The postings on this site are my own and do not represent my Employer's positions, advice or strategies.

LifeAsBob - Blog

 

Home

No Ads ever, except search!
Tuesday, April 23, 2024 Login
Public

4 problems - 1 Bug, CPU Drift 1/3/2008 2:55:22 PM

Chasing 4 problems.

  1. Polyserve failing / restarting SQL Instances due to no network traffic.
  2. Incorrect durations reporting in SQL Server traces (very large numbers).
  3. Information message logged in SQL Server 2005 log file, The time stamp counter of CPU on scheduler id 2 is not synchronized with other CPUs.
  4. Alert from Monitoring on, Windows cannot obtain the domain controller name for your computer network. (An unexpected network error occurred.). Group Policy processing aborted.
  5. I've been chasing these errors, without much support from network or other engineers, basically the same as the mechanic who tells you the engine light is on, but all is ok !

Well finally found a customer advisory from HP and all 4 problems roll up to the same issue:

----------------------------

SUPPORT COMMUNICATION - CUSTOMER ADVISORY

Document ID: c01075682

Version: 2
Advisory: (Revision) HP ProLiant Servers Using Dual-Core or More Than One Single-Core AMD Opteron Processor May Experience Incorrect Operating System Time When Running Systems That Use the System Time Stamp Counter
NOTICE: The information in this document, including products and software versions, is current as of the Release Date. This document is subject to change without notice.

Release Date: 2007-07-16

Last Updated: 2007-07-16


DESCRIPTIONDocument Version Release Date Details
2 07/16/2007 Added Sun Solaris information.
1 06/08/2007 Original Document Release.

HP ProLiant servers configured with Dual-Core or with more than one single-core AMD Opteron processor may encounter Time Stamp Counter (TSC) drift in certain conditions. The TSC is used by some operating systems as a timekeeping source. Each processor core, whether it is a single-core processor or a dual-core processor, includes a TSC. The condition where the TSC for different processor cores becomes unsynchronized is known as TSC drift.

Note : The potential for TSC drift if the proper recommendations are not applied when using AMD Opteron 200-series, Opteron 800-series, Opteron 1200-series, Opteron 2200-series and Opteron 8200-series processors is not specific to HP ProLiant servers.

Whether or not the system is affected by TSC drift depends on the specific ProLiant server generation, the number and type of AMD Opteron processors installed, the operating system, and whether the AMD PowerNow! feature is being utilized. TSC drift can result in different symptoms and behaviors based on the operating system environment, as detailed below:

Microsoft Windows Server 2003
This condition affects operations such as network communications and performance monitoring tasks that are sensitive to system time. For example, Microsoft Active Directory domain controllers can report an Unexpected Network Error (Event ID 1054) with the following description:

Event Description:
Windows cannot obtain the domain controller name for your computer network. (An unexpected network error occurred.). Group Policy processing aborted.

In addition, a negative PING time or larger than actual PING time may be returned after issuing the PING command. The negative PING time occurs because of a Time Stamp Counter drift occurring on AMD Opteron platforms which include more than one processor core.

Red Hat Enterprise Linux, SUSE Linux Enterprise Server and Sun Solaris
Earlier releases of Red Hat Enterprise Linux 4, SUSE Linux Enterprise Server 9 and Sun Solaris 10 will default to using the Time Stamp Counter as the default time source for gettimeofday() calls. When the time stamp counter is used, the server may exhibit some inconsistent timekeeping and the following symptoms may be observed:

When a command such as "date" is typed, an incorrect system time may be displayed.
The kernel may report an error similar to the following:
kernel: Your time source seems to be instable or some driver is hogging interrupts

Newer operating systems typically do not use the TSC by default if other timers are available in the system which can be used as a timekeeping source. Other available timers include the PM_Timer and the High Precision Event Timer (HPET). All HP ProLiant servers include the PM_Timer, and the latest generation of HP ProLiant servers supporting AMD Opteron 2200-series and 8200-series processors support HPET. These timers are not affected by this condition. New operating systems such as Red Hat Enterprise Linux (RHEL) 5, SUSE Linux Enterprise Server (SLES) 10, and Microsoft Windows Server 2008 (codename Longhorn) are not affected by this issue.

Note: Some applications (e.g., Microsoft SQL Server 2005) use the Time Stamp Counter even though the operating system is configured to use a different timer as the timekeeping source. To determine if a specific application uses the TSC as the timekeeping source, contact the software vendor.

