1. Forum
  2. FidoNet
  3. linux.debian.user

  • IO performance information taken out of kernel

    From C.T.F. Jansen@3:633/10 to All on Tuesday, July 21, 2026 04:10:01
    Greetings

    Someone has performed an act of hacking IO performance information
    collection from the kernel.

    iotop and iotop-c don't collect IO wait information anymore unless one
    "opts in" by setting up the relevant variable in /etc/default/grub. This
    is the wrong way round for running a server.

    If someone is that desperate for space/time then they can opt out. The
    rest need to know what is going on in a busy system and there is plenty
    of CPU etc to collect said information.

    It should be changed back to how it was and more oversight done to
    prevent such changes creeping in.

    frank.jansen@actrix.gen.nz

    --- PyGate Linux v1.5.18
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)
  • From Michel Verdier@3:633/10 to All on Tuesday, July 21, 2026 06:40:01
    On 2026-07-21, C.T.F. Jansen wrote:

    iotop and iotop-c don't collect IO wait information anymore unless one "opts in" by setting up the relevant variable in /etc/default/grub. This is the wrong way round for running a server.

    Which kernel version do you use ?
    Man page says "kernel.task_delayacct sysctl enabled". You can set it in /etc/sysctl.d/*. If you used /etc/sysctl.conf it could have been removed
    during procps upgrade.

    --- PyGate Linux v1.5.18
    * Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)