<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
	xmlns:georss="http://www.georss.org/georss" xmlns:geo="http://www.w3.org/2003/01/geo/wgs84_pos#" xmlns:media="http://search.yahoo.com/mrss/"
		>
<channel>
	<title>Comments on: How to kill an Oracle Process on Windows</title>
	<atom:link href="http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/feed/" rel="self" type="application/rss+xml" />
	<link>http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/</link>
	<description>What I learned about Oracle</description>
	<lastBuildDate>Tue, 18 Jun 2013 23:14:05 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.com/</generator>
	<item>
		<title>By: coskan</title>
		<link>http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-4316</link>
		<dc:creator><![CDATA[coskan]]></dc:creator>
		<pubDate>Wed, 12 Jan 2011 23:42:57 +0000</pubDate>
		<guid isPermaLink="false">http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-4316</guid>
		<description><![CDATA[SID of the db is basically the output of the query below. 

 select instance_name from v$instance]]></description>
		<content:encoded><![CDATA[<p>SID of the db is basically the output of the query below. </p>
<p> select instance_name from v$instance</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: John</title>
		<link>http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-4313</link>
		<dc:creator><![CDATA[John]]></dc:creator>
		<pubDate>Wed, 12 Jan 2011 04:24:19 +0000</pubDate>
		<guid isPermaLink="false">http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-4313</guid>
		<description><![CDATA[This tutorial is confusing if one doesn&#039;t know what an oracle instance id is. For the newbies out there like me this might be worth explaining.]]></description>
		<content:encoded><![CDATA[<p>This tutorial is confusing if one doesn&#8217;t know what an oracle instance id is. For the newbies out there like me this might be worth explaining.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Shirley S</title>
		<link>http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-3391</link>
		<dc:creator><![CDATA[Shirley S]]></dc:creator>
		<pubDate>Fri, 27 Feb 2009 16:25:54 +0000</pubDate>
		<guid isPermaLink="false">http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-3391</guid>
		<description><![CDATA[I have problem using orakill too. Every time it said successfully signaled. But the session is still there and I ended up reboot the box since this thread generate a huge trace file which I can’t delete. By the way what is “Process Explorer”?

Thanks,]]></description>
		<content:encoded><![CDATA[<p>I have problem using orakill too. Every time it said successfully signaled. But the session is still there and I ended up reboot the box since this thread generate a huge trace file which I can’t delete. By the way what is “Process Explorer”?</p>
<p>Thanks,</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: RichardG</title>
		<link>http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-3319</link>
		<dc:creator><![CDATA[RichardG]]></dc:creator>
		<pubDate>Thu, 03 Jul 2008 06:36:52 +0000</pubDate>
		<guid isPermaLink="false">http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-3319</guid>
		<description><![CDATA[Thanks for that info.

I think that I had the same problem, in that orakill returned success but the thread was still running.  

I think that Oracle has lost track of it&#039;s thread somehow.  I couldn&#039;t see it in v$process.

Anyways... after debating with myself for a while, I killed it using Process Explorer.  This appears to have worked.  At least... there&#039;s been no immediate problem.]]></description>
		<content:encoded><![CDATA[<p>Thanks for that info.</p>
<p>I think that I had the same problem, in that orakill returned success but the thread was still running.  </p>
<p>I think that Oracle has lost track of it&#8217;s thread somehow.  I couldn&#8217;t see it in v$process.</p>
<p>Anyways&#8230; after debating with myself for a while, I killed it using Process Explorer.  This appears to have worked.  At least&#8230; there&#8217;s been no immediate problem.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: cristiancudizio</title>
		<link>http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-2358</link>
		<dc:creator><![CDATA[cristiancudizio]]></dc:creator>
		<pubDate>Wed, 07 Nov 2007 10:59:58 +0000</pubDate>
		<guid isPermaLink="false">http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-2358</guid>
		<description><![CDATA[I would say two things:
1) orakill does not work as kill -9, i&#039;ve a windows 2003 standard edition sp1 machine with Oracle 10.2.0.2 used as development server. when i make orakill SID SPID it gives me the success message, but thread still remains. With kill -9 this never appens
2) Sessions never times out, if client does not cleanly close the connection session persist forever (until the instance is not stopped)
On this server i&#039;ve the problem that developers with debuggers don&#039;t close cleanly oracle sessions so they remain there for days , new sessions are opened and so number of processes grows until maximum processes number is reached. I&#039;ve tried with profiles and idle_time setting but sessions get status &quot;SNIPED&quot; and that does not resolve. With DCD (SQLNET.EXPIRE_TIME) sessions get closed, but shadow processes still remains so my problem remains. Killing threads with Process Explorer works fine, so i&#039;m searching a command line tool to kill thread that works in the same manner to schedule a batch to clean deadh sessions.]]></description>
		<content:encoded><![CDATA[<p>I would say two things:<br />
1) orakill does not work as kill -9, i&#8217;ve a windows 2003 standard edition sp1 machine with Oracle 10.2.0.2 used as development server. when i make orakill SID SPID it gives me the success message, but thread still remains. With kill -9 this never appens<br />
2) Sessions never times out, if client does not cleanly close the connection session persist forever (until the instance is not stopped)<br />
On this server i&#8217;ve the problem that developers with debuggers don&#8217;t close cleanly oracle sessions so they remain there for days , new sessions are opened and so number of processes grows until maximum processes number is reached. I&#8217;ve tried with profiles and idle_time setting but sessions get status &#8220;SNIPED&#8221; and that does not resolve. With DCD (SQLNET.EXPIRE_TIME) sessions get closed, but shadow processes still remains so my problem remains. Killing threads with Process Explorer works fine, so i&#8217;m searching a command line tool to kill thread that works in the same manner to schedule a batch to clean deadh sessions.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: coskan</title>
		<link>http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-1270</link>
		<dc:creator><![CDATA[coskan]]></dc:creator>
		<pubDate>Tue, 28 Aug 2007 14:03:11 +0000</pubDate>
		<guid isPermaLink="false">http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-1270</guid>
		<description><![CDATA[Try What ?

