Initially I have only 1 drive, and installed Rosetta in Drive C: it works fine.

Later I installed Drive D:. It was great.

But then I found that Rosetta consistently accessing my drive d: with no reason. This prevents my Drive D: from going into sleep mode. Or worst waking it up too frequent.

I have actually asked this question before, but found no answer. If rosetta continue to behave like that I will consider to un-install it soon.

Attached the log from Filemon

3:36:57 PM rosetta_beta_5.:8812 OPEN D: SUCCESS Options: Open Access: 00100180
3:36:57 PM rosetta_beta_5.:8812 QUERY INFORMATION D: BUFFER OVERFLOW FileFsVolumeInformation
3:37:03 PM rosetta_beta_5.:8812 OPEN D: SUCCESS Options: Open Directory Access: 00100001
3:37:03 PM rosetta_beta_5.:8812 OPEN D:.1fad5997entercept20411bbc SUCCESS Options: Open Directory Access: Read
3:37:03 PM rosetta_beta_5.:8812 CLOSE D:.1fad5997entercept20411bbc SUCCESS
3:37:03 PM rosetta_beta_5.:8812 DIRECTORY D: SUCCESS FileNamesInformation
3:37:03 PM rosetta_beta_5.:8812 READ D:$Directory SUCCESS Offset: 0 Length: 4096
3:37:03 PM rosetta_beta_5.:8812 DIRECTORY D: NO MORE FILES FileNamesInformation
3:37:13 PM svchost.exe:1324 OPEN D: SUCCESS Options: Open Access: 00020088
3:37:13 PM svchost.exe:1324 OPEN D:.1fad5997entercept20411bbc SUCCESS Options: Open Directory Access: Read
3:37:13 PM svchost.exe:1324 CLOSE D:.1fad5997entercept20411bbc SUCCESS

Initially I have only 1 drive, and installed Rosetta in Drive C: it works fine.

Later I installed Drive D:. It was great.

But then I found that Rosetta consistently accessing my drive d: with no reason. This prevents my Drive D: from going into sleep mode. Or worst waking it up too frequent.

I have actually asked this question before, but found no answer. If rosetta continue to behave like that I will consider to un-install it soon.

Attached the log from Filemon
{...}

I won't pretend to understand the log that you included, but I will offer another suggestion:

It is possible that Windows -- in its wisdom ;) -- has decided to include your drive D: for a pagefile. You can check this possibility by going through the Control Panel to get to System Properties (or right-click on My Computer), then:

System Properties - Advanced (tab) > Performance - Settings (button) > Performance Options - Advanced (tab) > Virtual Memory - Change (button)

Highlight the line with D: on it and read the info in the panels below.

If D: has a paging file allocated to it, that would be (at least part of the reason) why it never spins down.

You can choose to have "No paging file" on D:, but make sure that you have a sufficiently large pagefile on C: Windows sets (by default) a size equal to 1.5 times your physical RAM. As long as your drive C: isn't too crowded, I would suggest you leave it at the default value.

I might be completely off base here, and I bow to any contrary experience you may have, or better advice offered by others on this forum.

---

Rosie, Rosie, she's our gal,
If she can't do it, no one shall!

There is also a BOINC General Preference where you can put a cap on how frequently BOINC should write updates to disk. You didn't mention that, so I take it perhaps you were unaware of it. And I believe the default is 1 min. This doesn't mean BOINC WILL write to disk every minute. But it means Rosetta is basically being told to write as often as it would like.

So you might want to set the value higher, so your drive has some idle time to spin down. Keep in mind that the higher you set the time (i.e. the less frequently you allow data to be written to disk) the more processing time you will tend to lose when BOINC is exited or the computer is shutdown. This loss because BOINC was unable to preserve it's results by writing to disk.

Last I knew, not all BOINC applications honored the "write to disk at most every..." general preference. But I believe Rosetta does.

---

Rosetta Moderator: Mod.Sense

Last I knew, not all BOINC applications honored the "write to disk at most every..." general preference. But I believe Rosetta does.

I was just now reading over at Einstein@home forums that Einstein ignores this; at least, the app seems to honour it, but the debug output doesn't, so that the disk is getting written to anyhow. I don't think they'd worked out yet if the debug stuff was BOINC or the worker app though.

I'd like for my disks to power down, but there always seems to be -something- that wants its turn!! : )

---

Crunching Rosetta as a member of the Whirlpool BOINC Teams

Since the drive with the Windows swap file will be in use anyway, I would suggest installing BOINC on that drive. Unfortunately, there isn't yet a way to seperate the infrequently needed BOINC and project programs from the very frequently needed task-specific data.

---

Rosetta Moderator: Mod.Sense

i've seen random read requests from Rosetta as CKKwan has posted, including some of the programmers names included in the paths! I think these are erroneously left in from some debugging, and I don't think it's anything to do with the PF as I believe that's transparent to the program making the reads/writes and therefore also transparent to filemon (i.e. filemon would just see a read request for the file in its true c: location rather than its virtual pagefile location).

---

Since the drive with the Windows swap file will be in use anyway, I would suggest installing BOINC on that drive. Unfortunately, there isn't yet a way to seperate the infrequently needed BOINC and project programs from the very frequently needed task-specific data.

Yes there is :p , but BOINC Manager does not (at least in 5.10.20) honour this request, it still needs the 'boinc manager' log files and gui_rpc_auth.cfg to be in the same directory.


But everything else seems to work.

---

Team mauisun.org

In answer to the actual post though... I would check your pagefiles, VISTA would tend to use all the drives.

If you want to power it down you'll certainly need to take the pagefile off it.


If it's not that, post back though :-)

---

Team mauisun.org