Projects / mcelog

mcelog

mcelog is a daemon to handle machine check events (hardware errors) on x86-64 machines running an x86 Linux kernel. It accounts and logs CPU and memory errors, supports triggers on error thresholds, and can predictively offline memory pages and CPUs based on error trends. This daemon should run on all x86 Linux systems that want to handle hardware errors.

Tags
Licenses
Operating Systems
Implementation

Recent releases

  •  25 Dec 2009 14:42

    Release Notes: This is a test release towards 1.0. It should become 1.0 with minimal changes after some more testing. The 1.0 stream has lots of changes over older releases: new Intel CPU support, a daemon mode, memory DIMM error tracking, support for bad page offlining using predictive failure analysis (together with a 2.6.33 Linux kernel), a test suite, and much more. Note that the 0.9 stream has been abandoned.

    •  03 May 2006 12:39

      Release Notes: Modifier command line options are now allowed after --ascii. A decoded address is no longer printed twice for --ascii. The SMBIOS anchor scan was fixed to work on more machines and no longer crashes when no anchor is found. --ascii reparsing of mcelog output was fixed. NONE entries are not printed in SMBIOS.

      •  08 Feb 2006 15:31

        Release Notes: --ascii reparsing of mcelog output was fixed. --filter was added and known broken K8 GART errors are filtered out. The --ignorenodev argument was added and is used in the cron script (which avoids cron errors in Xen guest kernels). The new --dmi argument was added to resolve machine check addresses to DIMM numbers in SMBIOS (however, this is unreliable due to wide-spread BIOS bugs). Argument decoding was fixed to support -- and to allow arguments in any order. --ascii was clarified in the manpage.

        •  19 Dec 2005 19:21

          Release Notes: This release has a P4 decoder for Intel P4/Xeon machine checks. It has support for DRAM error threshold registers (needs 2.6.15+) in AMD K8 Revision F. The --ascii description in the man page has been clarified.

          •  26 Jan 2005 17:29

            Release Notes: The K8 decoder from the Linux 2.4 Kernel code was added to decode Opteron/Athlon64 logs. The --ascii function was added to decode fatal kernel output. The manpage was improved. 32-bit bugs were fixed. An uninitialized variable in check_cpu was fixed. Minor cleanups were made.

            Screenshot

            Project Spotlight

            OpenStack4j

            A Fluent OpenStack client API for Java.

            Screenshot

            Project Spotlight

            TurnKey TWiki Appliance

            A TWiki appliance that is easy to use and lightweight.