Re: OT: Yet Another Unhappy Customer for Vista

JackShephard wrote:

On Tue, 05 Jun 2007 16:38:00 -0700, John Larkin
<jjlarkin@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx> wrote:


It could have all sorts of files open, file control/directory blocks
partly coherent, disk blocks in mid-write, virtual memory tangled, all
sorts of nasty stuff. To its credit, NTFS seems to powerup gracefully,
unlike 95/98 FAT where it would insist on doing a scandisk sort of
thing on powerup if it thought the powerdown wasn't proper.


BOTH file system types get checked on bad shutdowns, and it has NOTHING
to do with the file system. It has EVERYTHING to do with Windows' OS
procedurals.

Lots of other OS's did this.


Journaling file systems DO check themselves independently of the OS
they run under. The support for them by the OS means that such routines
are built in. A bad shutdown makes a call to the checking routine upon
restart.


It's hard to design a file system that
can just be shut down any time.


It is hard to design a file system that can handle being shutdown while
a write operation is taking place safely.


It's done all the time in aeronautics. One scheme is several write plus majority decision upon read. This means some overhead but on a large jetliner on final approach into a busy place like LAX, with a few hundred people in there, a lengthy reboot after a tripped breaker is just not an option.


The system you describe would be easy. Simply set a flag on the drive
after any and all successful writes. Remove the flag during any write
op, and set it after.


Yep, that's another method.


If the flag is there on startup, the file system is considered whole
and valid.

It still has a safety concern, but very little. Older drives could
"spray" "bits" all over the drive's recording area as the head retracted
during a shutdown mid-write. Drive electronics on the drive itself has
since solved this problem. The exception would be a power surge passing
the whole way through, but such an animal would likely trounce the drive
electronics well before it hit the platters.


--
Regards, Joerg

http://www.analogconsultants.com
.

Relevant Pages

  • Re: OT: Yet Another Unhappy Customer for Vista
    ... It could have all sorts of files open, ... BOTH file system types get checked on bad shutdowns, ... It is hard to design a file system that can handle being shutdown while ... Older drives could ...
    (sci.electronics.design)
  • Re: hard drive scrubbing utility
    ... because it would have to write the *entire* disk. ... drives has increased, but not proportional to their density. ... But a scrubber has to actually *write* every sector of the hard drive. ... Some friends of mine worked on the Andrew File System at CMU, ...
    (microsoft.public.vc.mfc)
  • Re: Windows eXcresence Personified and disappearing HDs
    ... file system, and this file system is at risk. ... Configure your drives so the new one is Master and old one Slave. ... I'd cherry-pick the data off from DOS mode (thank goodness he wasn't ... So far, so good - no need to install XP, blah blah blah. ...
    (microsoft.public.windowsxp.hardware)
  • Re: Page fault in FreeBSD 5.3 on IBM e325, Dual AMD64 2.2GHz, 4GB RAM, ServeRAID 6M - debug logs
    ... Console: serial port ... <ACPI PCI bus> on pcib0 ... ips0: logical drives: 1 ... FILE SYSTEM CLEAN; SKIPPING CHECKS ...
    (freebsd-current)
  • Re: Dumb filesystem idea
    ... Im about to try a disklayout experiment and I wanted to ask everyone if ... SCSI is too expensive to waste on 10GB drives. ... losing the ability to have a high-capacicy file system. ... fault tolerance of RAID 1 and RAID 5 are nearly the same, ...
    (freebsd-questions)