[OpenAFS] OpenAFS in the ISP environment. A good idea...

[Leland J. Steinke]

Derrick J Brashear wrote:

>>When does the number of volumes per 
>>partition become problematic.  Is it a function (as you imply above) of 
>>partition size, volume size (and distribution thereof), and number of volumes in 
>>the partition?
> 
> 
> It becomes problematic when you need to vacate the disk because it's
> dying.


That's why we want to set it up on a RAID array, so individual disks can die and 
we can hot-swap them.  Heaven forbid two drives die at once!

> 
> Also, I should note, you can do a parallel salvage on partitions. If said
> does not become I/O bound, partitioning instead of "one big chunk" can be
> a win. I don't know offhand if it becomes I/O bound.
> 
As for the IO-boundedness I would think that RAID will make the number of drives 
a moot point, or at least change the nature of the beast since, to AFS, all the 
drives will look like a single large drive.

We will test different partition schemes and see if we can get extra servers for 
the cell.


Thanks for your insights!
Leland