Uploaded image for project: 'OpenVZ'
  1. OpenVZ
  2. OVZ-4891

Lots of "Uncharging too much" in 2.6.18-238.1.1.el5.028stab084.1ent

    XMLWordPrintable

    Details

    • Type: Bug
    • Status: Closed
    • Priority: Major
    • Resolution: Fixed
    • Fix Version/s: OpenVZ-legacy
    • Component/s: Containers::Kernel
    • Security Level: Public
    • Environment:
      Operating System: RHEL/CentOS 5
      Platform: x86 (i386)

      Description

      I've just booted 2.6.18-238.1.1.el5.028stab084.1ent on my test server and I'm seeing *lots* for "Uncharging too much" for kmemsize (and only for kmemsize).

      This happens as well on early stage of boot, where there are no CTs started (only CT0) as later, when CTs are running. The amount of messages is staggering, they fill dmesg buffer instantly. I'm attaching /var/log/messages contents, starting from reboot.

      Since it starts to happen even before any CT is up, I don't think that's related to my CTs configuration, but for reference, I am also attaching output of /proc/user_beancounters from 238.1.1.el5.028stab084.1ent (file bcnt.stab84) and 238.1.1.el5.028stab083.1ent, which I've been happily running earlier (file bcnt.stab83).

      If you need any more detail, please let me know. For now, I'm going back to stab083.1 as this uncharging too much is filling up my logs.

        Attachments

        1. ubc-fix-batched-page-tables-accounting-for-x86
          1 kB
        2. ubc-batch-non-4level-page-tables-kmem-acct
          0.9 kB
        3. d11.messages.txt
          963 kB
        4. bcnt.stab84
          11 kB
        5. bcnt.stab83
          11 kB

          Activity

            People

            Assignee:
            khlebnikov@openvz.org Konstantin Khlebnikov
            Reporter:
            shasta+ovz@toxcorp.com Jakub Jankowski
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

              Dates

              Created:
              Updated:
              Resolved: