Changing Solaris System Clock results in Unkillable Process'
edited Aug 20, 2013 8:30PM in Oracle Solaris Performance, Panics, Hangs, and Dtrace (MOSC) 2 commentsAnswered
Hi,
I have a Solaris 10 server ( SunOS myhost 5.10 Generic_147440-09 ) setup for the purpose of testing an application.
The requirement of the application testing requires that the system date needs to be regularly changed. ( into the future, mostly )
The application is a Weblogic appllication. When I change the system date I have noticed that sometimes the Weblogic process will become unkillable.
i.e. kill -9 does nothing, truss -p PID results in "unanticapited system error", requiring me to reboot the server in order to restart the application.
The server is running ~30zones, and rebooting because of this is not ideal. As a workaround, I have been shutting down all applications / zones, then resetting the clock and then starting everything back up which is also time consuming.
I have a Solaris 10 server ( SunOS myhost 5.10 Generic_147440-09 ) setup for the purpose of testing an application.
The requirement of the application testing requires that the system date needs to be regularly changed. ( into the future, mostly )
The application is a Weblogic appllication. When I change the system date I have noticed that sometimes the Weblogic process will become unkillable.
i.e. kill -9 does nothing, truss -p PID results in "unanticapited system error", requiring me to reboot the server in order to restart the application.
The server is running ~30zones, and rebooting because of this is not ideal. As a workaround, I have been shutting down all applications / zones, then resetting the clock and then starting everything back up which is also time consuming.
0