X-MimeOLE: Produced By Microsoft Exchange V6.5
Received: by onstor-exch02.onstor.net 
	id <01C8B7C9.DE93A498@onstor-exch02.onstor.net>; Fri, 16 May 2008 19:58:23 -0700
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="----_=_NextPart_001_01C8B7C9.DE93A498"
Content-class: urn:content-classes:message
Subject: RE: Defect  SW-Linux Opened TED00023817
Date: Fri, 16 May 2008 19:58:22 -0700
Message-ID: <BB375AF679D4A34E9CA8DFA650E2B04E09FCF877@onstor-exch02.onstor.net>
In-Reply-To: <ONSTOR-EXCH01WILhFs00000b7b@onstor-exch01.onstor.net>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Defect  SW-Linux Opened TED00023817
Thread-Index: Aci24jnA+xRoreQjTnWUc5X1mmYw/gA5xHig
References: <ONSTOR-EXCH01WILhFs00000b7b@onstor-exch01.onstor.net>
From: "Ron Bhanukitsiri" <ronb@onstor.com>
To: "Jonathan Goldick" <jonathan.goldick@onstor.com>
Cc: "Andy Sharp" <andy.sharp@onstor.com>,
	"Ron Bhanukitsiri" <ronb@onstor.com>

This is a multi-part message in MIME format.

------_=_NextPart_001_01C8B7C9.DE93A498
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

Good news on the cifsd massive memory leak Andy and I saw yesterday.
As I mentioned earlier to Jonathan, my source tree was about a week
old.  So today after I checked in the MMC fix, I rebuilt everything
in my cg_beta workspace from scratch and added the ldap open
trace back in.  No massive memory leak of any kind when I run my
test on my Cougar.  My test shows constant memory usage of 4702
bytes (similar to constant memory usage with my BC and thus consistent)
with MMC until eventually in ran into the connect() error.  Furthermore,
it makes no sense since there were no code changes in cifsd at all other
than the trace.

So I think we can now close the book on the massive memory leak
in cifsd.

Ron B[ee]
PS - Thanks Andy for your help yesterday.

-----Original Message-----
From: ron.bhanukitsiri@onstor.com [mailto:ron.bhanukitsiri@onstor.com]=20
Sent: Thursday, May 15, 2008 4:20 PM
To: Andy Sharp
Cc: Ron Bhanukitsiri
Subject: Defect SW-Linux Opened TED00023817

id: TED00023817
Headline: outgoing socket connect() returns error
Severity: 2-Major
Build: Submittal 21
Description: After some tireless;-) debugging, I chased the problem down
to openldap library os-ip.c module calling socket connect() call and got
EINTR error try to connect to the domain controller.

You may also see the following messages on the SSC console.  When this
happens, SSC is totally unresponsive to ssh session.
SiByte Watchdog in danger of initiating system reset in 8.3 seconds
SiByte Watchdog in danger of initiating system reset in 8.3 seconds
SiByte Watchdog in danger of initiating system reset in 8.3 seconds
SiByte Watchdog in danger of initiating system reset in 8.3 seconds
SiByte Watchdog in danger of initiating system reset in 8.3 seconds

I hasten to add that thousands of MSRPCs were processed.  Sometimes, the
problem happened quickly, sometimes it takes longer.  Some test runs
don't fail at all.

Just curious, I add some experimental code to retry "ldap connect" in
cifsd and it always seem to work after one retry.  However, if I get the
watchdog messages, MMC will eventually die.

I've ruled out any possible h/w issue.

Release_Project: Cougar


------_=_NextPart_001_01C8B7C9.DE93A498
Content-Type: text/html;
	charset="us-ascii"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3Dus-ascii">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
6.5.7653.38">
<TITLE>RE: Defect  SW-Linux Opened TED00023817</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/rtf format -->

