https://community.hpe.com/t5/operating-system-linux/clock-backward/m-p/4072167#M30072 Hi gurus,<BR /> I'm kind of lost on this one. Suppose I issue `date` every second. This is what I get:<BR /><BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:30 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:30 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:31 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:33 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:34 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:30 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:31 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:31 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:32 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:33 EDT 2007<BR /><BR />Its always +4 seconds -4 seconds and once in a while the system can go forward one more second. <BR /><BR />This is a newly installed box, running for about a week. The hardware is not new though.<BR /><BR />Some more infos:<BR /><BR />root@hostname:/boot# uname -a<BR />Linux "hostname" 2.6.15-28-386 #1 PREEMPT Wed Jul 18 22:50:32 UTC 2007 i686 GNU/Linux<BR /><BR />root@hostname:/boot# uptime<BR /> 11:25:55 up 3 days, 15:07, 2 users, load average: 1.77, 3.00, 2.32<BR /><BR /><BR /><BR />root@hostname:/boot# cat /proc/interrupts<BR /> CPU0<BR /> 0: 70562794 IO-APIC-edge timer<BR /> 1: 735 IO-APIC-edge i8042<BR /> 4: 1547 IO-APIC-edge serial<BR /> 7: 0 IO-APIC-edge parport0<BR /> 8: 4 IO-APIC-edge rtc<BR /> 9: 1 IO-APIC-level acpi<BR /> 14: 312227 IO-APIC-edge ide0<BR /> 15: 5244779 IO-APIC-edge ide1<BR />177: 0 IO-APIC-level uhci_hcd:usb3<BR />185: 38190485 IO-APIC-level uhci_hcd:usb1, uhci_hcd:usb4, i915@pci:0000:00:02.0<BR />193: 37 IO-APIC-level uhci_hcd:usb2<BR />201: 4 IO-APIC-level ehci_hcd:usb5<BR />209: 0 IO-APIC-level Intel ICH5<BR />217: 553389 IO-APIC-level eth0<BR />NMI: 0<BR />LOC: 131310810<BR />ERR: 0<BR />MIS: 0<BR /><BR /><BR />Somebody has any clue about what is going on?<BR /><BR />Btw, there is no cron job for ntpdate or else... Tue, 18 Sep 2007 11:20:33 GMT OlivierDrouin 2007-09-18T11:20:33Z https://community.hpe.com/t5/operating-system-linux/clock-backward/m-p/4072167#M30072 Hi gurus,<BR /> I'm kind of lost on this one. Suppose I issue `date` every second. This is what I get:<BR /><BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:30 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:30 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:31 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:33 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:34 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:30 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:31 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:31 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:32 EDT 2007<BR />root@autogouverneur01:/boot# date<BR />Tue Sep 18 11:24:33 EDT 2007<BR /><BR />Its always +4 seconds -4 seconds and once in a while the system can go forward one more second. <BR /><BR />This is a newly installed box, running for about a week. The hardware is not new though.<BR /><BR />Some more infos:<BR /><BR />root@hostname:/boot# uname -a<BR />Linux "hostname" 2.6.15-28-386 #1 PREEMPT Wed Jul 18 22:50:32 UTC 2007 i686 GNU/Linux<BR /><BR />root@hostname:/boot# uptime<BR /> 11:25:55 up 3 days, 15:07, 2 users, load average: 1.77, 3.00, 2.32<BR /><BR /><BR /><BR />root@hostname:/boot# cat /proc/interrupts<BR /> CPU0<BR /> 0: 70562794 IO-APIC-edge timer<BR /> 1: 735 IO-APIC-edge i8042<BR /> 4: 1547 IO-APIC-edge serial<BR /> 7: 0 IO-APIC-edge parport0<BR /> 8: 4 IO-APIC-edge rtc<BR /> 9: 1 IO-APIC-level acpi<BR /> 14: 312227 IO-APIC-edge ide0<BR /> 15: 5244779 IO-APIC-edge ide1<BR />177: 0 IO-APIC-level uhci_hcd:usb3<BR />185: 38190485 IO-APIC-level uhci_hcd:usb1, uhci_hcd:usb4, i915@pci:0000:00:02.0<BR />193: 37 IO-APIC-level uhci_hcd:usb2<BR />201: 4 IO-APIC-level ehci_hcd:usb5<BR />209: 0 IO-APIC-level Intel ICH5<BR />217: 553389 IO-APIC-level eth0<BR />NMI: 0<BR />LOC: 131310810<BR />ERR: 0<BR />MIS: 0<BR /><BR /><BR />Somebody has any clue about what is going on?<BR /><BR />Btw, there is no cron job for ntpdate or else... Tue, 18 Sep 2007 11:20:33 GMT https://community.hpe.com/t5/operating-system-linux/clock-backward/m-p/4072167#M30072 OlivierDrouin 2007-09-18T11:20:33Z https://community.hpe.com/t5/operating-system-linux/clock-backward/m-p/4072168#M30073 Is it NTP running on your system, what do you have in /var/log/messages? What is your configuration in /etc/sysconfig/clock. Once I had a similar problem and setting UTC and ARC to false solved the problem, if I remember well. Tue, 18 Sep 2007 12:19:48 GMT https://community.hpe.com/t5/operating-system-linux/clock-backward/m-p/4072168#M30073 Ivan Ferreira 2007-09-18T12:19:48Z https://community.hpe.com/t5/operating-system-linux/clock-backward/m-p/4072169#M30074 Thanks,<BR /> ntp is not running. /var/log/messages doesnt show anything useful.<BR /><BR /> Even if I sync the hardware clock and soft clock with hwclock --systohc it starts to drift right away.<BR /><BR /> I think it's acpi I need to turn off but a reboot seems to have fixed the problem. I'll keep an eye on this system to see if it happens again. Tue, 18 Sep 2007 12:59:13 GMT https://community.hpe.com/t5/operating-system-linux/clock-backward/m-p/4072169#M30074 OlivierDrouin 2007-09-18T12:59:13Z https://community.hpe.com/t5/operating-system-linux/clock-backward/m-p/4072170#M30075 A reboot solved the problem.<BR /> <BR />The system is responsive and the clock has not drifted in 5 hours uptime. Could Apic be the culprit? Tue, 18 Sep 2007 18:26:47 GMT https://community.hpe.com/t5/operating-system-linux/clock-backward/m-p/4072170#M30075 OlivierDrouin 2007-09-18T18:26:47Z