Dtrace writes in error log

Question

Hi all,
Here is the problem. There is a Soalris host with ZFS. There is some proprietary application which writes its error log. It has been noted that the error messages are not always written immediately, but only after processing some erroneous records. Plus in case of module terminating with SIGTERM, those errors which are still not in the log - they just would be lost, won't be written.
The experiment has been repeated on the different Solaris host with error log placed on local UFS and, second try, on NFS. In both cases error log writes immediately and correctly. So we are inclining to blame ZFS configuration.
I have an access to the "problematic" host and root privileges there. I would not like to restart the process which means I can't use truss. Hence I have a dtrace (in which I'm a nub yet).
Could somebody please give me a hint - which syscalls I should try to monitor to understand what is going on between the application and file system?
Any clue would be deeply appreciated.

Answer 1

Application release: 12.1.2
DB version: 11.2.0.2
OS: Linux 5.6 x86-64
Users can't launch login screen or get hang after input ID and password intermittently after migrate Oracle E-Business R12.1.2 from IBM AIX to Oracle Linux 5.6.
DB logs has errors:
Fatal NI connect error 12170.
VERSION INFORMATION:
TNS for Linux: Version 11.2.0.2.0 - Production
Oracle Bequeath NT Protocol Adapter for Linux: Version 11.2.0.2.0 - Production
TCP/IP NT Protocol Adapter for Linux: Version 11.2.0.2.0 - Production
Time: 05-SEP-2011 21:29:25
Tracing not turned on.
Tns error struct:
ns main err code: 12535
TNS-12535: TNS:operation timed out
ns secondary err code: 12560
nt main err code: 505
TNS-00505: Operation timed out
nt secondary err code: 110
nt OS err code: 0
Client address: (ADDRESS=(PROTOCOL=tcp)(HOST=10.36.152.208)(PORT=3763))
But these DB errors doesn't occurred in issue time.
Autoconfig can complete successfully.
Thanks,
Jackie