![]() | |
#21
| ||||||
| ||||||
|
|
Nan, 1. Could we possibly see the output of a 'setenv' before starting the MobiLink server? For now, try starting MobiLink from the shell command-line instead of your shell script. (Once we finish investigating the crash, you can go back to the script). my environment variables: |
|
2. Assume that the initial configuration is correct: - That /disk2/sqlanywhere10/MobiLink/setup/syncora.sql has been run against the Oracle server. - $ORACLE_HOME = /disk2/oracle9i/app/ora9.2 - $ODBCINI = /path/to/.odbc.ini - $ORACLE_HOME/network/admin/tnsnames.ora is a valid configuration |
.
. but I|
i.e. iasora9 = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = xxx.xxx.xxx.xxx)(PORT = xxxx)) ) (CONNECT_DATA = (SERVICE_NAME = orcl) ) ) - That the .odbc.ini reads: [iasora9] Driver=/disk2/sqlanywhere10/lib32/libdboraodbc10_r.so UserID=mobilink Password=mobilink SID=iasora9 ProcResults=No ArraySize=60000 ** NOTE: *** There should be a leading forward-slash to the DRIVER parameter. That wasn't present in the initial post information! ** NOTE: *** ProcResults=No, not "Yes". See: http://www.sybase.com/detail?id=1051962#_Oracle_8_and ================================================== ================ If all of that is correct, we probably need someone to investigate *what* is crashing exactly. Best way (in this scenario specifically, which is a MobiLink crash) is probably to gather a core dump. ================================================== ================ On AIX, to gather a core dump: ------------------------------- * Note: Users can set their ulimit maximums up to the hard limits defined in the /etc/security/limits file that is set by the sysadmin. If you need to exceed those limits, you must have root access to change that file. ------------------------------- 1. Check to see if full core dumps are enabled: % lsattr -El sys0 -a fullcore ** This should return: fullcore true Enable full CORE dump True *** If it doesn't, type (possibly as root): % chdev -l sys0 -a fullcore=true |
|
2. Check the output of "ulimit -a" % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) 2097151 nofiles(descriptors) 2000 *** Check to ensure that "coredump(blocks)" is set to a value big enough to hold the MobiLink memory space. If it isn't, try setting it to "unlimited": % ulimit -c unlimited % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) unlimited nofiles(descriptors) 2000 I changed the configuration of my AIX as above: |
|
3. Have MobiLink produce a core file: % mlsrv10 -c "DSN=iasora9" |
, the information on screen is:|
*** It crashes in the current directory, then should produce a file named "core" or "core.pid" or similar). 4. Submit the core file to Technical Support. If this is a priority issue, please call 1-800-8SYBASE to open a technical support case (using an existing support plan or open a pay-per-incident case). If this isn't a priority issue, you can submit it to http://case-express.sybase.com (under "Create Bug Report Case" on the left-hand side). Please submit the *core* file, the EXACT *version* and *build* of MobiLink (10.0.1.3609) and your operating system (uname -a). Please keep the core file, as we may need you to generate the stack traces and memory map on your AIX box. Regards, David Fishburn wrote: |
#22
| ||||||
| ||||||
|
|
Nan, 1. Could we possibly see the output of a 'setenv' before starting the MobiLink server? For now, try starting MobiLink from the shell command-line instead of your shell script. (Once we finish investigating the crash, you can go back to the script). my environment variables: |
|
2. Assume that the initial configuration is correct: - That /disk2/sqlanywhere10/MobiLink/setup/syncora.sql has been run against the Oracle server. - $ORACLE_HOME = /disk2/oracle9i/app/ora9.2 - $ODBCINI = /path/to/.odbc.ini - $ORACLE_HOME/network/admin/tnsnames.ora is a valid configuration |
.
