[CentOS] really large file systems with centos

Fri Jul 15 00:00:36 UTC 2011
John R Pierce <pierce at hogranch.com>

On 07/14/11 2:32 AM, Joseph L. Casale wrote:
> If I understand, it's not your only backup system, so I don't think it's that
> critical, but the rebuild time on each array versus the degraded IO
> capacity and its impact on serving content would be something interesting.
>
> Do you plan on making hotspares available? That many discs will likely
> have a higher rate of failure...

planning on about 5-10% hot spares in the array.   For example, my 
suggested layout for 192 disks is...

18 x 10 drive raid6, with 12 hot spares.   The 18 seperate raid sets 
would be merged into larger volume groups, probably as volume sets and 
NOT stripes so the idle disks could spin down, since most of this data, 
once written, will rarely if ever be looked at again.   With 3TB disks, 
this gives 18 x 8 x 3 TB == 432TB total usable capacity (really more 
like 400TB when you figure binary vs decimal, etc etc)

> What kind of discs and controllers do you intend on using?

Seagate Constellation ES.2 3TB SAS drives, in (one of a couple major 
label vendors) SAS chassis, with (major vendors' rebranded) LSI SAS 
HBA's (configured for JBOD).   Using SAS disks rather than SATA to gain 
the path redundancy from dual porting.

-- 
john r pierce                            N 37, W 122
santa cruz ca                         mid-left coast