> I have used ext2 and ReiserFS on my desktops. But when I put Reiser on my
> laptop, it keeps spinning up the disk periodically which will not allow hte
> machine to take full advantage of APM. Any parameters I should set, or is
> a different filesystem better?

Journaling filesystems have to write periodically to the harddrive
("commits"), but try ext3 with the latop-mode kernelpatch, this will
keep ext3 from doing commits every 10(?)seconds. If really not needed
the harddrive might stay quiet for up to 10 minutes.
See the readme for more.

I successfully use it on my iBook2 ATM where this patch came with the
2.4.21-benh2 patch.

-Thomas

Thomas Otto writes:
> Journaling filesystems have to write periodically to the harddrive
> ("commits"), but try ext3 with the latop-mode kernelpatch, this will
> keep ext3 from doing commits every 10(?)seconds. If really not needed
> the harddrive might stay quiet for up to 10 minutes.
> See the readme for more.
>
> I successfully use it on my iBook2 ATM where this patch came with the
> 2.4.21-benh2 patch.

I wonder whether it makes sense to journal to a flash card instead of
to the hard drive. Periodically the flash card contents could be
flush to the hard drive.

Oops! Paul Rubin was seen spray-painting on a wall:
> Thomas Otto writes:
>> Journaling filesystems have to write periodically to the harddrive
>> ("commits"), but try ext3 with the latop-mode kernelpatch, this will
>> keep ext3 from doing commits every 10(?)seconds. If really not needed
>> the harddrive might stay quiet for up to 10 minutes.
>> See the readme for more.
>>
>> I successfully use it on my iBook2 ATM where this patch came with the
>> 2.4.21-benh2 patch.
>
> I wonder whether it makes sense to journal to a flash card instead of
> to the hard drive. Periodically the flash card contents could be
> flush to the hard drive.

That's attractive, but not _quite_ right.

Unfortunately, flash cards have a limited lifetime (e.g. limited
number of writes), and this is going to absolutely _hammer_ the flash
card with writes.

So, yes, that's feasible, but what it means is you should get an 8MB
card (I have seen them available for about $5), configure for an 8MB
journal, and figure on replacing it fairly often.
--
If this was helpful, rate me
http://www3.sympatico.ca/cbbrowne/linuxxian.html
"IBM may have originated Not Invented Here, but Microsoft has managed
to squander their mountain of technical talent by not harvesting prior
art until the rain of flaming hardware makes it difficult for the
executives to drive to the Redmond campus.)" -- david parsons

On Sun, 14 Sep 2003 11:57:52 +0200, Thomas Otto
wrote in comp.os.linux.portable:

>Journaling filesystems have to write periodically to the harddrive
>("commits"), but try ext3 with the latop-mode kernelpatch, this will
>keep ext3 from doing commits every 10(?)seconds. If really not needed
>the harddrive might stay quiet for up to 10 minutes.
>See the readme for more.

Most journalled filesystems suffer from a recursive update problem because
each journal commit operation generates new meta-data which is then dumped
back into the queue.

You can substantially reduce the number of writes by mounting your
filesystems with atime updates turned off. This will almost eliminate the
recursive meta-data-update problem.

ISTR this breaks "make", however, so it may not be possible to turn off
atime on systems used for software compiling.

On Sun, 14 Sep 2003 16:36:31 GMT, Chris staggered into the Black Sun and
said:
> On Sun, 14 Sep 2003 11:57:52 +0200, Thomas Otto
> wrote in comp.os.linux.portable:
>
>>Journaling filesystems have to write periodically to the harddrive
>>("commits"), but try ext3 with the latop-mode kernelpatch, this will
>>keep ext3 from doing commits every 10(?)seconds. If really not needed
>
> Most journalled filesystems suffer from a recursive update problem
> because each journal commit operation generates new meta-data which is
> then dumped back into the queue.
>
> You can substantially reduce the number of writes by mounting your
> filesystems with atime updates turned off. This will almost eliminate
> the recursive meta-data-update problem. ISTR this breaks "make",
> however

It does? That doesn't jive with my experience at all:

mount | grep home
/dev/hda7 on /home type reiserfs (rw,noatime)
cd /home/me/packages/Steamband_031_src/src ; rm -f *.o ; make
(everything gets compiled since all the .o files are gone)
touch z-form.c ; make
(only z-form.c gets compiled)

> so it may not be possible to turn off atime on systems used for
> software compiling.

