MustGather: Data Collection for an application hang and system hangs on Windows
How do I collect the necessary output for support to analyze application and system hangs that occur on Microsoft Windows related to IBM Rational processes?
STEP 1: Identify the problem
- WHICH process is hanging? (Example: clearexplorer.exe, db_server.exe ....)
- WHAT operation was being performed when the hang occurred? (Example: shutting down GUI, performing X operation ....)
- HOW often do the errors occur? (Example: every night at X time, every day around X time ....)
- HOW many machines does this happen on? (Example: ALL, one or X number)
STEP 2: Collect relevant logs
Collect the logs from a good machine and a bad machine and compare the two for differences.
For example review technote 1131335 for more information on collecting ClearCase logs from Windows.
STEP 3: Collecting the User or Memory Dump for the hang
The following tools can be used to help collect data about application or system hangs. The instructions for each of the tools are provided below:
3a) Collecting application hangs (Send the dump file to IBM Rational Client Support for analysis)
- Collecting user dumps for Windows XP and Windows 2003:
Refer to the Microsoft KB241215 .
- Collecting user dumps for Microsoft Windows Vista and later:
- Using GenMiniDump.exe (Note: GenMiniDump.zip is attached below):
- Place genminidump.exe and dbghelp.dll in a folder
- cd into that folder
- Run the following from cmd: genminidump.exe <PID> or genminidump.exe <Process name> Example: C:\Temp\GenMiniDump\genminidump msdev.exe
- Using .dump command in windbg
- Download windbg:
- Attach to process and run .dump <filename> or .dump /ma <filename>
All source code and/or binaries attached to this document are referred to here as "the Program". IBM is not providing program services of any kind for the Program. IBM is providing the Program on an "AS IS" basis without warranty of any kind. IBM WILL NOT BE LIABLE FOR ANY ACTUAL, DIRECT, SPECIAL, INCIDENTAL, OR INDIRECT DAMAGES OR FOR ANY ECONOMIC CONSEQUENTIAL DAMAGES (INCLUDING LOST PROFITS OR SAVINGS), EVEN IF IBM, OR ITS RESELLER, HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.
Review MS KB article 254649 for an overview of memory dump file options for Windows Server 2003, Windows XP, and Windows 2000.
For Windows Vista, Server 2008, Windows 7 and later, refer to Forcing a System Crash from the Keyboard
- Click Start > Settings > Control Panel and double-click System
- On the Advanced tab, click Settings in the Startup and Recovery section.
- Ensure the Write Debugging Information drop down box has Kernel memory dump selected and note the Dump File: location window as this is where the dump file(s) will be saved.
Note: Only collect a complete memory dump when request by IBM Rational Client Support.
- Ensure Overwrite any existing file is checked.
- Ensure the Automatically restart box is checked.
Example Kernel memory dump screenshot (from XP):
Example Complete memory dump screenshot (from XP):
CONFIGURE STOP EVENT:
To configure the Stop Event, you will need to make a registry change on the host.
To add an entry to the registry:
- Click Start > Run > type: regedit
- Navigate to the subkey
- Select the Parameters subkey, and then click Add Value from the Edit menu.
- In the Add Value dialog box, in the Value Name box type CrashOnCtrlScroll, in the Data Type box select REG_DWORD, and then click OK.
- Double-click the CrashOnCtrlScroll entry. In the DWORD Editor dialog box, type 1 in the Data box and then click OK.
- Close Regedit, and restart your computer. After the computer has restarted, you can generate a stop event.
Refer to Microsoft KB article 244139 Windows feature lets you generate a memory dump file by using the keyboard for further details.
GENERATE STOP EVENT:
Press and hold the right Ctrl key (the left Ctrl key will not allow you to generate a stop event), and then press the Scroll Lock (ScrLk) key twice.
A stop event will occur, and a stop error will be displayed with the following message:
*** STOP: 0x000000E2 (0x00000000, 0x00000000, 0x00000000, 0x00000000)
The end-user manually generated the crashdump.
After generating a stop event, the computer will be inoperable until it has been restarted. After the restart, complete functionality will be restored.
Send the dump file to IBM Rational Client Support for analysis.
More support for:
Rational Customer Support
Software version: All Versions
Operating system(s): Windows
Reference #: 1366397
Modified date: 2013-12-05