RE: Subscribe_for_node_down

From: William Wagman <wjwagman_at_ucdavis.edu>
Date: Fri, 18 May 2007 16:47:40 -0700
Message-ID: <FE043305B38A0F448F3924429D650C2A02A79CF6@VEXBE2.ex.ad3.ucdavis.edu>

Fairlie,

Thanks for the response. Here is another piece to the puzzle which to me just adds to the confusion. On node 1...

$ srvctl status nodeapps -n brownshoe

hangs

$ onsctl ping

Number of onsconfiguration retrieved, numcfg = 2 onscfg[0]

{node = brownshoe.ucdavis.edu, port = 6200} Adding remote host brownshoe.ucdavis.edu:6200 onscfg[1]

{node = tenspeed.ucdavis.edu, port = 6200} Adding remote host tenspeed.ucdavis.edu:6200 ons is not running ...
$

on node 2 I see...

$ srvctl status nodeapps -n tenspeed

VIP is running on node: tenspeed
GSD is running on node: tenspeed
Listener is not running on node: tenspeed ONS daemon is running on node: tenspeed
$ onsctl ping

Number of onsconfiguration retrieved, numcfg = 2 onscfg[0]

{node = brownshoe.ucdavis.edu, port = 6200} Adding remote host brownshoe.ucdavis.edu:6200 onscfg[1]

{node = tenspeed.ucdavis.edu, port = 6200} Adding remote host tenspeed.ucdavis.edu:6200 ons is not running ...
$

Why does node 1 hang and why does srvctl show ons running but onsctl ping says no. I don't understand the discrepancy.

Thanks.

Bill Wagman
Univ. of California at Davis
IET Campus Data Center
wjwagman_at_ucdavis.edu
(530) 754-6208

From: fairlie rego [mailto:fairlie_r_at_yahoo.com] Sent: Friday, May 18, 2007 4:14 PM
To: William Wagman
Cc: oracle-l
Subject: Re: Subscribe_for_node_down_event

William,

In brief this is all a part of the FAN framework where a listener registers with ONS.
You shouldn't have this problem if you set SUBSCRIBE_FOR_NODE_DOWN_EVENT_<listener> to OFF which should disable ONS subscription
If you still have an issue can you paste/send the output of onsctl debug

Thanks
Fairlie

William Wagman <wjwagman_at_ucdavis.edu> wrote:

Greetings,

On a two node RAC cluster, Oracle 10.2.0.3.0 Enterprise Edition on RHEL4

64-bit using CRS.

	In my listener.log I see numerous occurrences of the message...
	WARNING: Subscription for node down event still pending
	In searching for the web I find it commonly suggested that the
	listener.ora file by updated to include the parameter
	SUBSCRIBE_FOR_NODE_DOWN_EVENT_LSNR=
	
	This is where I get lost. I am unable to find any documentation
as to
	what constitutes a node down event and what the ramifications
are. What
	is a node down event?
	
	I have searched the Oracle docs for this parameter, nothing is
found.
	Searching for listener.ora refers me to the net services guide
but I am
	unable to find specific syntax for this parameter, it doesn't
even
	appear in the index. Searching also refers me to the configuring
the
	listener.ora in the various clusterware guides but again, no
syntax for
	the parameter is given nor is it listed in the index. I'm

stumped.

What we are seeing is that our application is using web logic and it

appears that after a while connections and querying start to slow down

and restarting the listener temporarily appears to resolve the issue.

Some of the notes about the warning message indicates that the listener

encountering some memory issues but I really can't find any explanations

of this anywhere.

If anyone can shed any light on this or point me to some documentation

on what this means or how to troubleshoot this problem I would be very

grateful.

Thanks.

	Bill Wagman
	Univ. of California at Davis
	IET Campus Data Center
	wjwagman_at_ucdavis.edu
	(530) 754-6208
	--
	http://www.freelists.org/webpage/oracle-l

Fairlie Rego
Senior Oracle Consultant
http://el-caro.blogspot.com/
M: +61 402 792 405

Yahoo! oneSearch: Finally, mobile search that gives answers <http://us.rd.yahoo.com/evt=48252/*http://mobile.yahoo.com/mobileweb/one search?refer=1ONXIC> , not web links.

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

Received on Fri May 18 2007 - 18:47:40 CDT

RE: Subscribe_for_node_down_event