. but I|
i.e. iasora9 = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = xxx.xxx.xxx.xxx)(PORT = xxxx)) ) (CONNECT_DATA = (SERVICE_NAME = orcl) ) ) - That the .odbc.ini reads: [iasora9] Driver=/disk2/sqlanywhere10/lib32/libdboraodbc10_r.so UserID=mobilink Password=mobilink SID=iasora9 ProcResults=No ArraySize=60000 ** NOTE: *** There should be a leading forward-slash to the DRIVER parameter. That wasn't present in the initial post information! ** NOTE: *** ProcResults=No, not "Yes". See: http://www.sybase.com/detail?id=1051962#_Oracle_8_and ================================================== ================ If all of that is correct, we probably need someone to investigate *what* is crashing exactly. Best way (in this scenario specifically, which is a MobiLink crash) is probably to gather a core dump. ================================================== ================ On AIX, to gather a core dump: ------------------------------- * Note: Users can set their ulimit maximums up to the hard limits defined in the /etc/security/limits file that is set by the sysadmin. If you need to exceed those limits, you must have root access to change that file. ------------------------------- 1. Check to see if full core dumps are enabled: % lsattr -El sys0 -a fullcore ** This should return: fullcore true Enable full CORE dump True *** If it doesn't, type (possibly as root): % chdev -l sys0 -a fullcore=true |
|
2. Check the output of "ulimit -a" % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) 2097151 nofiles(descriptors) 2000 *** Check to ensure that "coredump(blocks)" is set to a value big enough to hold the MobiLink memory space. If it isn't, try setting it to "unlimited": % ulimit -c unlimited % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) unlimited nofiles(descriptors) 2000 I changed the configuration of my AIX as above: |
|
3. Have MobiLink produce a core file: % mlsrv10 -c "DSN=iasora9" |
, the information on screen is:|
*** It crashes in the current directory, then should produce a file named "core" or "core.pid" or similar). 4. Submit the core file to Technical Support. If this is a priority issue, please call 1-800-8SYBASE to open a technical support case (using an existing support plan or open a pay-per-incident case). If this isn't a priority issue, you can submit it to http://case-express.sybase.com (under "Create Bug Report Case" on the left-hand side). Please submit the *core* file, the EXACT *version* and *build* of MobiLink (10.0.1.3609) and your operating system (uname -a). Please keep the core file, as we may need you to generate the stack traces and memory map on your AIX box. Regards, David Fishburn wrote: |
#23
| ||||||
| ||||||
|
|
Nan, 1. Could we possibly see the output of a 'setenv' before starting the MobiLink server? For now, try starting MobiLink from the shell command-line instead of your shell script. (Once we finish investigating the crash, you can go back to the script). my environment variables: |
|
2. Assume that the initial configuration is correct: - That /disk2/sqlanywhere10/MobiLink/setup/syncora.sql has been run against the Oracle server. - $ORACLE_HOME = /disk2/oracle9i/app/ora9.2 - $ODBCINI = /path/to/.odbc.ini - $ORACLE_HOME/network/admin/tnsnames.ora is a valid configuration |
.
