home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!math.fu-berlin.de!mailgzrz.TU-Berlin.DE!mailgzrz.tu-berlin.de!nickel
- From: nickel@cs.tu-berlin.de (Juergen Nickelsen)
- Newsgroups: alt.folklore.computers
- Subject: sync; sync; sync; was Re: Computer Superstition
- Date: 18 Nov 1992 17:17:10 GMT
- Organization: STONE Project, Technical University of Berlin, Germany
- Lines: 34
- Message-ID: <NICKEL.92Nov18181710@desaster.cs.tu-berlin.de>
- References: <10292323.7317.11306@kcbbs.gen.nz> <1992Nov18.132507.22731@bnr.uk>
- Reply-To: nickel@cs.tu-berlin.de
- NNTP-Posting-Host: desaster.cs.tu-berlin.de
- In-reply-to: agc@bmdhh286.bnr.ca's message of Wed, 18 Nov 1992 13:25:07 GMT
-
- In article <1992Nov18.132507.22731@bnr.uk> agc@bmdhh286.bnr.ca (Alan
- Carter) writes:
-
- > How about the way that UNIX System Administrators type
- >
- > sync
- > sync
- > sync
- >
- > at shutdown time, when one sync would be quite enough? When questioned
- > they sometimes give the lame excuse that this gives the first one time
- > to work! I was taught to do this, and still do, even though I have
- > figured it out for myself that it is nonsense.
-
- Hmm, perhaps this *has* to appear regularly.
-
- The sync(1) program does nothing but invoke the sync(2) system call.
- This call writes the contents of modified disk buffers back to the
- disk. The manual (SunOS, in this case) of sync(2) states:
-
- BUGS
- The writing, although scheduled, is not necessarily complete
- upon return from sync().
-
- An todays fast machines with fast disks this won't matter -- just type
- sync; and before you've had time to do something else the disk buffers
- will be safe.
-
- But back in the old days, it was not a bad idea to type sync three
- times (and not as "sync; sync; sync") to make sure the first call was
- completed.
-
- --
- Juergen Nickelsen
-