right, the oracle system uses a second "low latency" bus to
manage locking information (at the block level) via a
distributed lock manager. (but this is slightly different
albeit related to a clustered file system and OS-managed
locking, eg)

-----Original Message-----
From: pgsql-performance-owner@postgresql.org [mailtogsql-performance-owner@postgresql.org] On Behalf Of Dawid Kuroczko
Sent: Wednesday, April 20, 2005 4:56 AM
To: pgsql-performance@postgresql.org
Subject: Re: [PERFORM] How to improve db performance with $7K?


On 4/19/05, Mohan, Ross <RMohan@arbinet.com> wrote:
> Clustered file systems is the first/best example that
> comes to mind. Host A and Host B can both request from diskfarm, eg.

Something like a Global File System?

http://www.redhat.com/software/rha/gfs/

(I believe some other company did develop it some time in the past; hmm, probably the guys doing LVM stuff?).

Anyway the idea is that two machines have same filesystem mounted and they share it. The locking I believe is handled by communication between computers using "host to host" SCSI commands.

I never used it, I've only heard about it from a friend who used to work with it in CERN.

Regards,
Dawid

---------------------------(end of broadcast)---------------------------
TIP 8: explain analyze is your friend

---------------------------(end of broadcast)---------------------------
TIP 5: Have you checked our extensive FAQ?

http://www.postgresql.org/docs/faq