<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=us-ascii">
<META content="MSHTML 6.00.2900.2802" name=GENERATOR></HEAD>
<BODY>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN class=931510918-06042006>Yes it
should.</SPAN></FONT></DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN
class=931510918-06042006></SPAN></FONT> </DIV>
<DIV><FONT face=Arial color=#0000ff size=2><SPAN
class=931510918-06042006>Ishai</SPAN></FONT></DIV><BR>
<DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left>
<HR tabIndex=-1>
<FONT face=Tahoma size=2><B>From:</B> Woodruff, Robert J
[mailto:robert.j.woodruff@intel.com] <BR><B>Sent:</B> Thursday, April 06, 2006
1:31 AM<BR><B>To:</B> Ishai Rabinovitz;
openfabrics-ewg@openib.org<BR><B>Subject:</B> RE: [openfabrics-ewg] Changes in
SM for the new SRP daemon<BR></FONT><BR></DIV>
<DIV></DIV>
<DIV dir=ltr align=left><SPAN class=988263022-05042006><FONT face=Arial
color=#0000ff size=2>Shouldn't this discussion be on openib-general
?</FONT></SPAN></DIV><BR>
<DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left>
<HR tabIndex=-1>
<FONT face=Tahoma size=2><B>From:</B> openfabrics-ewg-bounces@openib.org
[mailto:openfabrics-ewg-bounces@openib.org] <B>On Behalf Of </B>Ishai
Rabinovitz<BR><B>Sent:</B> Wednesday, April 05, 2006 2:14 PM<BR><B>To:</B>
openfabrics-ewg@openib.org<BR><B>Subject:</B> [openfabrics-ewg] Changes in SM
for the new SRP daemon<BR></FONT><BR></DIV>
<DIV></DIV>
<DIV><FONT face=Arial color=#0000ff size=2></FONT> </DIV>
<DIV class=OutlookMessageHeader lang=en-us dir=ltr align=left><FONT face=Arial
size=2><SPAN class=396313908-04042006>Summary:</SPAN></DIV>
<DIV>
<BLOCKQUOTE dir=ltr style="MARGIN-RIGHT: 0px">
<P><SPAN class=396313908-04042006>In the next release of SRP we want to add a
daemon that is executed in each initiator and finds out which targets exist in
the fabric.</SPAN></P>
<P><SPAN class=396313908-04042006>The new daemon can use SA capabilities
from IBTA 1.2 errata (details at the bottom) to improve performance. If
you are using SM other then openIB's openSM please support this feature in
your SM.</SPAN></P></BLOCKQUOTE>
<P><SPAN class=396313908-04042006>Details:</SPAN></P>
<BLOCKQUOTE dir=ltr style="MARGIN-RIGHT: 0px">
<P><SPAN class=396313908-04042006>When one wants to find all SRP targets and
their information in the fabric, he/she currently run "ibsrpdm"<BR>"ibsrpdm"
uses the following procedure to discover all SRP targets available in the
fabric</SPAN></P>
<OL>
<LI><SPAN class=396313908-04042006>"ibsrpdm" sends a query <SPAN
class=396313908-04042006>get<SPAN class=006224120-05042006>_</SPAN>table
node info with node type is CA</SPAN>. </SPAN>
<LI><SPAN class=396313908-04042006></SPAN><SPAN class=396313908-04042006>It
gets quite a big table and then for <SPAN
class=006224120-05042006>every </SPAN>node it sends a <SPAN
class=006224120-05042006>port_info </SPAN>query. </SPAN>
<LI><SPAN class=396313908-04042006>From the response to this query the
initiator can check if this port is an SRP target. (dm <SPAN
class=006224120-05042006>bit </SPAN>capability is set)</SPAN></LI></OL>
<P>The problem with <SPAN class=396313908-04042006>this</SPAN> <SPAN
class=396313908-04042006>procedure</SPAN> is that it <SPAN
class=396313908-04042006>may </SPAN>create too much traffic on the fabric.</P>
<P>Let's assume there is a cluster of 4096 nodes<SPAN
class=396313908-04042006> booting together</SPAN>. Each of this 4096 nodes is
