Need prog. to prevent hdd going into idle (such as a perdiodic access?)

  • Thread starter Thread starter Louis Grant
  • Start date Start date
true. I wonder how xp manages to run the prog. for long intervals
without hdd usage...
Click to expand...

BTW, check the output. The first number is the offset inside the file
where data is read; the first number is the actual number of bytes
read. It should always be 1.
 
Fabien said:
BTW, check the output. The first number is the offset inside the file
where data is read; the first number is the actual number of bytes
read. It should always be 1.
Click to expand...

this is what I get:

12019800 1 381503867 1 404924282 1 611237925 1
797221510 1
312804581 1 124076904 1 642393369 1 188035310 1
780366693 1
942117838 1 249787590 1 1084369507 1 1050949681 1
513427362 1
753195145 1 625151701 1 129729201 1 138249442 1
540589241 1
642299677 1 988887002 1 204696356 1 790871716 1
612782060 1
709611824 1 896099530 1 524848883 1 428941512 1
686936930 1
194183739 1 335797905 1 465162359 1 471235202 1
562457128 1
1855399 1 79380977 1 289513197 1 404236025 1
224420206 1
240957732


looks good, right?

Louis
 
looks good, right?
Click to expand...

*Looks* good, yes.

Unfortunately, we have no way to know whether the file is actually
read.

Delete all the files you got from me, and get this one:
<http://perso.edulang.com/csiphs/access_hard_drive.v2.zip>

access_hard_drive.create.exe will create a non-trivial 1.5 GB
"c:\dummy" file.

The new "access_hard_drive.exe" will read random values in the file
(as did the previous version), and check that what it reads is
consistent.

The value to be read is nontrivial (17 * offset), so no caching system
can predict it.
And, no caching system is able to cache a 1.5 GB file with only 512 MB
RAM.

If you've got something like "16608188 ok 527137596 ok
559498424 ok 844569396 ok 1101549596 ok 432213324 ok", then
I'm pretty sure the file is being read (4 bytes at a time).
Otherwise, there's some kind of error message.
 
Fabien said:
Delete all the files you got from me, and get this one:
<http://perso.edulang.com/csiphs/access_hard_drive.v2.zip>

access_hard_drive.create.exe will create a non-trivial 1.5 GB
"c:\dummy" file.

The new "access_hard_drive.exe" will read random values in the file
(as did the previous version), and check that what it reads is
consistent.

The value to be read is nontrivial (17 * offset), so no caching system
can predict it.
And, no caching system is able to cache a 1.5 GB file with only 512 MB
RAM.

If you've got something like "16608188 ok 527137596 ok
559498424 ok 844569396 ok 1101549596 ok 432213324 ok", then
I'm pretty sure the file is being read (4 bytes at a time).
Otherwise, there's some kind of error message.
Click to expand...

ok... it seems to work when I run your programm the first time after a
reboot, I tested it for about 20 minutes, will see though when I work
on the laptop tomorrow.

yet, when i close down your programm and then restart it, it seems to
not read periodically anymore and it goes into idle... though the
programm keeps returning numbers.

when I reboot and restart the programm, it appears to work again.

what could that be? some *other* form of caching? can one clear it
before starting the programm again?

Ingolf
 
yet, when i close down your programm and then restart it, it seems to
not read periodically anymore and it goes into idle... though the
programm keeps returning numbers.
Click to expand...

Actually, it's got to do with the random number generator: each time
you start the program, it's reset, so it will give the same list of
offsets -- which have already been read.

This third version fix this, by initializing properly the random
number generator.
<http://perso.edulang.com/csiphs/access_hard_drive.v3.zip>
 
Fabien said:
Actually, it's got to do with the random number generator: each time
you start the program, it's reset, so it will give the same list of
offsets -- which have already been read.

This third version fix this, by initializing properly the random
number generator.
<http://perso.edulang.com/csiphs/access_hard_drive.v3.zip>
Click to expand...

aha! ;)

seems to work now also on second run.

i'll report back tomorrow to see if it works for more than 30
minutes...

thanks your all your effort, very much appreciated!

Louis

PS: just for my peace of mind, all this program is doing is to read
some bits every x milliseconds, ... this shouldnt be bad for the drive,
right?
 
hello again,

I let the laptop run overnight and this morning it was still reading
periodically. havent had a chance to work for several hours but I
imagine it''ll be fine.

should I come across a problem after all (more work next week), I'll
post here...

thanks again,
Louis

PS: just for my peace of mind, all this program is doing is to read
some bits every x milliseconds, ... this shouldnt be bad for the drive,
right?
Click to expand...

any clue?
 
Arno Wagner said:
Not likely. There is nothing Fujitsu can do, since APM is not a function
of the disk, but rather the OS. Unless you are talking about the auto
spindown of the drive? That can be changed. Don't know the right
software for Windows. Mybe the Hitachi HDD feature tool can help.
Maybe Sisoft Sandra can do it.
Click to expand...

Babble, babble, babble, rant.
Just the usual cluelessness form our resident babblehead.
 
Back
Top