Re: Explaining single level store to non i people

On Mon, Aug 17, 2009 at 16:20, Nathan Andelin<nandelin@xxxxxxxxx> wrote:

No. ÂI was referring to the "c:" drive running out of space. ÂSay a spammer floods a mail server with garbage that fills the drive unexpectedly.

C:\ is the system drive. It should never contain any data except the
operating system - not even queues for your mail server. This is a
best practice that is rather different from the IBM i.

My understanding is that under Windows you allocate space to a "drive", then reference that drive from applications. The logical drive is a fixed size - and doesn't extend automatically, when another physical drive is added.

That's correct. Just like on the IBM i (where you have to add the
physical drive to the ASP you want to use), you will need to add the
newly configured hard drive (or RAID array) to the partition (~= ASP)
you want to use it with.

Under IBM i, no space is specifically allocated to libraries, files, file systems, or databases. ÂAdding a physical drive automatically extends any space available to everything on the server. ÂNo need to partition anything, no need to designate a logical drive, no need to allocate space to anything, no need to move anything, no need to change applications to reference a different "drive", for example.

If you only use a single ASP, this is true. However, the same applies
if you're using your Windows servers the same way - namely create one
giant C:\ drive and put all your data on it. Then it behaves exactly
the same as an ASP.

There's a difference on best practices here - on Windows, Linux, Unix,
AIX, you usually try to segregate the different workloads onto
separate sets of platters.

On the IBM i, this is usually not done - many shops just run a single,
giant ASP 1 with all the drives in it.

If i had to guess a reason for this, i'd say it's related to disk
price - a 147GB 15kRPM SAS Drive on the System x side is ~500$, on the
Power Side it's ~1500$.