. but I|
i.e. iasora9 = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = xxx.xxx.xxx.xxx)(PORT = xxxx)) ) (CONNECT_DATA = (SERVICE_NAME = orcl) ) ) - That the .odbc.ini reads: [iasora9] Driver=/disk2/sqlanywhere10/lib32/libdboraodbc10_r.so UserID=mobilink Password=mobilink SID=iasora9 ProcResults=No ArraySize=60000 ** NOTE: *** There should be a leading forward-slash to the DRIVER parameter. That wasn't present in the initial post information! ** NOTE: *** ProcResults=No, not "Yes". See: http://www.sybase.com/detail?id=1051962#_Oracle_8_and ================================================== ================ If all of that is correct, we probably need someone to investigate *what* is crashing exactly. Best way (in this scenario specifically, which is a MobiLink crash) is probably to gather a core dump. ================================================== ================ On AIX, to gather a core dump: ------------------------------- * Note: Users can set their ulimit maximums up to the hard limits defined in the /etc/security/limits file that is set by the sysadmin. If you need to exceed those limits, you must have root access to change that file. ------------------------------- 1. Check to see if full core dumps are enabled: % lsattr -El sys0 -a fullcore ** This should return: fullcore true Enable full CORE dump True *** If it doesn't, type (possibly as root): % chdev -l sys0 -a fullcore=true |
|
2. Check the output of "ulimit -a" % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) 2097151 nofiles(descriptors) 2000 *** Check to ensure that "coredump(blocks)" is set to a value big enough to hold the MobiLink memory space. If it isn't, try setting it to "unlimited": % ulimit -c unlimited % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) unlimited nofiles(descriptors) 2000 I changed the configuration of my AIX as above: |
|
3. Have MobiLink produce a core file: % mlsrv10 -c "DSN=iasora9" |
, the information on screen is:|
*** It crashes in the current directory, then should produce a file named "core" or "core.pid" or similar). 4. Submit the core file to Technical Support. If this is a priority issue, please call 1-800-8SYBASE to open a technical support case (using an existing support plan or open a pay-per-incident case). If this isn't a priority issue, you can submit it to http://case-express.sybase.com (under "Create Bug Report Case" on the left-hand side). Please submit the *core* file, the EXACT *version* and *build* of MobiLink (10.0.1.3609) and your operating system (uname -a). Please keep the core file, as we may need you to generate the stack traces and memory map on your AIX box. Regards, David Fishburn wrote: |
#24
| ||||||
| ||||||
|
|
Nan, 1. Could we possibly see the output of a 'setenv' before starting the MobiLink server? For now, try starting MobiLink from the shell command-line instead of your shell script. (Once we finish investigating the crash, you can go back to the script). my environment variables: |
|
2. Assume that the initial configuration is correct: - That /disk2/sqlanywhere10/MobiLink/setup/syncora.sql has been run against the Oracle server. - $ORACLE_HOME = /disk2/oracle9i/app/ora9.2 - $ODBCINI = /path/to/.odbc.ini - $ORACLE_HOME/network/admin/tnsnames.ora is a valid configuration |
.
. but I|
i.e. iasora9 = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = xxx.xxx.xxx.xxx)(PORT = xxxx)) ) (CONNECT_DATA = (SERVICE_NAME = orcl) ) ) - That the .odbc.ini reads: [iasora9] Driver=/disk2/sqlanywhere10/lib32/libdboraodbc10_r.so UserID=mobilink Password=mobilink SID=iasora9 ProcResults=No ArraySize=60000 ** NOTE: *** There should be a leading forward-slash to the DRIVER parameter. That wasn't present in the initial post information! ** NOTE: *** ProcResults=No, not "Yes". See: http://www.sybase.com/detail?id=1051962#_Oracle_8_and ================================================== ================ If all of that is correct, we probably need someone to investigate *what* is crashing exactly. Best way (in this scenario specifically, which is a MobiLink crash) is probably to gather a core dump. ================================================== ================ On AIX, to gather a core dump: ------------------------------- * Note: Users can set their ulimit maximums up to the hard limits defined in the /etc/security/limits file that is set by the sysadmin. If you need to exceed those limits, you must have root access to change that file. ------------------------------- 1. Check to see if full core dumps are enabled: % lsattr -El sys0 -a fullcore ** This should return: fullcore true Enable full CORE dump True *** If it doesn't, type (possibly as root): % chdev -l sys0 -a fullcore=true |
|
