Shared Servers stay in Status Quit

Shared Servers stay in Status Quit (Doc ID 1072534.1)

Last updated on MARCH 22, 2010

Applies to:

Oracle Server - Enterprise Edition - Version: 10.2.0.1 - Release: 10.2
Information in this document applies to any platform.
Reported for 10.2.0.4

Symptoms

Shared servers stay in QUIT status and never goto status TERMINATED.

SQL> select name, status from shared_server;

NAME STATUS
---- ----------------
S000 EXEC
S001 WAIT(COMMON)
S002 WAIT(COMMON)
S003 WAIT(COMMON)
S004 WAIT(COMMON)
S005 QUIT
S006 QUIT
S007 QUIT
S008 QUIT
S009 QUIT
S010 QUIT

NAME STATUS
---- ----------------
S011 QUIT
S012 QUIT
S013 QUIT
S014 QUIT
S015 QUIT
S019 QUIT
S022 QUIT

18 rows selected.

This means the resources allocated to the shared server (memory) memory is never released.Shared server process are supposed to terminate if the following manner:

Shared server is finished with and is marked EXEC
It is now excess to requirements so it is marked QUIT
If it receives a new request within 30 seconds or so, then it will not terminate and will process that request (back to EXEC)
If it receives no new request within about 30 seconds then it should change from QUIT to TERMINATED and the OS process exits.

Oracle Net server trace of shared server is status QUIT shows the following is repeated:

[000001 04-DEC-2009 10:30:12:603] nsevwait: 0 registered connection(s)
[000001 04-DEC-2009 10:30:12:603] nsevwait: 0 pre-posted event(s)
[000001 04-DEC-2009 10:30:12:603] nsevwait: waiting for transport event (0 thru 0)...
[000001 04-DEC-2009 10:30:42:604] nsevwait: 0 newly-posted event(s)
[000001 04-DEC-2009 10:30:42:604] nsevwait: 0 posted event(s)
[000001 04-DEC-2009 10:30:42:604] nsevwait: exit (0)

Tusc of the shared server process in QUIT status shows:

sigprocmask(SIG_BLOCK, 0x9ffffffffffedbc0, NULL) ......... = 0
sigprocmask(SIG_UNBLOCK, 0x9ffffffffffedbc0, NULL) ....... = 0
times(0x60000000001c19cc) ................................ = 2331651118
times(0x60000000001c19bc) ................................ = 2331651118
Received signal 23, SIGWINCH, in poll(), [caught]
Siginfo: sent by pid 13183 (uid 1001), si_errno: 0
poll(0x9fffffffffff2e00, 2, 1000) ........................ ERR#4 EINTR
sigprocmask(SIG_BLOCK, 0x9ffffffffffedbc0, NULL) ......... = 0
sigprocmask(SIG_UNBLOCK, 0x9ffffffffffedbc0, NULL) ....... = 0
times(0x60000000001c19cc) ................................ = 2331651123
times(0x60000000001c19bc) ................................ = 2331651123
poll(0x9fffffffffff2160, 2, 0) ........................... = 0
times(0x60000000001c19cc) ................................ = 2331651123
times(0x60000000001c19bc) ................................ = 2331651123
poll(0x9fffffffffff2160, 2, 0) ........................... = 0
times(0x60000000001c19cc) ................................ = 2331651123
times(0x60000000001c19bc) ................................ = 2331651123
poll(0x9fffffffffff2160, 2, 2500) ........................ = 1
recvmsg(400, "9fffffffffff% p \0\0\010\0\0\0\0".., O_RDONLY) = 2152

Shutting down shared servers with the command

SQL>alter system set shared_severs='0';

Does not always release the operating system process for the shared server. Kill -9 on the PID sometimes has to be run.

Cause

	Sign In with your My Oracle Support account
	Don't have a My Oracle Support account? Click to get started

My Oracle Support provides customers with access to over a
Million Knowledge Articles and hundreds of Community platforms

Applies to:

Symptoms

Cause

Sign In with your My Oracle Support account

Don't have a My Oracle Support account? Click to get started