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">Good =
news on the</FONT></SPAN><SPAN LANG=3D"en-us"><FONT SIZE=3D2 =
FACE=3D"Tahoma"> cifsd</FONT></SPAN><SPAN LANG=3D"en-us"><FONT SIZE=3D2 =
FACE=3D"Tahoma"> massive memory leak Andy and I saw =
yesterday.</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">As I =
mentioned earlier to Jonathan, my source tree was about =
a</FONT></SPAN><SPAN LANG=3D"en-us"> <FONT SIZE=3D2 =
FACE=3D"Tahoma">week</FONT></SPAN><SPAN LANG=3D"en-us"></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 =
FACE=3D"Tahoma">old.&nbsp; So today after I checked in the MMC fix, I =
rebuilt everything</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">in my =
cg_beta workspace from scratch and added the ldap open</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">trace =
back in.&nbsp; No massive memory leak of any kind</FONT></SPAN><SPAN =
LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma"> when I run =
my</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">test on =
my Cougar</FONT></SPAN><SPAN LANG=3D"en-us"><FONT SIZE=3D2 =
FACE=3D"Tahoma">.&nbsp; My test shows</FONT></SPAN><SPAN =
LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma"></FONT></SPAN><SPAN =
LANG=3D"en-us"> <FONT SIZE=3D2 FACE=3D"Tahoma">constant memory usage of =
4</FONT></SPAN><SPAN LANG=3D"en-us"><FONT SIZE=3D2 =
FACE=3D"Tahoma">702</FONT></SPAN><SPAN LANG=3D"en-us"></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">bytes =
(similar to constant memory</FONT></SPAN><SPAN LANG=3D"en-us"><FONT =
SIZE=3D2 FACE=3D"Tahoma"></FONT></SPAN><SPAN LANG=3D"en-us"> <FONT =
SIZE=3D2 FACE=3D"Tahoma">usage with my BC and thus =
consistent)</FONT></SPAN><SPAN LANG=3D"en-us"></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">with =
MMC until eventually</FONT></SPAN><SPAN LANG=3D"en-us"><FONT SIZE=3D2 =
FACE=3D"Tahoma"></FONT></SPAN><SPAN LANG=3D"en-us"> <FONT SIZE=3D2 =
FACE=3D"Tahoma">in ran into the connect() error.</FONT></SPAN><SPAN =
LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">&nbsp; =
Furthermore,</FONT></SPAN><SPAN LANG=3D"en-us"></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">it =
makes no sense</FONT></SPAN><SPAN LANG=3D"en-us"><FONT SIZE=3D2 =
FACE=3D"Tahoma"></FONT></SPAN><SPAN LANG=3D"en-us"> <FONT SIZE=3D2 =
FACE=3D"Tahoma">since there were no code changes in cifsd at all =
other</FONT></SPAN><SPAN LANG=3D"en-us"></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">than =
the</FONT></SPAN><SPAN LANG=3D"en-us"><FONT SIZE=3D2 =
FACE=3D"Tahoma"></FONT></SPAN><SPAN LANG=3D"en-us"> <FONT SIZE=3D2 =
FACE=3D"Tahoma">trace.</FONT></SPAN><SPAN LANG=3D"en-us"></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">So I =
think we can now close the book on the massive memory =
leak</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">in =
cifsd.</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">Ron =
B[ee]</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">PS - =
Thanks Andy for your help yesterday.</FONT></SPAN><SPAN =
LANG=3D"en-us"></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 =
FACE=3D"Tahoma">-----Original Message-----<BR>
From: ron.bhanukitsiri@onstor.com [<A =
HREF=3D"mailto:ron.bhanukitsiri@onstor.com">mailto:ron.bhanukitsiri@onsto=
r.com</A>]<BR>
Sent: Thursday, May 15, 2008 4:20 PM<BR>
To: Andy Sharp<BR>
Cc: Ron Bhanukitsiri<BR>
Subject: Defect SW-Linux Opened TED00023817</FONT></SPAN><SPAN =
LANG=3D"en-us"></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">id: =
TED00023817</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 =
FACE=3D"Tahoma">Headline: outgoing socket connect() returns =
error</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 =
FACE=3D"Tahoma">Severity: 2-Major</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">Build: =
Submittal 21</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 =
FACE=3D"Tahoma">Description: After some tireless;-) debugging, I chased =
the problem down to openldap library os-ip.c module calling socket =
connect() call and got EINTR error try to connect to the domain =
controller.</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">You may =
also see the following messages on the SSC console.&nbsp; When this =
happens, SSC is totally unresponsive to ssh session.</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">SiByte =
Watchdog in danger of initiating system reset in 8.3 =
seconds</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">SiByte =
Watchdog in danger of initiating system reset in 8.3 =
seconds</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">SiByte =
Watchdog in danger of initiating system reset in 8.3 =
seconds</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">SiByte =
Watchdog in danger of initiating system reset in 8.3 =
seconds</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">SiByte =
Watchdog in danger of initiating system reset in 8.3 =
seconds</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">I =
hasten to add that thousands of MSRPCs were processed.&nbsp; Sometimes, =
the problem happened quickly, sometimes it takes longer.&nbsp; Some test =
runs don't fail at all.</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">Just =
curious, I add some experimental code to retry &quot;ldap connect&quot; =
in cifsd and it always seem to work after one retry.&nbsp; However, if I =
get the watchdog messages, MMC will eventually die.</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 FACE=3D"Tahoma">I've =
ruled out any possible h/w issue.</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"><FONT SIZE=3D2 =
FACE=3D"Tahoma">Release_Project: Cougar</FONT></SPAN></P>

<P DIR=3DLTR><SPAN LANG=3D"en-us"></SPAN></P>

</BODY>
</HTML>
------_=_NextPart_001_01C8B7C9.DE93A498--