2. Check the output of "ulimit -a" % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) 2097151 nofiles(descriptors) 2000 *** Check to ensure that "coredump(blocks)" is set to a value big enough to hold the MobiLink memory space. If it isn't, try setting it to "unlimited": % ulimit -c unlimited % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) unlimited nofiles(descriptors) 2000 I changed the configuration of my AIX as above: |
|
3. Have MobiLink produce a core file: % mlsrv10 -c "DSN=iasora9" |
, the information on screen is:|
*** It crashes in the current directory, then should produce a file named "core" or "core.pid" or similar). 4. Submit the core file to Technical Support. If this is a priority issue, please call 1-800-8SYBASE to open a technical support case (using an existing support plan or open a pay-per-incident case). If this isn't a priority issue, you can submit it to http://case-express.sybase.com (under "Create Bug Report Case" on the left-hand side). Please submit the *core* file, the EXACT *version* and *build* of MobiLink (10.0.1.3609) and your operating system (uname -a). Please keep the core file, as we may need you to generate the stack traces and memory map on your AIX box. Regards, David Fishburn wrote: |
#25
| ||||||
| ||||||
|
|
Nan, 1. Could we possibly see the output of a 'setenv' before starting the MobiLink server? For now, try starting MobiLink from the shell command-line instead of your shell script. (Once we finish investigating the crash, you can go back to the script). my environment variables: |
|
2. Assume that the initial configuration is correct: - That /disk2/sqlanywhere10/MobiLink/setup/syncora.sql has been run against the Oracle server. - $ORACLE_HOME = /disk2/oracle9i/app/ora9.2 - $ODBCINI = /path/to/.odbc.ini - $ORACLE_HOME/network/admin/tnsnames.ora is a valid configuration |
.
. but I|
i.e. iasora9 = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = xxx.xxx.xxx.xxx)(PORT = xxxx)) ) (CONNECT_DATA = (SERVICE_NAME = orcl) ) ) - That the .odbc.ini reads: [iasora9] Driver=/disk2/sqlanywhere10/lib32/libdboraodbc10_r.so UserID=mobilink Password=mobilink SID=iasora9 ProcResults=No ArraySize=60000 ** NOTE: *** There should be a leading forward-slash to the DRIVER parameter. That wasn't present in the initial post information! ** NOTE: *** ProcResults=No, not "Yes". See: http://www.sybase.com/detail?id=1051962#_Oracle_8_and ================================================== ================ If all of that is correct, we probably need someone to investigate *what* is crashing exactly. Best way (in this scenario specifically, which is a MobiLink crash) is probably to gather a core dump. ================================================== ================ On AIX, to gather a core dump: ------------------------------- * Note: Users can set their ulimit maximums up to the hard limits defined in the /etc/security/limits file that is set by the sysadmin. If you need to exceed those limits, you must have root access to change that file. ------------------------------- 1. Check to see if full core dumps are enabled: % lsattr -El sys0 -a fullcore ** This should return: fullcore true Enable full CORE dump True *** If it doesn't, type (possibly as root): % chdev -l sys0 -a fullcore=true |
|
2. Check the output of "ulimit -a" % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) 2097151 nofiles(descriptors) 2000 *** Check to ensure that "coredump(blocks)" is set to a value big enough to hold the MobiLink memory space. If it isn't, try setting it to "unlimited": % ulimit -c unlimited % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) unlimited nofiles(descriptors) 2000 I changed the configuration of my AIX as above: |
|
3. Have MobiLink produce a core file: % mlsrv10 -c "DSN=iasora9" |
, the information on screen is:|
*** It crashes in the current directory, then should produce a file named "core" or "core.pid" or similar). 4. Submit the core file to Technical Support. If this is a priority issue, please call 1-800-8SYBASE to open a technical support case (using an existing support plan or open a pay-per-incident case). If this isn't a priority issue, you can submit it to http://case-express.sybase.com (under "Create Bug Report Case" on the left-hand side). Please submit the *core* file, the EXACT *version* and *build* of MobiLink (10.0.1.3609) and your operating system (uname -a). Please keep the core file, as we may need you to generate the stack traces and memory map on your AIX box. Regards, David Fishburn wrote: |
#26
| ||||||
| ||||||
|
|
Nan, 1. Could we possibly see the output of a 'setenv' before starting the MobiLink server? For now, try starting MobiLink from the shell command-line instead of your shell script. (Once we finish investigating the crash, you can go back to the script). my environment variables: |
|
2. Assume that the initial configuration is correct: - That /disk2/sqlanywhere10/MobiLink/setup/syncora.sql has been run against the Oracle server. - $ORACLE_HOME = /disk2/oracle9i/app/ora9.2 - $ODBCINI = /path/to/.odbc.ini - $ORACLE_HOME/network/admin/tnsnames.ora is a valid configuration |
.
