![]() | |
![]() |
| | Thread Tools | Display Modes |
#1
| |||
| |||
|
#2
| |||
| |||
|
|
Hello, I see very frequent crashes of dbsrv11 on Windows 7 32 Bit (11.0.1.2355) and on Linux (Debian 64Bit 11.0.1.2324) when profiling is turned on. It occurs when we activate the detailed database tracing and select the same database as target. (I know, it's not the best way to do it, but it should still work) Here a example of the server logfile: I. 01/26 20:50:10. Diagnostic tracing has been started I. 01/26 20:50:10. Diagnostic tracing is being sent to 'links=tcpip{host=127.0.0.1};ENG=SQLSV18;DBN=xxxxx xxx' I. 01/26 20:55:10. Diagnostic tracing is being sent to 'links=tcpip{host=127.0.0.1};ENG=SQLSV18;DBN=xxxxx xxx' I. 01/26 20:57:28. SQL Anywhere Network Server Version 11.0.1.2324 I. 01/26 20:57:28. Workgroup edition I. 01/26 20:57:28. I. 01/26 20:57:28. Copyright (c) 2001-2009, iAnywhere Solutions, Inc. I. 01/26 20:57:28. Portions copyright (c) 1988-2009, Sybase, Inc. All rights reserved. I. 01/26 20:57:28. Use of this software is governed by the Sybase License Agreement. Refer to http://www.sybase.com/softwarelicenses I. 01/26 20:57:28. I. 01/26 20:57:28. 8 logical processor(s) on 1 physical processor(s) detected. I. 01/26 20:57:28. Processor limit (Workgroup Edition): 2 I. 01/26 20:57:28. Processor limit (licensed processors): 1 I. 01/26 20:57:28. Maximum number of physical processors the server will use: 1 I. 01/26 20:57:28. This server is licensed to: I. 01/26 20:57:28. xxxxxxxx I. 01/26 20:57:28. xxxxxxxx I. 01/26 20:57:28. Running Linux 2.6.26-2-amd64 #1 SMP Wed Aug 19 22:33:18 UTC 2009 on X86_64 I. 01/26 20:57:28. Server built for X86_64 processor architecture I. 01/26 20:57:28. 4194304K of memory used for caching I. 01/26 20:57:28. Minimum cache size: 4194304K, maximum cache size: 11094168K I. 01/26 20:57:28. Using a maximum page size of 8192 bytes As you can see, the dbsrv11 crashed about 60 seconds after starting tracing. it is not a problem for us, it's just nasty and probably shows a general problem since it occurs under windows and linux as well. André |
#3
| |||
| |||
|
|
Hi Andre, Thanks for reporting the problem; I am unaware of any outstanding stability issues with Application Profiling in recent builds of 11.0.1. Since this appears to be straightforward for you to reproduce, is it at all possible for you to send me steps to reproduce the problem, which may (or may not) require a copy of this database? |
|
Thanks Glenn André Schild wrote: Hello, I see very frequent crashes of dbsrv11 on Windows 7 32 Bit (11.0.1.2355) and on Linux (Debian 64Bit 11.0.1.2324) when profiling is turned on. It occurs when we activate the detailed database tracing and select the same database as target. (I know, it's not the best way to do it, but it should still work) Here a example of the server logfile: I. 01/26 20:50:10. Diagnostic tracing has been started I. 01/26 20:50:10. Diagnostic tracing is being sent to 'links=tcpip{host=127.0.0.1};ENG=SQLSV18;DBN=xxxxx xxx' I. 01/26 20:55:10. Diagnostic tracing is being sent to 'links=tcpip{host=127.0.0.1};ENG=SQLSV18;DBN=xxxxx xxx' I. 01/26 20:57:28. SQL Anywhere Network Server Version 11.0.1.2324 I. 01/26 20:57:28. Workgroup edition I. 01/26 20:57:28. I. 01/26 20:57:28. Copyright (c) 2001-2009, iAnywhere Solutions, Inc. I. 01/26 20:57:28. Portions copyright (c) 1988-2009, Sybase, Inc. All rights reserved. I. 01/26 20:57:28. Use of this software is governed by the Sybase License Agreement. Refer to http://www.sybase.com/softwarelicenses I. 01/26 20:57:28. I. 01/26 20:57:28. 8 logical processor(s) on 1 physical processor(s) detected. I. 01/26 20:57:28. Processor limit (Workgroup Edition): 2 I. 01/26 20:57:28. Processor limit (licensed processors): 1 I. 01/26 20:57:28. Maximum number of physical processors the server will use: 1 I. 01/26 20:57:28. This server is licensed to: I. 01/26 20:57:28. xxxxxxxx I. 01/26 20:57:28. xxxxxxxx I. 01/26 20:57:28. Running Linux 2.6.26-2-amd64 #1 SMP Wed Aug 19 22:33:18 UTC 2009 on X86_64 I. 01/26 20:57:28. Server built for X86_64 processor architecture I. 01/26 20:57:28. 