Instructions on how to troubleshoot application crashes
This article applies to:
Introduction Under specific circumstances software may crash. It is very important to report a crash instead of trying to workaround it. Normal workflow for a crash problem is the following:
Crash happens. Support reports crash to developers and provides them crash dumps. Developers investigate the dumps and find root causes. Fixes are implemented. What information needs to be gathered:
Step by step description of the issue Log of the operation which caused the crash Crash dump. Collect System Report from the machine in question: see instruction for Acronis Backup & Recovery 11.5 Generating System Report , if this fails for any reason, you can use this AcronisInfo Utility Solution Follow these steps to troubleshoot application crashes depending on the operating system you are on:
How to spot a crash:
Windows Application log will have a similar error for an Acronis process or service: Operations fail with similar errors: The activity state has been repaired after unexpected failure. The server was probably unavailable. An underlying layer has disconnected by peer. Product closes unexpectedly Service starts and stops unexpectedly. For the list Acronis processes please check the following articles: Acronis True Image 2014: Processes and Services Acronis Backup & Recovery 11.5: Windows Services (plus service_process.exe and mediabuilder.exe)
How to collect a crash dump:
Use the ProcDump Utility:
Download ProcDump from Windows Sysinternals site; Create a folder where dumps will be stored (e.g. C:\Dumps\); Unzip the archive and put the procdump.exe in to the created directory; Open Windows command-line: Hit Start -> Run and type in cmd ; In CMD, switch to the newly created folder using the cd command:
cd <path_to_folder>
For example: cd C:\Dumps
Issue the following command:
procdump -e -w -ma <process_name>
Reproduce the crash. Collect crash dump in the same folder where you put ProcDump. Sometimes there are issues when you need to collect dump of "service_process" that crashes randomly. If you use the ProcDump utility in this case and the service_process does not crash on the first run, ProcDump will exit with "process exited" without collecting the dump. In this case you should use automatic dump creation, see Acronis Backup & Recovery 11.5: Enabling Automatic Dump Creation for Acronis Processes or True Image 2013 by Acronis and Acronis True Image 2014: Collecting Application Dump .
How to spot a crash:
Operations fail with similar errors: The activity state has been repaired after unexpected failure. The server was probably unavailable. Underlying layer has disconnected by peer.
Syslog (/var/log/messages) of the machine MUST have something similar: Nov 22 08:39:32 localhost kernel: mms[15295]: segfault at 0 ip 00007ffe6210fa79 sp 00007fff87d2a7e0 error 4 in libDiskBundle.so[7ffe6143c000+240f000] or
How to collect a crash dump:
To prepare the operating system to create a core dump user should run the following command under the root account: #ulimit -c unlimited After an application crash, a core dump will be automatically created in the CURRENT directory. For more information see Acronis Linux Products: Collecting Core Dumps of Acronis Services in Linux
How to spot a crash:
Appliance reboots.
Appliance will have a file named “core” (in Acronis Backup & Recovery in the "/" folder, in vmProtect it is in /tmp).
How to collect a crash dump:
Install the WinSCP utility which you can get from http://winscp.net/eng/download.php . Log in to vSphere Client and open the console of Acronis Virtual Appliance. Write down or remember the IP address of it. Press CTRL+SHIFT+ALT+SPACE+F1 in the Virtual Appliance of Acronis Backup & Recovery 11.5 (or CTRL+SHIFT+ALT+SPACE+F1 and then ALT+F2 in vmProtect). This will bring you to the command line interface; In the console please type this command: /bin/sshd Launch the WinSCP utility on the machine where you installed it; In the host machine field type your Acronis Virtual Appliance IP (recorded on Step 2), the same as in the vSphere console. The default login:password is root:root (the password may differ in Acronis vmProtect if you changed it via web console at Configure -> Agent Password ). The protocol should be set to SCP and then click on log in. You will get a window with all the directories of the Appliance (Windows Commander-like interface), collect the core file from "/" in the Acronis Backup & Recovery Virtual Appliance (/tmp/core file in the vmProtect Appliance).
How to spot a crash:
The media will unexpectedly reboot.
How to collect a crash dump:
If the media crashes do the following:
Boot to the Acronis Boot Selection screen: Click F11 Remove the word “quiet”. Boot the appropriate media. The media will boot to shell. /bin/sshd to start ssh demon. /bin/product It will start product GUI. Reproduce the crash. Now the media will not reboot but hang or exit to shell. When the crash is reproduced (media gets unresponsive), try to click CTRL+SHIFT+F2 to get to shell. Note: If that does not work, that is not an application crash, but a kernel crash (follow this article) ls / Check if you can see any “core” files in the root folder. If yes, please use WinSCP to download the core file: Install the WinSCP utility which you can get from http://winscp.net/eng/download.php Launch the WinSCP utility on the machine where you installed it. In the host machine field type your Acronis Bootable Media IP. The protocol should be set to SCP. Credentials are root:root Collect the core file from the root directory ("/")
With the information collected contact Acronis Customer Central .