. but I|
i.e. iasora9 = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = xxx.xxx.xxx.xxx)(PORT = xxxx)) ) (CONNECT_DATA = (SERVICE_NAME = orcl) ) ) - That the .odbc.ini reads: [iasora9] Driver=/disk2/sqlanywhere10/lib32/libdboraodbc10_r.so UserID=mobilink Password=mobilink SID=iasora9 ProcResults=No ArraySize=60000 ** NOTE: *** There should be a leading forward-slash to the DRIVER parameter. That wasn't present in the initial post information! ** NOTE: *** ProcResults=No, not "Yes". See: http://www.sybase.com/detail?id=1051962#_Oracle_8_and ================================================== ================ If all of that is correct, we probably need someone to investigate *what* is crashing exactly. Best way (in this scenario specifically, which is a MobiLink crash) is probably to gather a core dump. ================================================== ================ On AIX, to gather a core dump: ------------------------------- * Note: Users can set their ulimit maximums up to the hard limits defined in the /etc/security/limits file that is set by the sysadmin. If you need to exceed those limits, you must have root access to change that file. ------------------------------- 1. Check to see if full core dumps are enabled: % lsattr -El sys0 -a fullcore ** This should return: fullcore true Enable full CORE dump True *** If it doesn't, type (possibly as root): % chdev -l sys0 -a fullcore=true |
|
2. Check the output of "ulimit -a" % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) 2097151 nofiles(descriptors) 2000 *** Check to ensure that "coredump(blocks)" is set to a value big enough to hold the MobiLink memory space. If it isn't, try setting it to "unlimited": % ulimit -c unlimited % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) unlimited nofiles(descriptors) 2000 I changed the configuration of my AIX as above: |
|
3. Have MobiLink produce a core file: % mlsrv10 -c "DSN=iasora9" |
, the information on screen is:|
*** It crashes in the current directory, then should produce a file named "core" or "core.pid" or similar). 4. Submit the core file to Technical Support. If this is a priority issue, please call 1-800-8SYBASE to open a technical support case (using an existing support plan or open a pay-per-incident case). If this isn't a priority issue, you can submit it to http://case-express.sybase.com (under "Create Bug Report Case" on the left-hand side). Please submit the *core* file, the EXACT *version* and *build* of MobiLink (10.0.1.3609) and your operating system (uname -a). Please keep the core file, as we may need you to generate the stack traces and memory map on your AIX box. Regards, David Fishburn wrote: |
#27
| ||||||
| ||||||
|
|
Nan, 1. Could we possibly see the output of a 'setenv' before starting the MobiLink server? For now, try starting MobiLink from the shell command-line instead of your shell script. (Once we finish investigating the crash, you can go back to the script). my environment variables: |
|
2. Assume that the initial configuration is correct: - That /disk2/sqlanywhere10/MobiLink/setup/syncora.sql has been run against the Oracle server. - $ORACLE_HOME = /disk2/oracle9i/app/ora9.2 - $ODBCINI = /path/to/.odbc.ini - $ORACLE_HOME/network/admin/tnsnames.ora is a valid configuration |
.
