On Wed, Oct 11, 2000 at 09:17:46AM -0400, rpjday wrote:
| > | p.s. dd if=/dev/hda bs=96k | gzip > test.gz for efficiency
| > Though of course that still wastes time chucking data through the pipe
| > between two programs. This:
| > gzip </dev/hda >test.gz
| > should win. (Where'd the 96k come from - is it just an arbirary big chunk or
| > do you have something specific in mind?)
|
| i see your first point -- is there any slowdown based on lack of blocking?
On many OSes, yes (because typically you dd to/from raw devices to
bypass the buffer poll, through which blocking factors are a little
meaningless). Linux doesn't provide raw devices - everything goes
through its buffer pool. I expect the device driver does the blocking
usually, though it must be meaningful in some sense or you'd have no
control over blocking on things like tapes, where it really matters.
Dd defaults to 512 bytes. The travesty of a manual entry the GNU people
provide says little, but a manual entry on my trusty Solaris box has the
info. On Linux, though, the default blocksize may be no big issue
because Linux is fairly aggressive about this - you'll be making more
system calls [read(0,buf,512)] but the overhead probably isn't
enormous.
Cheers,
--
Cameron Simpson, DoD#743 [EMAIL PROTECTED] http://www.zip.com.au/~cs/
We would've believed it was an accidental shooting if he hadn't changed
magazines ......TWICE - [EMAIL PROTECTED]
_______________________________________________
Redhat-list mailing list
[EMAIL PROTECTED]
https://listman.redhat.com/mailman/listinfo/redhat-list
