home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!cis.ohio-state.edu!rutgers!att!ucbvax!VNET.IBM.COM!lybrand
- From: lybrand@VNET.IBM.COM (David Lybrand)
- Newsgroups: comp.sys.pen
- Subject: PenPoint slow to wake on NCR 3125 ?
- Message-ID: <9208281143.AA00565@ucbvax.Berkeley.EDU>
- Date: 28 Aug 92 11:33:57 GMT
- References: <1992Aug27.212924.20842@morrow.stanford.edu>
- Sender: daemon@ucbvax.BERKELEY.EDU
- Organization: IBM Corp, Boca Raton, FL
- Lines: 31
-
- In <1992Aug27.212924.20842@morrow.stanford.edu> Richard Acuff writes:
- >
- >1. Why does PenPoint take an extra 7 seconds to wake up over DOS?
- >
- >2. What does PenPoint need to do that takes 11.7 seconds and lots of
- > disk rattling when standing by?
- >
- >3. Why does the disk have to spin at all for wake-up?
- >
- PenPoint takes great pains to protect the user from data loss while in
- standby, since the user may very well walk away from the machine while
- it's in that mode. Thus it flushes all file-system buffers before entering
- standby. If the disk had spun-down before the attempt to go to standby,
- there would be time spent spinning it up before the buffers can be flushed.
- I still think the time is a bit longer than you'd expect, and I think that
- GO will be investigating these performance concerns SOMEday...
-
- The disk spin-up on wake-up is really a function of the MIL. Probably
- that the NCR MIL is automatically doing a spin-up on the resume call
- to the disk device. In all fairness, it's hard for the MIL to decide WHEN
- it should spin-up the disk, since the disk had been spun up at suspend
- time (going into standby). Thus it's probably returning it to the same
- state it was in prior to the suspend.
-
- But it's possible that PenPoint is refreshing it's filesystem buffers
- at that time, so this is probably another "safety feature" resulting from
- the previous flushing of the buffers.
- DPL
- Additionally, PenPoint repaints the screen on resume, which takes a bit
- longer than a DOS resume which "assumes" that the VGA memory is intact.
- DPL
-