. but I|
i.e. iasora9 = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = xxx.xxx.xxx.xxx)(PORT = xxxx)) ) (CONNECT_DATA = (SERVICE_NAME = orcl) ) ) - That the .odbc.ini reads: [iasora9] Driver=/disk2/sqlanywhere10/lib32/libdboraodbc10_r.so UserID=mobilink Password=mobilink SID=iasora9 ProcResults=No ArraySize=60000 ** NOTE: *** There should be a leading forward-slash to the DRIVER parameter. That wasn't present in the initial post information! ** NOTE: *** ProcResults=No, not "Yes". See: http://www.sybase.com/detail?id=1051962#_Oracle_8_and ================================================== ================ If all of that is correct, we probably need someone to investigate *what* is crashing exactly. Best way (in this scenario specifically, which is a MobiLink crash) is probably to gather a core dump. ================================================== ================ On AIX, to gather a core dump: ------------------------------- * Note: Users can set their ulimit maximums up to the hard limits defined in the /etc/security/limits file that is set by the sysadmin. If you need to exceed those limits, you must have root access to change that file. ------------------------------- 1. Check to see if full core dumps are enabled: % lsattr -El sys0 -a fullcore ** This should return: fullcore true Enable full CORE dump True *** If it doesn't, type (possibly as root): % chdev -l sys0 -a fullcore=true |
|
2. Check the output of "ulimit -a" % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) 2097151 nofiles(descriptors) 2000 *** Check to ensure that "coredump(blocks)" is set to a value big enough to hold the MobiLink memory space. If it isn't, try setting it to "unlimited": % ulimit -c unlimited % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) unlimited nofiles(descriptors) 2000 I changed the configuration of my AIX as above: |
|
3. Have MobiLink produce a core file: % mlsrv10 -c "DSN=iasora9" |
, the information on screen is:|
*** It crashes in the current directory, then should produce a file named "core" or "core.pid" or similar). 4. Submit the core file to Technical Support. If this is a priority issue, please call 1-800-8SYBASE to open a technical support case (using an existing support plan or open a pay-per-incident case). If this isn't a priority issue, you can submit it to http://case-express.sybase.com (under "Create Bug Report Case" on the left-hand side). Please submit the *core* file, the EXACT *version* and *build* of MobiLink (10.0.1.3609) and your operating system (uname -a). Please keep the core file, as we may need you to generate the stack traces and memory map on your AIX box. Regards, David Fishburn wrote: |
#28
| ||||||
| ||||||
|
|
Nan, 1. Could we possibly see the output of a 'setenv' before starting the MobiLink server? For now, try starting MobiLink from the shell command-line instead of your shell script. (Once we finish investigating the crash, you can go back to the script). my environment variables: |
|
2. Assume that the initial configuration is correct: - That /disk2/sqlanywhere10/MobiLink/setup/syncora.sql has been run against the Oracle server. - $ORACLE_HOME = /disk2/oracle9i/app/ora9.2 - $ODBCINI = /path/to/.odbc.ini - $ORACLE_HOME/network/admin/tnsnames.ora is a valid configuration |
.
