DPDK community structure changes
 help / color / Atom feed
* [dpdk-moving] Minutes from "Moving DPDK to Linux Foundation" call, February 14th
@ 2017-02-16 12:15 O'Driscoll, Tim
  0 siblings, 0 replies; only message in thread
From: O'Driscoll, Tim @ 2017-02-16 12:15 UTC (permalink / raw)
  To: moving

[-- Attachment #1: Type: text/plain, Size: 4696 bytes --]

Here are my notes from yesterday's call. Please feel free to correct any errors or to add additional details.



Attendees: Bruce Richardson (Intel), Erez Scop (Mellanox), George Zhao (Huawei), Jan Blunck (Brocade), Jaswinder Singh (NXP), Jim St. Leger (Intel), John Bromhead (Cavium), John McNamara (Intel), Keith Wiles (Intel), Kevin Traynor (Red Hat), Thomas Monjalon (6WIND), Tim O'Driscoll (Intel), Vincent Jardin (6WIND).



Firstly, here are some links to help keep track of things:

Project Charter: https://docs.google.com/document/d/1x43ycfW3arJNX-e6NQt3OVzAuNXtD7dppIhrY48FoGs

Budget estimate: https://docs.google.com/spreadsheets/d/1-3686Xb_jf4FtxdX8Mus9UwIxUb2vI_ppmJV5GnXcLg/edit#gid=302618256

Reference Lab Proposal: http://dpdk.org/ml/archives/moving/2017-February/000177.html

Technical governance, including info on Maintainers and sub-trees: http://dpdk.org/doc/guides/contributing/index.html & http://dpdk.org/dev.

Summary of discussion at Userspace event in Dublin: http://dpdk.org/ml/archives/dev/2016-October/049259.html

Minutes of January 10th call: http://dpdk.org/ml/archives/moving/2017-January/000136.html

Minutes of January 17th call: http://dpdk.org/ml/archives/moving/2017-January/000157.html

Minutes of January 24th call: http://dpdk.org/ml/archives/moving/2017-January/000162.html

Minutes of January 31st call: http://dpdk.org/ml/archives/moving/2017-February/000176.html

Minutes of February 7th call: http://dpdk.org/ml/archives/moving/2017-February/000179.html



Open Reference Lab:

- Agreed that an Open Reference Lab is a good idea. It would help to identify performance issues and would improve quality and stability.

- We discussed whether all performance numbers should be made public. It was felt that there may be cases where some vendors do not want absolute numbers published, and that in those cases we could publish delta numbers to show the difference between one performance test and another. This could happen in cases where DPDK is being supported on a new platform (e.g. as is happening now with NXP) and performance has not yet reached the expected level. This would just be a temporary situation until performance tuning is complete. Thomas suggested that we give each vendor the option, and let them decide when they're ready to publish absolute numbers.

- Agreed that the lab should be referred to as the Open Reference Lab.

- Agreed that the scope is to identify performance regression, not to performance test new features and/or new hardware. Vendors will still want to performance test DPDK on their latest hardware, but that should be done internally in their own labs.

- Discussed the frequency of testing. Ideally this would be done per patch so that we know the performance impacts of each patch and can take that into account when agreeing whether the patch should be accepted into DPDK. Testing per patch will increase the resource requirements, so we'll need to take that into account.

- Discussed how we want to maintain the lab. This is related to the decision on where we want to host it. Ideally, it should be administered by people familiar with DPDK and familiar with the hardware.

- Jan suggested looking at how other projects have handled admin for open labs. We'll look into this.

- Agreed that it would be useful to allow people to submit software applications to be run in the lab, as Matt suggested in a previous discussion. This will add some complexity.

- Discussed how we ensure that the lab doesn't become a competitive environment for vendors to show that their performance is the best. We need to agree some rules to prevent this.

- We need to look at hosting options (not just the LF, but other options too). To do this, we need to know hardware and power requirements. We haven't got to that level of detail yet.

- We'll discuss further in a smaller group. Anybody interested in participating in those discussions should let me know.



Tech Board:

- A tech board meeting was held yesterday. Minutes will be published. Making tech board meetings public was one of the agenda items.





Next Meeting:

Tuesday February 21st at 3pm GMT, 4pm CET, 10am EST, 7am PST. Access numbers for the call are:

  Skype meeting: https://meet.intel.com/tim.odriscoll/97F0Q1HF

  France: +33 1588 77298

  UK: +44 179340 2663

  USA: +1 916 356 2663

  Bridge Number: 5

  Conference ID: 120705852

We need to discuss how membership discussions are progressing and if our target date of a project launch to coincide with ONS (http://events.linuxfoundation.org/events/open-networking-summit, April 3-6) is realistic.