sending the first query and gets a list of 4096 nodes. This list is divided
into a long number of UD messages and may cause retransmit. After getting this
list each node sends 4096 queries for the port of each node.</P>
<P>This traffic is <SPAN class=396313908-04042006>quite </SPAN>huge.</P>
<P><SPAN class=396313908-04042006>T</SPAN>he SA <SPAN
class=396313908-04042006>has a new </SPAN>capability to answer the query:
"please return a list of the ports that has the dm bit set" (meaning return
ports of SRP targets). (This capability of the SA is part of Errata Release
Version: 1.2 1/26/2006 Chapter/Subsection: 15.2.5.3 <SPAN
class=396313908-04042006> - quoted at the bottom</SPAN>). </P>
<P><SPAN class=396313908-04042006>Using this capability we can use the
following procedure:</SPAN></P>
<OL>
<LI><SPAN class=396313908-04042006>The daemon will send "get table port_info
of ports that has dm bit set" query and gets a table of small number of
port_info<SPAN class=006224120-05042006>.</SPAN></SPAN>
<LI><SPAN class=396313908-04042006>For each port</SPAN> It queries for
the guid of this ports.</LI></OL>
<P>This will significantly reduce the traffic on the fabric. </P>
<P><FONT size=+0><FONT size=2>Actually, in this solution there is so little
traffic that <SPAN class=396313908-04042006>the new daemon will run it
periodically (every minute) to look for changes (There will be
less traffic than </SPAN><SPAN class=396313908-04042006>registering
</SPAN>to Trap 64 and Trap 65.<SPAN
class=396313908-04042006>)</SPAN></FONT></FONT></P>
<P><FONT size=+0><FONT size=2><SPAN
class=396313908-04042006></SPAN></FONT></FONT> </P><FONT size=+0><FONT
face=Arial><SPAN class=396313908-04042006></BLOCKQUOTE>
<P dir=ltr><SPAN class=396313908-04042006><FONT size=2>Quoting the
errata:</FONT></SPAN></P>
<BLOCKQUOTE dir=ltr style="MARGIN-RIGHT: 0px">
<P><FONT
size=2>----------------------------------------------------------------------------</FONT></P>
<P><FONT size=2>Errata Tracking Number: MGTWG8372 </FONT></P>
<P><FONT size=2>Sub-Case Number: 0 </FONT></P>
<P><FONT size=2>Reference ID: 4291 </FONT></P>
<P><FONT size=2>Title: Enhanced SA PortInfoRecord searches </FONT></P>
<P><FONT size=2>Submitter: Livingston, James (James.Livingston@necsam.com)
</FONT></P>
<P><FONT size=2>Volume: 1 </FONT></P>
<P><FONT size=2>Revision: 1.2 </FONT></P>
<P><FONT size=2>Errata Release Version: 1.2 1/26/2006 </FONT></P>
<P><FONT size=2>Chapter/Subsection: 15 </FONT></P>
<P><FONT size=2>Page: 885 </FONT></P>
<P><FONT size=2>Line: 20 </FONT></P>
<P><FONT size=2>AssignedIntensity: </FONT></P>
<P><FONT size=2>Status/Disposition: WG_Approved </FONT></P>
<P><FONT size=2>Problem Description: Add a new row to Table 186 SA-Specific
</FONT></P>
<P><FONT size=2>ClassPortInfo:CapabilityMask Bits: </FONT></P>
<P><FONT size=2>Problem Resolution: Add a new row to Table 186 SA-Specific
</FONT></P>
<P><FONT size=2>ClassPortInfo:CapabilityMask Bits </FONT></P>
<P><FONT size=2>Original Text: <none> </FONT></P>
<P><FONT size=2>Corrected Text: <Name> </FONT></P>
<P><FONT size=2>IsPortInfoCapMaskMatchSupported </FONT></P>
<P><FONT size=2><Bit> </FONT></P>
<P><FONT size=2>13 </FONT></P>
<P><FONT size=2><Description> </FONT></P>
<P><FONT size=2>If this value is 1, SA shall support matching the </FONT></P>
<P><FONT size=2>PortInfo:CapabilityMask component as described in <ref to
section </FONT></P>