. but I|
i.e. iasora9 = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = xxx.xxx.xxx.xxx)(PORT = xxxx)) ) (CONNECT_DATA = (SERVICE_NAME = orcl) ) ) - That the .odbc.ini reads: [iasora9] Driver=/disk2/sqlanywhere10/lib32/libdboraodbc10_r.so UserID=mobilink Password=mobilink SID=iasora9 ProcResults=No ArraySize=60000 ** NOTE: *** There should be a leading forward-slash to the DRIVER parameter. That wasn't present in the initial post information! ** NOTE: *** ProcResults=No, not "Yes". See: http://www.sybase.com/detail?id=1051962#_Oracle_8_and ================================================== ================ If all of that is correct, we probably need someone to investigate *what* is crashing exactly. Best way (in this scenario specifically, which is a MobiLink crash) is probably to gather a core dump. ================================================== ================ On AIX, to gather a core dump: ------------------------------- * Note: Users can set their ulimit maximums up to the hard limits defined in the /etc/security/limits file that is set by the sysadmin. If you need to exceed those limits, you must have root access to change that file. ------------------------------- 1. Check to see if full core dumps are enabled: % lsattr -El sys0 -a fullcore ** This should return: fullcore true Enable full CORE dump True *** If it doesn't, type (possibly as root): % chdev -l sys0 -a fullcore=true |
|
2. Check the output of "ulimit -a" % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) 2097151 nofiles(descriptors) 2000 *** Check to ensure that "coredump(blocks)" is set to a value big enough to hold the MobiLink memory space. If it isn't, try setting it to "unlimited": % ulimit -c unlimited % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) unlimited nofiles(descriptors) 2000 I changed the configuration of my AIX as above: |
|
3. Have MobiLink produce a core file: % mlsrv10 -c "DSN=iasora9" |
, the information on screen is:|
*** It crashes in the current directory, then should produce a file named "core" or "core.pid" or similar). 4. Submit the core file to Technical Support. If this is a priority issue, please call 1-800-8SYBASE to open a technical support case (using an existing support plan or open a pay-per-incident case). If this isn't a priority issue, you can submit it to http://case-express.sybase.com (under "Create Bug Report Case" on the left-hand side). Please submit the *core* file, the EXACT *version* and *build* of MobiLink (10.0.1.3609) and your operating system (uname -a). Please keep the core file, as we may need you to generate the stack traces and memory map on your AIX box. Regards, David Fishburn wrote: |
#29
| ||||||
| ||||||
|
|
Nan, 1. Could we possibly see the output of a 'setenv' before starting the MobiLink server? For now, try starting MobiLink from the shell command-line instead of your shell script. (Once we finish investigating the crash, you can go back to the script). my environment variables: |
|
2. Assume that the initial configuration is correct: - That /disk2/sqlanywhere10/MobiLink/setup/syncora.sql has been run against the Oracle server. - $ORACLE_HOME = /disk2/oracle9i/app/ora9.2 - $ODBCINI = /path/to/.odbc.ini - $ORACLE_HOME/network/admin/tnsnames.ora is a valid configuration |
.
. but I|
i.e. iasora9 = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = xxx.xxx.xxx.xxx)(PORT = xxxx)) ) (CONNECT_DATA = (SERVICE_NAME = orcl) ) ) - That the .odbc.ini reads: [iasora9] Driver=/disk2/sqlanywhere10/lib32/libdboraodbc10_r.so UserID=mobilink Password=mobilink SID=iasora9 ProcResults=No ArraySize=60000 ** NOTE: *** There should be a leading forward-slash to the DRIVER parameter. That wasn't present in the initial post information! ** NOTE: *** ProcResults=No, not "Yes". See: http://www.sybase.com/detail?id=1051962#_Oracle_8_and ================================================== ================ If all of that is correct, we probably need someone to investigate *what* is crashing exactly. Best way (in this scenario specifically, which is a MobiLink crash) is probably to gather a core dump. ================================================== ================ On AIX, to gather a core dump: ------------------------------- * Note: Users can set their ulimit maximums up to the hard limits defined in the /etc/security/limits file that is set by the sysadmin. If you need to exceed those limits, you must have root access to change that file. ------------------------------- 1. Check to see if full core dumps are enabled: % lsattr -El sys0 -a fullcore ** This should return: fullcore true Enable full CORE dump True *** If it doesn't, type (possibly as root): % chdev -l sys0 -a fullcore=true |
|
2. Check the output of "ulimit -a" % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) 2097151 nofiles(descriptors) 2000 *** Check to ensure that "coredump(blocks)" is set to a value big enough to hold the MobiLink memory space. If it isn't, try setting it to "unlimited": % ulimit -c unlimited % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) unlimited nofiles(descriptors) 2000 I changed the configuration of my AIX as above: |
|
3. Have MobiLink produce a core file: % mlsrv10 -c "DSN=iasora9" |
, the information on screen is:|
*** It crashes in the current directory, then should produce a file named "core" or "core.pid" or similar). 4. Submit the core file to Technical Support. If this is a priority issue, please call 1-800-8SYBASE to open a technical support case (using an existing support plan or open a pay-per-incident case). If this isn't a priority issue, you can submit it to http://case-express.sybase.com (under "Create Bug Report Case" on the left-hand side). Please submit the *core* file, the EXACT *version* and *build* of MobiLink (10.0.1.3609) and your operating system (uname -a). Please keep the core file, as we may need you to generate the stack traces and memory map on your AIX box. Regards, David Fishburn wrote: |
#30
| ||||||
| ||||||
|
|
Nan, 1. Could we possibly see the output of a 'setenv' before starting the MobiLink server? For now, try starting MobiLink from the shell command-line instead of your shell script. (Once we finish investigating the crash, you can go back to the script). my environment variables: |
|
2. Assume that the initial configuration is correct: - That /disk2/sqlanywhere10/MobiLink/setup/syncora.sql has been run against the Oracle server. - $ORACLE_HOME = /disk2/oracle9i/app/ora9.2 - $ODBCINI = /path/to/.odbc.ini - $ORACLE_HOME/network/admin/tnsnames.ora is a valid configuration |
.
