Messages in this thread | | | Date | Mon, 12 Nov 2001 13:43:37 -0800 | From | Andrew Morton <> | Subject | Re: File System Performance |
| |
Steve Lord wrote: > > ... > > I tried an experiment which puzzled me somwhat: > > > mount /xfs > > cd /xfs/lord/xfs-linux > > time tar cf /dev/null linux > > real 0m7.743s > user 0m0.510s > sys 0m1.380s > > > hdparm -t /dev/sda5 > > /dev/sda5: > Timing buffered disk reads: 64 MB in 3.76 seconds = 17.02 MB/sec > > > du -sk linux > 173028 linux > > The tar got ~21 Mbytes/sec. >
It's tar. It cheats. It somehow detects that the output is /dev/null, and so it doesn't read the input files. I think.
akpm-1:/opt> l total 562884 -rw-r--r-- 1 akpm akpm 575785370 Apr 8 2001 backup
akpm-1:/opt> time tar cf /dev/null backup tar cf /dev/null backup 0.02s user 0.00s system 4% cpu 0.422 total
That's 570 megs in 0.4 seconds. Impressive.
So I just use time (find dir -type f | xargs cat > /dev/null)
> In our cvs tree cmd/xfsprogs/tests/src/fsstress.c
OK, thanks.
- - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |