[Lustre-discuss] MDS server and storage sizing

Klaus Steden klaus.steden at thomson.net
Fri Nov 2 13:52:31 PDT 2007


The docs recommend an MDS volume around 1-2% of the size of your total data
volume.

There¹s a lot of useful information in the manual about sizing, etc., so
you¹re probably better off reading them to see what suits your environment ‹
but the 1-2% rule of thumb nonetheless applies.

The key element if you¹re setting up an active/passive MDS pair is that you
need to be able to see the same device from both MDS nodes; the active one
will mount it, the passive one will wait until it¹s needed (i.e. the active
MDS encounters a failure).

It¹s essential that they not mount it simultaneously, though, or wicked
corruption is likely to occur.

hth,
Klaus

On 11/2/07 1:45 PM, "Mertol Ozyoney" <Mertol.Ozyoney at Sun.COM>did etch on
stone tablets:

> Hi All ;
>  
> What kind of server and storage would you suggest for a 600 TB (Net) lustre
> system?
>  
> I plan on using a 2 or 4 socket Quad Core Server with a lot of RAM (active
> passive cluster) with 24 x 146 GB 15k RPM drives. (Around 2 TB usable capacity
> ) 
>  
>  
>  
>  
>    <http://www.sun.com/>    Mertol Ozyoney
>  Storage Practice - Sales Manager
>  
>  Sun Microsystems, TR
>  Istanbul TR
>  Phone +902123352200
>  Mobile +905339310752
>  Fax +902123352222
>  Email mertol.ozyoney at Sun.COM <mailto:Ayca.Yalcin at Sun.COM>
>  
>  
> 
> 
> _______________________________________________
> Lustre-discuss mailing list
> Lustre-discuss at clusterfs.com
> https://mail.clusterfs.com/mailman/listinfo/lustre-discuss


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lustre.org/pipermail/lustre-discuss-lustre.org/attachments/20071102/f4625dac/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/gif
Size: 1257 bytes
Desc: not available
URL: <http://lists.lustre.org/pipermail/lustre-discuss-lustre.org/attachments/20071102/f4625dac/attachment.gif>


More information about the lustre-discuss mailing list