Subject Re: [ib-support] When do FB/IB 'flush' to disk?
Author Claudio Valderrama C.
"Phil Shrimpton" <phil@...> wrote in message
> On Tuesday 27 November 2001 16:17, you wrote:
> NT4, 3 x SCSI disks in RAID array. Everything is good apart from the OS
> I am going to try the test again tonight wiith W2K, if the same results I
> shall urge the customer to upgrade to Linux, or implement the period
> disconnect/reconnect routine.

Phil, please don't tell us that you have one of those monster RAID
controllers, able to cache in RAM from 32 up to 256 MB!!! Some (or several?)
of those artifacts don't have backup batteries, so they lose all cache on
power lost time.

Also, there's a thing called NTFS undo-redo log. NTFS itself is able to
rollback inconsistent, incomplete operations. I think that in theory, it
should be able to figure out when an operation was completed by not marked
as safe and then mark it instead of discarding its effects next time the OS

Claudio Valderrama C. - -
Independent developer
Owner of the Interbase® WebRing