SCOPE

Any HP ProLiant server configured with more than one single-core AMD Opteron processor or configured with one (or more) dual-core AMD Opteron processors running the following operating systems:

Microsoft Windows Server 2003 (any edition)
Microsoft Windows Server 2003 x64 Edition (any edition)
Red Hat Enterprise Linux 4(x86) or earlier
Red Hat Enterprise Linux 4 (AMD64/EM64T) or earlier
SUSE Linux Enterprise Server 9 32-bit (x86) or earlier
SUSE Linux Enterprise Server 9 64-bit (AMD64/EM64T) or earlier
Sun Solaris 9
Sun Solaris 10 3/05 (32/64 bit)
VMware ESX Server 2.5.4 (or earlier)

Note: VMware ESX Server 2.5.4 with the January 2007 (or later) patch is not affected. VMware ESX Server 3.0.0 (or later) uses an alternate mechanism for timekeeping and is not affected by the potential TSC drift.

Note : The issue does not affect systems with only one single-core processor installed.

The following servers are affected when running an affected operating system:

HP ProLiant BL465c Blade Server
HP ProLiant BL685c Blade Server
HP ProLiant BL25p G2 server
HP ProLiant BL45p G2 server
HP ProLiant DL145 G3 server
HP ProLiant DL385 G2 server
HP ProLiant DL585 G2 server
HP ProLiant DL365 server
HP ProLiant ML115 server

The following servers are affected ONLY when using the AMD PowerNow! feature and running an affected operating system:

ProLiant BL25p Blade Server
HP ProLiant BL45p Blade Server
HP ProLiant DL145 G2 server
HP ProLiant DL385 server
HP ProLiant DL585 server

The following operating systems are not affected by TSC drift because these operating systems do not use the TSC as a timekeeping source:

Microsoft Windows Server 2008 (codename Longhorn)
Red Hat Enterprise Linux 5 (x86)
Red Hat Enterprise Linux 5 (AMD64/EM64T)
SUSE Linux Enterprise Server 10 (x86)
SUSE Linux Enterprise Server 10 (AMD64/EM64T)
VMware ESX Server 3.0.0 (or later)

RESOLUTION

To ensure proper operation of tasks sensitive to system time, perform either of the following actions, based on the operating system environment:

Microsoft Windows Server 2003 (any edition)
Edit the BOOT.ini file and add the parameter "/usepmtimer," then reboot the server. Adding the "/usepmtimer" parameter to the BOOT.INI file configures the Windows operating system to use the PM_TIMER, rather than the Time Stamp Counter.

Note: When installing the AMD Opteron Processor with AMD PowerNow! Technology driver Version 1.3.2.16 (or later) from AMD, the BOOT.INI file will automatically be updated with the "/usepmtimer" parameter. While the driver itself does not resolve this issue, the installation process will make the necessary changes to the BOOT.INI file to prevent the issue from occurring.

Red Hat Enterprise Linux 4 or SUSE Linux
Add the boot parameter "clock=pmtmr" to the /boot/grub/menu.lst file. Adding the "clock=pmtmr" to the /boot/grub/menu.lst file configures the operating system to use the PM_TIMER, rather than the Time Stamp Counter.

Sun Solaris
If using Sun Solaris 10 3/05 apply the 1/06 (Update 1) Patch (or later). To locate the latest version of the Solaris 10 patch, click on the following Sun Microsystems URL, and click on the desired patch:

http://www.sun.com/downloads

VMware
If using VMware ESX Server 2.5.4, update to the January 2007 Patch (or later). To locate the latest version of the ESX Server 2.5.4 patch, click on the following VMware URL, and click on the desired patch.

http://www.vmware.com/download/esx/esx2_patches.html#c4317 RECEIVE PROACTIVE UPDATES : Receive support alerts (such as Customer Advisories), as well as updates on drivers, software, firmware, and customer replaceable components, proactively via e-mail through HP Subscriber's Choice. Sign up for Subscriber's Choice at the following URL:
http://www.hp.com/go/myadvisory

SEARCH TIP : For hints on locating similar documents on HP.com, refer to the Search Tips document: http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?objectID=c00638154 .
To search for additional advisories related to System Time, use the following search string:
+ProLiant +Advisory +System Time
KEYWORDS: time sync, clock, track time

-------------------------------------

http://support.microsoft.com/kb/931279/en-us 


 

 


Blog Home