4194304K of memory used for caching I. 01/26 20:57:28. Minimum cache size: 4194304K, maximum cache size: 11094168K I. 01/26 20:57:28. Using a maximum page size of 8192 bytes As you can see, the dbsrv11 crashed about 60 seconds after starting tracing. it is not a problem for us, it's just nasty and probably shows a general problem since it occurs under windows and linux as well. André |
#4
| |||
| |||
|
|
Am 26.01.2010 21:59, schrieb Glenn Paulley [Sybase iAnywhere]: Hi Andre, Thanks for reporting the problem; I am unaware of any outstanding stability issues with Application Profiling in recent builds of 11.0.1. Since this appears to be straightforward for you to reproduce, is it at all possible for you to send me steps to reproduce the problem, which may (or may not) require a copy of this database? The "problem" is that the workload is generated by our application and dbsrv11 crashes often when we try to profile the DB part of it. The steps we do: - Start dbsrv11 with the database file (8k page size, one log file) - Connect with sybase central to the DB Sorry, now the sybase central is in german: -> Right click on the started DB and select "Protokollierung" -> Select checkbox "Hohe Detailierung" -> Click "Next" -> Don't change anything, just click "Next" -> Now use the default "Keine neue Datenbank erstellen, Vorhandene Datenbank verwenden" (Use existing DB) -> Click "Next" -> Select "Protokollierungsdaten in dieser Datenbank speichern" (Store data in same database) -> Click "Fertigstellen" (Finish) Now we do put some load on the DB server by out application and after ~1-2 minutes it just terminates/crashes, apparently even without a assertion... It does not crash always, I did just had it 3-4 times now, always in the same condition, at least as far as I can tell... André Thanks Glenn André Schild wrote: Hello, I see very frequent crashes of dbsrv11 on Windows 7 32 Bit (11.0.1.2355) and on Linux (Debian 64Bit 11.0.1.2324) when profiling is turned on. It occurs when we activate the detailed database tracing and select the same database as target. (I know, it's not the best way to do it, but it should still work) Here a example of the server logfile: I. 01/26 20:50:10. Diagnostic tracing has been started I. 01/26 20:50:10. Diagnostic tracing is being sent to 'links=tcpip{host=127.0.0.1};ENG=SQLSV18;DBN=xxxxx xxx' I. 01/26 20:55:10. Diagnostic tracing is being sent to 'links=tcpip{host=127.0.0.1};ENG=SQLSV18;DBN=xxxxx xxx' I. 01/26 20:57:28. SQL Anywhere Network Server Version 11.0.1.2324 I. 01/26 20:57:28. Workgroup edition I. 01/26 20:57:28. I. 01/26 20:57:28. Copyright (c) 2001-2009, iAnywhere Solutions, Inc. I. 01/26 20:57:28. Portions copyright (c) 1988-2009, Sybase, Inc. All rights reserved. I. 01/26 20:57:28. Use of this software is governed by the Sybase License Agreement. Refer to http://www.sybase.com/softwarelicenses I. 01/26 20:57:28. I. 01/26 20:57:28. 8 logical processor(s) on 1 physical processor(s) detected. I. 01/26 20:57:28. Processor limit (Workgroup Edition): 2 I. 01/26 20:57:28. Processor limit (licensed processors): 1 I. 01/26 20:57:28. Maximum number of physical processors the server will use: 1 I. 01/26 20:57:28. This server is licensed to: I. 01/26 20:57:28. xxxxxxxx I. 01/26 20:57:28. xxxxxxxx I. 01/26 20:57:28. Running Linux 2.6.26-2-amd64 #1 SMP Wed Aug 19 22:33:18 UTC 2009 on X86_64 I. 01/26 20:57:28. Server built for X86_64 processor architecture I. 01/26 20:57:28. 4194304K of memory used for caching I. 01/26 20:57:28. Minimum cache size: 4194304K, maximum cache size: 11094168K I. 01/26 20:57:28. Using a maximum page size of 8192 bytes As you can see, the dbsrv11 crashed about 60 seconds after starting tracing. it is not a problem for us, it's just nasty and probably shows a general problem since it occurs under windows and linux as well. André |
