Hanging Report Requests In UNIX
Often when duplicate processes run simultaneously in a UNIX system report requests from report servers hang up and can lead to serious performance issues. What can be a workaround this process?
Often when duplicate processes run simultaneously in a UNIX system report requests from report servers hang up and can lead to serious performance issues. What can be a workaround this process?
Report request crashes are based majorly on two issues. Either the master job hangs up or freezes before finishing or report server stops responding when running report requests. We may approach the problems as follows:
a) Master Job hangs: To start with we must check the engineResponseTimeOut in the engine element of rwserver.conf file, from here we can set all attributes to prevent server instability. We can also try showjobs command on the console to know which jobs are hanging up, by ending the hanging processes we can enable other processes in the priority queue to run.
b) Report server stops responding: We can try restarting the report server and if that does not work or is not possible we can get the dump for debugging by killing the current process and getting the dump to a separate file. On Solaris the command “kill -3 server_pid” provides the required diagnostic information on the console. To redirect the thread information we can modify the rwserver.sh file in $ORACLE_INSTANCE/config/reports/bin directory.