e4defrag (1.42.7) is "slow" (RAID6)

Anselm D
2013-03-14
2013-03-14
  • Anselm D
    Anselm D
    2013-03-14

    I did some performance tests with my RAID6 and e4defrag and i am wondering, why e4defrag is so "slow".
    The System: No Page file, 20GB main memory, raid 6

    Summary of the test.

    1. Creating big file with dd: 81920000000 bytes (82 GB) copied, 87.019 s, 941 MB/s
    2. Copying the file: 236 s
    3. Copying the file two time in parallel to produce something to defrag: 306 s
    4. e4defrag one of the in 3. copied files: 39:18 min = 2358 s

    Some detailed measurements

    Using /usr/bin/time with the following parameters (taken from man pages):

    • I Number of file system inputs by the process.
    • O Number of file system outputs by the process.
    • P Percentage of the CPU that this job got. This is just user + system times divided by the total running time. It also prints a percentage sign.
    • R Number of minor, or recoverable, page faults. These are pages that are not valid (so they fault) but which have not yet been claimed by other virtual pages. Thus the data in the page is still valid but the system tables must be updated.
    • w Number of times that the program was context-switched voluntarily, for instance while waiting for an I/O operation to complete.

    Output for 1. create file with dd
    1:27.03 real, 2.56 user, 78.88 sys
    M=928KiB
    i=40
    o=160000000
    P=93%
    R=286
    w=9732

    Output for 2. copy file
    3:56.04 real, 0.42 user, 90.76 sys
    M=956KiB
    i=160000504
    o=160000000
    P=38%
    R=292
    w=14210

    Output for 4. e4defrag
    39:18.06 real, 0.05 user, 391.27 sys
    M=1000KiB
    i=160007256
    o=160000000
    P=16%
    R=308
    w=20572253

    Does anyone have a good explanation for this?

     
    Last edit: Anselm D 2013-03-14
  • Anselm D
    Anselm D
    2013-03-14

    Wrong Forum, can anyone move it to open discussions, please?