#5
| |||
| |||
|
|
All right. I presume the server leaves behind a crash log and possibly a minidump. Have you sent these in to iAnywhere via dbsupport? If so, please email me the error report numbers? Glenn André Schild wrote: Am 26.01.2010 21:59, schrieb Glenn Paulley [Sybase iAnywhere]: Hi Andre, Thanks for reporting the problem; I am unaware of any outstanding stability issues with Application Profiling in recent builds of 11.0.1. Since this appears to be straightforward for you to reproduce, is it at all possible for you to send me steps to reproduce the problem, which may (or may not) require a copy of this database? The "problem" is that the workload is generated by our application and dbsrv11 crashes often when we try to profile the DB part of it. The steps we do: - Start dbsrv11 with the database file (8k page size, one log file) - Connect with sybase central to the DB Sorry, now the sybase central is in german: -> Right click on the started DB and select "Protokollierung" -> Select checkbox "Hohe Detailierung" -> Click "Next" -> Don't change anything, just click "Next" -> Now use the default "Keine neue Datenbank erstellen, Vorhandene Datenbank verwenden" (Use existing DB) -> Click "Next" -> Select "Protokollierungsdaten in dieser Datenbank speichern" (Store data in same database) -> Click "Fertigstellen" (Finish) Now we do put some load on the DB server by out application and after ~1-2 minutes it just terminates/crashes, apparently even without a assertion... It does not crash always, I did just had it 3-4 times now, always in the same condition, at least as far as I can tell... André Thanks Glenn André Schild wrote: Hello, I see very frequent crashes of dbsrv11 on Windows 7 32 Bit (11.0.1.2355) and on Linux (Debian 64Bit 11.0.1.2324) when profiling is turned on. It occurs when we activate the detailed database tracing and select the same database as target. (I know, it's not the best way to do it, but it should still work) Here a example of the server logfile: I. 01/26 20:50:10. Diagnostic tracing has been started I. 01/26 20:50:10. Diagnostic tracing is being sent to 'links=tcpip{host=127.0.0.1};ENG=SQLSV18;DBN=xxxxx xxx' I. 01/26 20:55:10. Diagnostic tracing is being sent to 'links=tcpip{host=127.0.0.1};ENG=SQLSV18;DBN=xxxxx xxx' I. 01/26 20:57:28. SQL Anywhere Network Server Version 11.0.1.2324 I. 01/26 20:57:28. Workgroup edition I. 01/26 20:57:28. I. 01/26 20:57:28. Copyright (c) 2001-2009, iAnywhere Solutions, Inc. I. 01/26 20:57:28. Portions copyright (c) 1988-2009, Sybase, Inc. All rights reserved. I. 01/26 20:57:28. Use of this software is governed by the Sybase License Agreement. Refer to http://www.sybase.com/softwarelicenses I. 01/26 20:57:28. I. 01/26 20:57:28. 8 logical processor(s) on 1 physical processor(s) detected. I. 01/26 20:57:28. Processor limit (Workgroup Edition): 2 I. 01/26 20:57:28. Processor limit (licensed processors): 1 I. 01/26 20:57:28. Maximum number of physical processors the server will use: 1 I. 01/26 20:57:28. This server is licensed to: I. 01/26 20:57:28. xxxxxxxx I. 01/26 20:57:28. xxxxxxxx I. 01/26 20:57:28. Running Linux 2.6.26-2-amd64 #1 SMP Wed Aug 19 22:33:18 UTC 2009 on X86_64 I. 01/26 20:57:28. Server built for X86_64 processor architecture I. 01/26 20:57:28. 4194304K of memory used for caching I. 01/26 20:57:28. Minimum cache size: 4194304K, maximum cache size: 11094168K I. 01/26 20:57:28. Using a maximum page size of 8192 bytes As you can see, the dbsrv11 crashed about 60 seconds after starting tracing. it is not a problem for us, it's just nasty and probably shows a general problem since it occurs under windows and linux as well. André |
![]() |
| Thread Tools | |
| Display Modes | |
| |