Hello, My Windows 2008 R2x64 server is running a SA11.0.1.2627 database server. For no reason, the server crashed with no log information about the crash. There is nothing at console log file and no mini dump file was generated. The only log available is the Windows Event Viewer Application Log, copied as follows: Faulting application name: dbsrv11.exe, version: 11.0.1.2627, time stamp: 0x4e0536dd Faulting module name: dbserv11.dll, version: 11.0.1.2627, time stamp: 0x4e0536c2 Exception code: 0xc0000417 Fault offset: 0x00000000005a6c38 Faulting process id: 0xaf0 Faulting application start time: 0x01cc39b5b67152d6 Faulting application path: C:\Program Files\SQL Anywhere 11\Bin64\dbsrv11.exe Faulting module path: C:\Program Files\SQL Anywhere 11\Bin64\dbserv11.dll Report Id: 56adb6c4-bc59-11e0-ae97-00155d830172 Any Clue? Thanks. asked 01 Aug '11, 13:46 Thiago Reis Mark Culp |
The problem was solved by Sybase Engineering as folows: QTS #681398 - Server could crash while using UNLOAD statement to unload data to a file. The correction was addressed at build #2661. Thanks!!! EDIT: Full description (with interesting details, methinks):
answered 22 Sep '11, 06:46 Thiago Reis Volker Barth Further details about this resolution can be found here: http://search.sybase.com/kbx/changerequests?bug_id=681398
(26 Sep '11, 10:43)
Jeff Albion
Thiago, you should accept this post as the answer to clearly show others what the solution was.
(26 Sep '11, 10:49)
Martin
|
Thiago, This may sound a bit old-school, but hardware failures can cause otherwise-unexplainable application and system crashes. Unhandled memory errors are the most common hardware source, but blown-out capacitors on the system board have also caused problems. If you can afford some system down-time, try testing the memory and visually inspecting for failed capacitors (look for bulged ends or any sign of fluid leakage). answered 17 Sep '11, 21:00 William Clardy |
This is a STATUS_INVALID_CRUNTIME_PARAMETER exception. You say, that no dump is created, so it might be, that the exception is thrown as a result of trying to call _invoke_watson to dump the real exception. Which would mean, that the information necessary for Sybase to identify any flaw is obscurred by this subsequent error. answered 20 Sep '11, 03:00 Martin |
Hello Thiago,
How often is this issue occurring? Is it reproducible?
This is definitely a server crash and we will need to obtain a memory dump from the server in order to provide more context about the reason for the crash. We unfortunately will not be able to diagnose the issue from the above posted information.
Upon a crash, SQL Anywhere 11 should automatically create a (mini) memory dump file in %SADIAGDIR% (usually "%ALLUSERSPROFILE%Application DataSQL Anywhere 11diagnostics" ). If dbsupport is not installed / configured, it will attempt to create the dump file in the same directory as the database file, or possibly the TEMP directory.
See: http://dcx.sybase.com/index.html#1101/en/dbadmin_en11/da-envvar-s-4186730.html
http://dcx.sybase.com/index.html#1101/en/dbadmin_en11/dbserver-s-3394455.html
You can submit the dump file using "dbsupport -sa" and post the submission ID here so we can quickly check for you if this is a known issue.
If this is an urgent issue, we would encourage you to work directly with technical support by opening a technical support case and upload the crash files directly to us.
Hi Jeff,
In two months running on production, this is the first time the server has crashed WITHOUT generating a Mini Dump file.
In other circunstances the mini dump file was generated sucessfully. But not this time.
Unfortunatelly I can´t reproduce the problem.
I guess I´ll have to wait another crash and try to reproduce it.
Thank you!