[-- Attachment #2: Type: text/html, Size: 9439 bytes --]

<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=us-ascii">
<meta name="Generator" content="Microsoft Word 15 (filtered medium)">
<style><!--
/* Font Definitions */
@font-face
	{font-family:"Cambria Math";
	panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
	{font-family:Calibri;
	panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0in;
	margin-bottom:.0001pt;
	font-size:11.0pt;
	font-family:"Calibri",sans-serif;}
a:link, span.MsoHyperlink
	{mso-style-priority:99;
	color:#0563C1;
	text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{mso-style-priority:99;
	color:#954F72;
	text-decoration:underline;}
p.MsoPlainText, li.MsoPlainText, div.MsoPlainText
	{mso-style-priority:99;
	mso-style-link:"Plain Text Char";
	margin:0in;
	margin-bottom:.0001pt;
	font-size:11.0pt;
	font-family:"Courier New";}
span.EmailStyle17
	{mso-style-type:personal-compose;
	font-family:"Calibri",sans-serif;
	color:windowtext;}
span.PlainTextChar
	{mso-style-name:"Plain Text Char";
	mso-style-priority:99;
	mso-style-link:"Plain Text";
	font-family:"Courier New";}
.MsoChpDefault
	{mso-style-type:export-only;
	font-family:"Calibri",sans-serif;}
@page WordSection1
	{size:8.5in 11.0in;
	margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
	{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]-->
</head>
<body lang="EN-US" link="#0563C1" vlink="#954F72">
<div class="WordSection1">
<p class="MsoPlainText">Here are my notes from yesterday's call. Please feel free to correct any errors or to add additional details.<o:p></o:p></p>
<p class="MsoPlainText"><o:p>&nbsp;</o:p></p>
<p class="MsoPlainText">Attendees: Bruce Richardson (Intel), Erez Scop (Mellanox), George Zhao (Huawei), Jan Blunck (Brocade), Jaswinder Singh (NXP), Jim St. Leger (Intel), John Bromhead (Cavium), John McNamara (Intel), Keith Wiles (Intel), Kevin Traynor (Red
 Hat), Thomas Monjalon (6WIND), Tim O'Driscoll (Intel), Vincent Jardin (6WIND).<o:p></o:p></p>
<p class="MsoPlainText"><o:p>&nbsp;</o:p></p>
<p class="MsoPlainText">Firstly, here are some links to help keep track of things:<o:p></o:p></p>
<p class="MsoPlainText">Project Charter: <a href="https://docs.google.com/document/d/1x43ycfW3arJNX-e6NQt3OVzAuNXtD7dppIhrY48FoGs">
https://docs.google.com/document/d/1x43ycfW3arJNX-e6NQt3OVzAuNXtD7dppIhrY48FoGs</a><o:p></o:p></p>
<p class="MsoPlainText">Budget estimate: <a href="https://docs.google.com/spreadsheets/d/1-3686Xb_jf4FtxdX8Mus9UwIxUb2vI_ppmJV5GnXcLg/edit#gid=302618256">
https://docs.google.com/spreadsheets/d/1-3686Xb_jf4FtxdX8Mus9UwIxUb2vI_ppmJV5GnXcLg/edit#gid=302618256</a><o:p></o:p></p>
<p class="MsoPlainText">Reference Lab Proposal: <a href="http://dpdk.org/ml/archives/moving/2017-February/000177.html">
http://dpdk.org/ml/archives/moving/2017-February/000177.html</a><o:p></o:p></p>
<p class="MsoPlainText">Technical governance, including info on Maintainers and sub-trees:
<a href="http://dpdk.org/doc/guides/contributing/index.html">http://dpdk.org/doc/guides/contributing/index.html</a> &amp;
<a href="http://dpdk.org/dev">http://dpdk.org/dev</a>.<o:p></o:p></p>
<p class="MsoPlainText">Summary of discussion at Userspace event in Dublin: <a href="http://dpdk.org/ml/archives/dev/2016-October/049259.html">
http://dpdk.org/ml/archives/dev/2016-October/049259.html</a><o:p></o:p></p>
<p class="MsoPlainText">Minutes of January 10th call: <a href="http://dpdk.org/ml/archives/moving/2017-January/000136.html">
http://dpdk.org/ml/archives/moving/2017-January/000136.html</a><o:p></o:p></p>
<p class="MsoPlainText">Minutes of January 17th call: <a href="http://dpdk.org/ml/archives/moving/2017-January/000157.html">
http://dpdk.org/ml/archives/moving/2017-January/000157.html</a><o:p></o:p></p>
<p class="MsoPlainText">Minutes of January 24th call: <a href="http://dpdk.org/ml/archives/moving/2017-January/000162.html">
http://dpdk.org/ml/archives/moving/2017-January/000162.html</a><o:p></o:p></p>
<p class="MsoPlainText">Minutes of January 31st call: <a href="http://dpdk.org/ml/archives/moving/2017-February/000176.html">
http://dpdk.org/ml/archives/moving/2017-February/000176.html</a><o:p></o:p></p>
<p class="MsoPlainText">Minutes of February 7th call: <a href="http://dpdk.org/ml/archives/moving/2017-February/000179.html">
http://dpdk.org/ml/archives/moving/2017-February/000179.html</a><o:p></o:p></p>
<p class="MsoPlainText"><o:p>&nbsp;</o:p></p>
<p class="MsoPlainText">Open Reference Lab:<o:p></o:p></p>
<p class="MsoPlainText">- Agreed that an Open Reference Lab is a good idea. It would help to identify performance issues and would improve quality and stability.<o:p></o:p></p>
