Re: [gentoo-user] [OT] Fragmentation of my drives. Curious mostly

[Joshua Murphy]

On Mon, Dec 1, 2008 at 3:11 PM, Daniel Iliev <daniel.iliev@×××××.com> wrote:
> On Sun, 30 Nov 2008 14:27:22 +0200
> Alan McKinnon <alan.mckinnon@×××××.com> wrote:
>
> [-snip-]
>
>>
>> You are the one making them claims! You do the test!
>
>
> Me? No, my dear. It is you who claims that fragmentation only matters
> on MS file systems while Linux FS remain unaffected w/o providing
> any proof. The common sense says otherwise: fragmentation leads to
> seeks, seeks hurt performance.
>
> Anyways, I'm going to do the tests because I want to see the numbers.
> Should the outcome prove you were right I'd be the first to
> congratulate and thank you for opening my eyes.
>
> Since I'm going to use my workstation at work to do the job at night
> I'll limit the repetitions to 400 which would take about 13 hours with
> two files, 2GB each. The machine has 1GB RAM installed and I'll run the
> test in Gentoo's "boot" rc level. I've closed everything except sshd
> and screen (see "ps" outpput below) and have taken care of the cache
> by clearing it before each run via /proc (see the script). The file
> system is ext3 and during the test it will be used in about 30% of its
> capacity:
>
> localhost test # df -h | grep home
> /dev/sdc1   15G  718M   14G   5% /home
>
> localhost test # mount -v | grep home
> /dev/sdc1 on /home/ type ext3
> (rw,noatime,nodiratime,data=journal,commit=1)
>
>
> localhost test # rc-status
> Runlevel: boot
> bootmisc        [ started  ]
> checkfs         [ started  ]
> checkroot       [ started  ]
> clock           [ started  ]
> hostname        [ started  ]
> iptables        [ started  ]
> localmount      [ started  ]
> modules         [ started  ]
> net.eth0        [ started  ]
> net.lo          [ started  ]
> sshd            [ started  ]
> udev-postmount  [ started  ]
>
>
> localhost test # ps af
>  PID TTY      STAT   TIME COMMAND
> 12191 pts/0    Ss     0:00 -bash
> 12227 pts/0    S      0:00  \_ su -
> 12231 pts/0    S      0:00      \_ -su
> 12266 pts/0    S+     0:00          \_ screen -x
>  8575 pts/3    Ss+    0:00 -/bin/bash
>  4967 pts/2    Ss     0:00 -/bin/bash
> 12830 pts/2    R+     0:00  \_ ps af
>  4895 pts/1    Ss+    0:03 -/bin/bash
>  4755 tty6     Ss+    0:00 /sbin/agetty 38400 tty6 linux
>  4754 tty5     Ss+    0:00 /sbin/agetty 38400 tty5 linux
>  4753 tty4     Ss+    0:00 /sbin/agetty 38400 tty4 linux
>  4752 tty3     Ss+    0:00 /sbin/agetty 38400 tty3 linux
>  4751 tty2     Ss+    0:00 /sbin/agetty 38400 tty2 linux
>  4750 tty1     Ss+    0:00 /sbin/agetty 38400 tty1 linux
>
>
> The script:
>
> #!/bin/bash
>
> cat /usr/portage/distfiles/* > test1
> cp test1 test2
>
> filefrag test*
>
> sync
>
> for (( i=0 ; i<=400 ; i++ ))
> do
>
>  sync
>  echo 3 > /proc/sys/vm/drop_caches
>
>  echo -n "try_no=$i "
>
>  /usr/bin/time -f "\
> command=%C|\
> real_t=%e|\
> kernel_t=%S|\
> user_t=%U|\
> major_faults=%F|\
> minor_faults=%R|\
> context_sw=%c|\
> io_waits=%w|\
> fs_reads=%I|\
> fs_writes=%O" cp test1 /dev/null
>
>  sync
>  echo 3 > /proc/sys/vm/drop_caches
>
>  echo -n "try_no=$i "
>
>  /usr/bin/time -f "\
> command=%C|\
> real_t=%e|\
> kernel_t=%S|\
> user_t=%U|\
> major_faults=%F|\
> minor_faults=%R|\
> context_sw=%c|\
> io_waits=%w|\
> fs_reads=%I|\
> fs_writes=%O" cp test2 /dev/null
>
>
> done
>
> ###########EOF
>
>
>
> I hope all this will be enough for you to accept the results.
>
>
>
> --
> Best regards,
> Daniel
>
>

One thing that comes to mind... if one of those files is primarily
closer to the start of the disk than the other, and the drive tapers
down on speed as bad as some... well, that's one variable you can't
easily resolve. Whether that would influence enough to matter,
compared to seeks, is questionable though.

-- 
Poison [BLX]
Joshua M. Murphy