Killing a session running compiled  PL-SQL by orakill ?]]></description>
		<content:encoded><![CDATA[<p>Try What ?</p>
<p>Killing a session running compiled  PL-SQL by orakill ?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: CliffP</title>
		<link>http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-1269</link>
		<dc:creator><![CDATA[CliffP]]></dc:creator>
		<pubDate>Tue, 28 Aug 2007 13:54:54 +0000</pubDate>
		<guid isPermaLink="false">http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-1269</guid>
		<description><![CDATA[Don&#039;t try this with compiled PL/SQL code.  PL/SQL in Oracle 9 and 10 both get upset if you try.

The only way to do this is to call Java - look in asktom for the article covering this, and be very carefull to pay attention to the Java security part.]]></description>
		<content:encoded><![CDATA[<p>Don&#8217;t try this with compiled PL/SQL code.  PL/SQL in Oracle 9 and 10 both get upset if you try.</p>
<p>The only way to do this is to call Java &#8211; look in asktom for the article covering this, and be very carefull to pay attention to the Java security part.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: coskan</title>
		<link>http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-1257</link>
		<dc:creator><![CDATA[coskan]]></dc:creator>
		<pubDate>Tue, 28 Aug 2007 08:23:51 +0000</pubDate>
		<guid isPermaLink="false">http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-1257</guid>
		<description><![CDATA[Thanks for the enlightening comment Yasin 

As I mentioned this were the web site owners thoughts and The web site owner might mean if sometimes,  not every time. 

I think they want to mean, There can be a possibility of a hanged session which stays there after killed. like marked as killed for rolling back. 

Also test below shows us that alter system kill session could not immediately release the locks 

---User Session
11:49:48 USER_SESSION&gt;select sid from v$mystat where rownum=1;

       SID
----------
       157

11:49:51 USER_SESSION&gt;SELECT sid,serial#,username,status from v$session where sid=157;

       SID    SERIAL# USERNAME                       STATUS
---------- ---------- ------------------------------ --------
       157        710 SYS                            ACTIVE

11:54:04 USER_SESSION&gt;insert into t1  select * from dba_objects union all select * from t1;
insert into t1  select * from dba_objects union all select * from t1
*
ERROR at line 1:
ORA-00028: your session has been killed




----Admin session 

11:49:25 ADMIN_SESSION&gt;alter system kill session &#039;157,710&#039;;

System altered.


11:54:27 ADMIN_SESSION&gt;SELECT sid,serial#,username,status from v$session where sid=157;

       SID    SERIAL# USERNAME                       STATUS
---------- ---------- ------------------------------ --------
       157        712                                ACTIVE


11:54:52 ADMIN_SESSION&gt;select sid,type from v$lock where sid=157;

       SID TY
---------- --
       157 PS

11:55:10 ADMIN_SESSION&gt;select sid,type from v$lock where sid=157;

no rows selected




Session is killed at 11:54:04  but lock still held for rolling back purposes at 11:54:52


  