<P><FONT size=2>15.2.5.3>. </FONT></P>
<P><FONT size=2>Comment History: Dec 14, 2005 08:11:26 PM Old: New:Pending
By:Benner, Alan </FONT></P>
<P><FONT size=2>(bennera@us.ibm.com) </FONT></P>
<P><FONT
size=2>----------------------------------------------------------------------------</FONT></P>
<P><FONT size=2>Errata Tracking Number: MGTWG8372 </FONT></P>
<P><FONT size=2>Sub-Case Number: 1 </FONT></P>
<P><FONT size=2>Reference ID: 4292 </FONT></P>
<P><FONT size=2>Title: Enhanced SA PortInfoRecord searches </FONT></P>
<P><FONT size=2>Submitter: Livingston, James (James.Livingston@necsam.com)
</FONT></P>
<P><FONT size=2>Volume: 1 </FONT></P>
<P><FONT size=2>Revision: 1.2 </FONT></P>
<P><FONT size=2>Errata Release Version: 1.2 1/26/2006 </FONT></P>
<P><FONT size=2>Chapter/Subsection: 15.2.5.3 </FONT></P>
<P><FONT size=2>Page: 891 </FONT></P>
<P><FONT size=2>Line: 36 </FONT></P>
<P><FONT size=2>AssignedIntensity: </FONT></P>
<P><FONT size=2>Status/Disposition: WG_Approved </FONT></P>
<P><FONT size=2>Problem Description: Add optional compliance statement for new
capability. </FONT></P>
<P><FONT size=2>Problem Resolution: Make the proposed change to the spec text.
</FONT></P>
<P><FONT size=2>Original Text: <none> </FONT></P>
<P><FONT size=2>Corrected Text: o15-0.x.y: If SA's </FONT></P>
<P><FONT size=2>ClassPortInfo:CapabilityMask.IsPortInfoCapMaskMatchSupported
is 1, </FONT></P>
<P><FONT size=2>then the AttributeModifier of the SubnAdmGet() and
SubnAdmGetTable() </FONT></P>
<P><FONT size=2>methods affects the matching behavior on the
PortInfo:CapabilityMask </FONT></P>
<P><FONT size=2>component. If the high-order bit (bit 31) of the
AttributeModifier </FONT></P>
<P><FONT size=2>is set to 1, matching on the CapabilityMask component will not
be an </FONT></P>
<P><FONT size=2>exact bitwise match as described in <ref to 15.4.4>.
Instead, </FONT></P>
<P><FONT size=2>matching will only be performed on those bits which are set to
1 in </FONT></P>
<P><FONT size=2>the PortInfo:CapabilityMask embedded in the query. </FONT></P>
<P><FONT size=2></FONT></P>
<P><FONT size=2>In <ref to o15-0.x.y>, bits in the
PortInfo:CapabilityMask embedded </FONT></P>
<P><FONT size=2>in the query that are set to 0 are bitwise wildcards for
purposes of </FONT></P>
<P><FONT size=2>matching. </FONT></P>
<P><FONT size=2></FONT></P>
<P><FONT size=2>This gives a requester the ability to select desired
capabilities </FONT></P>
<P><FONT size=2>and query for ports which support those capabilities.
</FONT></P>
<P><FONT size=2></FONT></P>
<P><FONT size=2>If SA's
ClassPortInfo:CapabilityMask.IsPortInfoCapMaskMatchSupported </FONT></P>
<P><FONT size=2>is 0, or if bit 31 of the AttributeModifier is 0, then any
matching </FONT></P>
<P><FONT size=2>performed on the PortInfo:CapabilityMask component is as
described </FONT></P>
<P><FONT size=2>in <ref to 15.4.4>. </FONT></P>
<P><FONT size=2>Comment History: Dec 14, 2005 08:21:47 PM Old: New:Pending
By:Benner, Alan </FONT></P>
<P><FONT size=2>(bennera@us.ibm.com) </FONT></P>
<P><FONT
size=2>----------------------------------------------------------------------------</FONT></P><FONT
face="Palatino Linotype" color=#0000ff></FONT></SPAN></FONT></FONT>
<P><FONT size=+0><FONT face=Arial><FONT size=2><SPAN
class=396313908-04042006></SPAN></FONT></FONT></FONT> </P>
<P><FONT face=Arial
size=2>Ishai</FONT></P></BLOCKQUOTE></FONT></DIV></BODY></HTML>