<p class="MsoPlainText">- We discussed whether all performance numbers should be made public. It was felt that there may be cases where some vendors do not want absolute numbers published, and that in those cases we could publish delta numbers to show the difference
 between one performance test and another. This could happen in cases where DPDK is being supported on a new platform (e.g. as is happening now with NXP) and performance has not yet reached the expected level. This would just be a temporary situation until
 performance tuning is complete. Thomas suggested that we give each vendor the option, and let them decide when they&#8217;re ready to publish absolute numbers.<o:p></o:p></p>
<p class="MsoPlainText">- Agreed that the lab should be referred to as the Open Reference Lab.<o:p></o:p></p>
<p class="MsoPlainText">- Agreed that the scope is to identify performance regression, not to performance test new features and/or new hardware. Vendors will still want to performance test DPDK on their latest hardware, but that should be done internally in
 their own labs.<o:p></o:p></p>
<p class="MsoPlainText">- Discussed the frequency of testing. Ideally this would be done per patch so that we know the performance impacts of each patch and can take that into account when agreeing whether the patch should be accepted into DPDK. Testing per
 patch will increase the resource requirements, so we&#8217;ll need to take that into account.<o:p></o:p></p>
<p class="MsoPlainText">- Discussed how we want to maintain the lab. This is related to the decision on where we want to host it. Ideally, it should be administered by people familiar with DPDK and familiar with the hardware.<o:p></o:p></p>
<p class="MsoPlainText">- Jan suggested looking at how other projects have handled admin for open labs. We&#8217;ll look into this.<o:p></o:p></p>
<p class="MsoPlainText">- Agreed that it would be useful to allow people to submit software applications to be run in the lab, as Matt suggested in a previous discussion. This will add some complexity.<o:p></o:p></p>
<p class="MsoPlainText">- Discussed how we ensure that the lab doesn&#8217;t become a competitive environment for vendors to show that their performance is the best. We need to agree some rules to prevent this.<o:p></o:p></p>
<p class="MsoPlainText">- We need to look at hosting options (not just the LF, but other options too). To do this, we need to know hardware and power requirements. We haven&#8217;t got to that level of detail yet.<o:p></o:p></p>
<p class="MsoPlainText">- We&#8217;ll discuss further in a smaller group. Anybody interested in participating in those discussions should let me know.<o:p></o:p></p>
<p class="MsoPlainText"><o:p>&nbsp;</o:p></p>
<p class="MsoPlainText">Tech Board:<o:p></o:p></p>
<p class="MsoPlainText">- A tech board meeting was held yesterday. Minutes will be published. Making tech board meetings public was one of the agenda items.<o:p></o:p></p>
<p class="MsoPlainText"><o:p>&nbsp;</o:p></p>
<p class="MsoPlainText"><o:p>&nbsp;</o:p></p>
<p class="MsoPlainText">Next Meeting:<o:p></o:p></p>
<p class="MsoPlainText">Tuesday February 21st at 3pm GMT, 4pm CET, 10am EST, 7am PST. Access numbers for the call are:<o:p></o:p></p>
<p class="MsoPlainText">&nbsp; Skype meeting: <a href="https://meet.intel.com/tim.odriscoll/97F0Q1HF">
https://meet.intel.com/tim.odriscoll/97F0Q1HF</a><o:p></o:p></p>
<p class="MsoPlainText">&nbsp; France: &#43;33 1588 77298<o:p></o:p></p>
<p class="MsoPlainText">&nbsp; UK: &#43;44 179340 2663<o:p></o:p></p>
<p class="MsoPlainText">&nbsp; USA: &#43;1 916 356 2663&nbsp;&nbsp;&nbsp;&nbsp; <o:p></o:p></p>
<p class="MsoPlainText">&nbsp; Bridge Number: 5<o:p></o:p></p>
<p class="MsoPlainText">&nbsp; Conference ID: 120705852<o:p></o:p></p>
<p class="MsoPlainText">We need to discuss how membership discussions are progressing and if our target date of a project launch to coincide with ONS (<a href="http://events.linuxfoundation.org/events/open-networking-summit">http://events.linuxfoundation.org/events/open-networking-summit</a>,
 April 3-6) is realistic.<o:p></o:p></p>
<p class="MsoPlainText"><o:p>&nbsp;</o:p></p>
<p class="MsoNormal"><o:p>&nbsp;</o:p></p>
</div>
</body>
</html>

^ permalink raw reply	[flat|nested] only message in thread

only message in thread, back to index

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-02-16 12:15 [dpdk-moving] Minutes from "Moving DPDK to Linux Foundation" call, February 14th O'Driscoll, Tim

DPDK community structure changes

Archives are clonable:
	git clone --mirror http://inbox.dpdk.org/moving/0 moving/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 moving moving/ http://inbox.dpdk.org/moving \
		moving@dpdk.org
	public-inbox-index moving


Newsgroup available over NNTP:
	nntp://inbox.dpdk.org/inbox.dpdk.moving


AGPL code for this site: git clone https://public-inbox.org/ public-inbox