Re: Why doesn't Oracle care about Linux as IBM does?

On Thu, 16 Aug 2001 16:48:19 -0400, Blair Kenneth Adamache <adamache_at_ca.ibm.com> wrote:

just another $0.02 worth:

>Oracle 9i: data is shared amongst nodes (shared data)
More appropriately: data may be shared if needed. Not mandatory, not necessary. But it's there if you need it. No need to re-design anything if you started out single node, then you need to *scale* to clustering. Minor detail. But very, very important when we talk scalability.

>DB2 EEE, Informix XPS: data is shipped between nodes, but never shared; function
>is shipped; disk can be shared on large SMP's or disk can be private to nodes on
>clustered small SMPs (like 4-ways)

I find the "disk can be shared on large SMP's" part a bit disturbing, to say the least. I'd say that's a given of ANY database running in a large SMP system.
At least it's been since I've been using O in that environment, for quite a while now. And I don't need the OPS thing active for that.

Funny enough, in the Numa-Q architecture (clustered small 4-way SMPs), I don't even need the OPS to run effectively in all clusters. Although I can use it, I don't need it at all. It runs quite well in all clusters, sharing data, disks, memory, you name it.

>Teradata: data and disks are never shared.

Sounds awfully like the M$ solution. Dunno, never been involved with them.

Cheers
Nuno Souto
nsouto_at_optushome.com.au.nospam Received on Fri Aug 17 2001 - 06:16:45 CDT