Anyway, this tool is cool for the ones who are in trouble and have nothing to do.]]></description>
		<content:encoded><![CDATA[<p>Thanks for the enlightening comment Yasin </p>
<p>As I mentioned this were the web site owners thoughts and The web site owner might mean if sometimes,  not every time. </p>
<p>I think they want to mean, There can be a possibility of a hanged session which stays there after killed. like marked as killed for rolling back. </p>
<p>Also test below shows us that alter system kill session could not immediately release the locks </p>
<p>&#8212;User Session<br />
11:49:48 USER_SESSION&gt;select sid from v$mystat where rownum=1;</p>
<p>       SID<br />
&#8212;&#8212;&#8212;-<br />
       157</p>
<p>11:49:51 USER_SESSION&gt;SELECT sid,serial#,username,status from v$session where sid=157;</p>
<p>       SID    SERIAL# USERNAME                       STATUS<br />
&#8212;&#8212;&#8212;- &#8212;&#8212;&#8212;- &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212; &#8212;&#8212;&#8211;<br />
       157        710 SYS                            ACTIVE</p>
<p>11:54:04 USER_SESSION&gt;insert into t1  select * from dba_objects union all select * from t1;<br />
insert into t1  select * from dba_objects union all select * from t1<br />
*<br />
ERROR at line 1:<br />
ORA-00028: your session has been killed</p>
<p>&#8212;-Admin session </p>
<p>11:49:25 ADMIN_SESSION&gt;alter system kill session &#8217;157,710&#8242;;</p>
<p>System altered.</p>
<p>11:54:27 ADMIN_SESSION&gt;SELECT sid,serial#,username,status from v$session where sid=157;</p>
<p>       SID    SERIAL# USERNAME                       STATUS<br />
&#8212;&#8212;&#8212;- &#8212;&#8212;&#8212;- &#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212; &#8212;&#8212;&#8211;<br />
       157        712                                ACTIVE</p>
<p>11:54:52 ADMIN_SESSION&gt;select sid,type from v$lock where sid=157;</p>
<p>       SID TY<br />
&#8212;&#8212;&#8212;- &#8211;<br />
       157 PS</p>
<p>11:55:10 ADMIN_SESSION&gt;select sid,type from v$lock where sid=157;</p>
<p>no rows selected</p>
<p>Session is killed at 11:54:04  but lock still held for rolling back purposes at 11:54:52</p>
<p>Anyway, this tool is cool for the ones who are in trouble and have nothing to do.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Yas</title>
		<link>http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-1256</link>
		<dc:creator><![CDATA[Yas]]></dc:creator>
		<pubDate>Tue, 28 Aug 2007 08:05:27 +0000</pubDate>
		<guid isPermaLink="false">http://coskan.wordpress.com/2007/08/27/how-to-kill-oracle-process-from-windows/#comment-1256</guid>
		<description><![CDATA[&lt;i&gt;1. The alter system statement will not clear any locks that exist. Instead, the session will remain connected until it times out, then the session is killed and the locks are released.&lt;/i&gt;

That is not true. &quot;alter system kill session&quot; releases the locks held by the killed session.


SQL&gt; create table test as select * from all_objects;

Table created.

SQL&gt; select sid from v$mystat where rownum=1;

       SID
----------
       341

SQL&gt; select sid,serial# from v$session where sid=341;

       SID    SERIAL#
---------- ----------
       341      21620

SQL&gt; update test set object_id=1;

61657 rows updated.



From another session:


SQL&gt; select sid,type from v$lock where sid=341;

       SID TY
---------- --
       341 TX
       341 TM

SQL&gt; alter system kill session &#039;341,21620&#039;;

System altered.

SQL&gt; select sid,type from v$lock where sid=341;

no rows selected


The related documentation also confirms that : http://download.oracle.com/docs/cd/B19306_01/server.102/b14200/statements_2013.htm#i2065117]]></description>
		<content:encoded><![CDATA[<p><i>1. The alter system statement will not clear any locks that exist. Instead, the session will remain connected until it times out, then the session is killed and the locks are released.</i></p>
<p>That is not true. &#8220;alter system kill session&#8221; releases the locks held by the killed session.</p>
<p>SQL&gt; create table test as select * from all_objects;</p>
<p>Table created.</p>
<p>SQL&gt; select sid from v$mystat where rownum=1;</p>
<p>       SID<br />
&#8212;&#8212;&#8212;-<br />
       341</p>
<p>SQL&gt; select sid,serial# from v$session where sid=341;</p>
<p>       SID    SERIAL#<br />
&#8212;&#8212;&#8212;- &#8212;&#8212;&#8212;-<br />
       341      21620</p>
<p>SQL&gt; update test set object_id=1;</p>
<p>61657 rows updated.</p>
<p>From another session:</p>
<p>SQL&gt; select sid,type from v$lock where sid=341;</p>
<p>       SID TY<br />
&#8212;&#8212;&#8212;- &#8211;<br />
       341 TX<br />
       341 TM</p>
<p>SQL&gt; alter system kill session &#8217;341,21620&#8242;;</p>
<p>System altered.</p>
<p>SQL&gt; select sid,type from v$lock where sid=341;</p>
<p>no rows selected</p>
<p>The related documentation also confirms that : <a href="http://download.oracle.com/docs/cd/B19306_01/server.102/b14200/statements_2013.htm#i2065117" rel="nofollow">http://download.oracle.com/docs/cd/B19306_01/server.102/b14200/statements_2013.htm#i2065117</a></p>
]]></content:encoded>
	</item>
</channel>
</rss>
