> I dropped that idea, but I'd read a post somewhere about dd, I tried...
> dd if=/dev/ad2 of=/dev/da0
> ...now the IDE activity LED has been on solid for about 4hours

dd's performance is highly depending on the size of its buffers, which are
abysmally small (512 bytes) by default:

    [EMAIL PROTECTED]:~$ dd if=/dev/zero of=/tmp/foo count=1
    1+0 records in
    1+0 records out
    512 bytes transferred in 0.000074 secs (6916211 bytes/sec)

You can increase this by specifying your own buffer size:

    [EMAIL PROTECTED]:~$ dd if=/dev/zero of=/tmp/foo count=1 bs=16384
    1+0 records in
    1+0 records out
    16384 bytes transferred in 0.000173 secs (94654927 bytes/sec)

Before starting the full transfer, do something like:

dd if=/dev/ad2 of=/dev/da0 bs=512 count=16384

and keep doubling the size of the bs argument until the throughput values
stop increasing noticably, then use that value to duplicate your drive.  It
will still be somewhat slow, but I guarantee you can speed it up by at least
5 times.
Kirk Strauser
In Googlis non est, ergo non est.
<< attach3 >>

Here's the results of doubling the buffer size as you suggested:

BS      bytes/sec
512     60808 (default)
1024    121614
2048    240212
4096    470312
8192    898897
16384   1652020
32768   2848452
65536   4490871*
131072  3811765
262144  3438026...

iostat now reports a steady 4+MB/s (75x increase).

Thanks alot, Kirk!

MSN 8 helps eliminate e-mail viruses. Get 2 months FREE*. http://join.msn.com/?page=features/virus

To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-questions" in the body of the message

Reply via email to