. but I|
i.e. iasora9 = (DESCRIPTION = (ADDRESS_LIST = (ADDRESS = (PROTOCOL = TCP)(HOST = xxx.xxx.xxx.xxx)(PORT = xxxx)) ) (CONNECT_DATA = (SERVICE_NAME = orcl) ) ) - That the .odbc.ini reads: [iasora9] Driver=/disk2/sqlanywhere10/lib32/libdboraodbc10_r.so UserID=mobilink Password=mobilink SID=iasora9 ProcResults=No ArraySize=60000 ** NOTE: *** There should be a leading forward-slash to the DRIVER parameter. That wasn't present in the initial post information! ** NOTE: *** ProcResults=No, not "Yes". See: http://www.sybase.com/detail?id=1051962#_Oracle_8_and ================================================== ================ If all of that is correct, we probably need someone to investigate *what* is crashing exactly. Best way (in this scenario specifically, which is a MobiLink crash) is probably to gather a core dump. ================================================== ================ On AIX, to gather a core dump: ------------------------------- * Note: Users can set their ulimit maximums up to the hard limits defined in the /etc/security/limits file that is set by the sysadmin. If you need to exceed those limits, you must have root access to change that file. ------------------------------- 1. Check to see if full core dumps are enabled: % lsattr -El sys0 -a fullcore ** This should return: fullcore true Enable full CORE dump True *** If it doesn't, type (possibly as root): % chdev -l sys0 -a fullcore=true |
|
2. Check the output of "ulimit -a" % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) 2097151 nofiles(descriptors) 2000 *** Check to ensure that "coredump(blocks)" is set to a value big enough to hold the MobiLink memory space. If it isn't, try setting it to "unlimited": % ulimit -c unlimited % ulimit -a time(seconds) unlimited file(blocks) unlimited data(kbytes) 131072 stack(kbytes) 32768 memory(kbytes) 32768 coredump(blocks) unlimited nofiles(descriptors) 2000 I changed the configuration of my AIX as above: |
|
3. Have MobiLink produce a core file: % mlsrv10 -c "DSN=iasora9" |
, the information on screen is:|
*** It crashes in the current directory, then should produce a file named "core" or "core.pid" or similar). 4. Submit the core file to Technical Support. If this is a priority issue, please call 1-800-8SYBASE to open a technical support case (using an existing support plan or open a pay-per-incident case). If this isn't a priority issue, you can submit it to http://case-express.sybase.com (under "Create Bug Report Case" on the left-hand side). Please submit the *core* file, the EXACT *version* and *build* of MobiLink (10.0.1.3609) and your operating system (uname -a). Please keep the core file, as we may need you to generate the stack traces and memory map on your AIX box. Regards, David Fishburn wrote: |
![]() |
| Thread Tools | |
| Display Modes | |
| |