RE: Defunct processes in 10g(mmon and qmnc)

From: Jesse, Rich <Rich.Jesse_at_qg.com>
Date: Thu, 27 Apr 2006 11:44:03 -0500
Message-ID: <FB5D3CCFCECC2948B5DCF4CABDBE669754614B@QTEX1.qg.com>

Yup. 10.1.0.4 on Solaris 9. The defuncts death spiraled until we ran out of OS process slots and I had to bounce the instance. All 10000+ processes that I saw were from MMON. And this was on a production box. From the SR: "Development indicates that the MMON messages occur when manageability features like automatic snapshots stop if one executes 'alter system enable restricted session' and then 'alter system disable restricted session'."  

I had not touched any of the ADDM stuff in this database, so I'm not exactly buying the answer, but YMMV.  

GL!  

Rich

	-----Original Message-----
	From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Sriram Kumar
	Sent: Thursday, April 27, 2006 9:25 AM
	To: ORACLE-L
	Subject: Defunct processes in 10g(mmon and qmnc)
	
	
	Hello Gurus,
	 
	Environment: 10.1.0.4
	OS : Solaris 9 (Sunfire V440)
	 
	Was casually checking process listing of oracle id and found a

lot (about 70-80 processes) of <defunct> processes and most of these processes have thier PPID as qmnc' s (queue monitor) or mmon's (mem mon) process id. being a test box I had killed qmnc and mmon and these processes disappeared.          

        I am closely monitoring the systems for these symptoms again and plan to raise a SR with metablink. Have any one of you seen this before?          

        Best Regards          

        Sriram Kumar                              

--
http://www.freelists.org/webpage/oracle-l

Received on Thu Apr 27 2006 - 11:44:03 CDT