Nah. From "man make":

The make program uses the makefile data base and the last-modification
times of the files to decide which of the files need to be updated.

....make uses mtime, not atime, in other words. I hate to think what a
make that used atime would do on a big project!

--
Matt G|There is no Darkness in Eternity/But only Light too dim for us to see
Brainbench MVP for Linux Admin /
http://www.brainbench.com / "He is a rhythmic movement of the
-----------------------------/ penguins, is Tux." --MegaHAL

Frank schrieb:

> machine to take full advantage of APM. Any parameters I should set,
> or is a different filesystem better?

Try out JFS, it seems to commit only if there are changes to the
filesystem. The harddisc on my thinkpad goes to sleep and stay until
pressing a key.

Bye Eric

Christopher Browne wrote:

>> I wonder whether it makes sense to journal to a flash card instead of
>> to the hard drive. Periodically the flash card contents could be
>> flush to the hard drive.
>
> That's attractive, but not quite right.
>
> Unfortunately, flash cards have a limited lifetime (e.g. limited
> number of writes), and this is going to absolutely hammer the flash
> card with writes.

why not use a ramdisk instead?

Lukas Kraehenbuehl writes:
> >> I wonder whether it makes sense to journal to a flash card instead of
> >> to the hard drive. Periodically the flash card contents could be
> >> flush to the hard drive.
> >
> > That's attractive, but not quite right.
> >
> > Unfortunately, flash cards have a limited lifetime (e.g. limited
> > number of writes), and this is going to absolutely hammer the flash
> > card with writes.
>
> why not use a ramdisk instead?

The idea of a flash card is it doesn't lose its contents on system
crashes.

Flash cards are good for millions of write cycles, but yeah, if
there's an update every few seconds, maybe write wear starts to be an
issue.

There are also PCMCIA SRAM cards, which are very nice, but have much
lower capacity and $/MB than flash cards.

Lukas Kraehenbuehl wrote:

>> Unfortunately, flash cards have a limited lifetime (e.g. limited
>> number of writes), and this is going to absolutely hammer the flash
>> card with writes.
>
> why not use a ramdisk instead?
>

And when the power dies?

--

Fundamentalism is fundamentally wrong.

To reply to this message, replace everything to the left of "@" with
james.knott.

On Sun, 14 Sep 2003 02:49:10 -0400, Frank wrote:

> I have used ext2 and ReiserFS on my desktops. But when I put Reiser on my
> laptop, it keeps spinning up the disk periodically which will not allow hte
> machine to take full advantage of APM. Any parameters I should set, or is
> a different filesystem better?
I have the Ext3 on both desktop and a laptop, had a few power outages
without destroying the file system

Paul

After takin a swig o' Arrakan spice grog, Lukas Kraehenbuehl belched out...:
> Christopher Browne wrote:
>>> I wonder whether it makes sense to journal to a flash card instead of
>>> to the hard drive. Periodically the flash card contents could be
>>> flush to the hard drive.
>>
>> That's attractive, but not quite right.
>>
>> Unfortunately, flash cards have a limited lifetime (e.g. limited
>> number of writes), and this is going to absolutely hammer the flash
>> card with writes.
>
> why not use a ramdisk instead?

That's a thought; you'll lose ALL the files every time you reboot, and
eliminate all the annoying things about reading them in.
--
(reverse (concatenate 'string "gro.mca" "@" "enworbbc"))
http://www3.sympatico.ca/cbbrowne/advocacy.html
"Computers are like air conditioners: They stop working properly if
you open windows."