTs per OSS in case of Failover<br>
<br></span>
*Option B*<span class=""><br>
10+10+10+10 Disk RAID60 Array with 128KB Chunk Size i.e. 4MB Stripe<br>
Width<br>
720 Disks / (10+10+10+10) = 18 Arrays<br>
6 OSTs per OSS<br>
9 OSTs per OSS in case of Failover<br>
4MB RPC and I/O<br>
<br></span>
*Questions*<span class=""><br>
1. Would it be better to let Lustre do most of the striping / file<br>
distribution (as in Option A) OR would it be better to let the RAID<br>
Controllers do it (as in Option B)<br>
<br>
2. Will Option B allow us to have lesser CPU/RAM than Option A?<br>
<br>
Regards,<br>
<br>
<br>
Indivar Nair<br>
<br>
<br>
<br></span>
_______________________________________________<br>
HPDD-discuss mailing list<br>
<a href="mailto:HPDD-discuss@lists.01.org" target="_blank">HPDD-discuss@lists.01.org</a><br>
<a href="https://lists.01.org/mailman/listinfo/hpdd-discuss" rel="noreferrer" target="_blank">https://lists.01.org/mailman/listinfo/hpdd-discuss</a><br>
<br>
</blockquote>
<br>
<br>
<br>_______________________________________________<br>
lustre-discuss mailing list<br>
<a href="mailto:lustre-discuss@lists.lustre.org">lustre-discuss@lists.lustre.org</a><br>
<a href="http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org" rel="noreferrer" target="_blank">http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org</a><br>
<br></blockquote></div><br></div>