From owner-Big-Internet@munnari.OZ.AU Sun Apr 16 05:38:34 1995 Received: from murtoa.cs.mu.OZ.AU by munnari.oz.au with SMTP (5.83--+1.3.1+0.50) id AA12809; Sun, 16 Apr 1995 05:38:34 +1000 (from owner-Big-Internet@munnari.OZ.AU) Return-Path: Received: from mailing-list by murtoa.cs.mu.OZ.AU (8.6.9/1.0) id FAA24773; Sun, 16 Apr 1995 05:38:14 +1000 Received: from munnari.oz.au by murtoa.cs.mu.OZ.AU (8.6.9/1.0) with SMTP id FAA24757; Sun, 16 Apr 1995 05:21:57 +1000 Precedence: list Received: from Mordor.Stanford.EDU by munnari.oz.au with SMTP (5.83--+1.3.1+0.50) id AA12348; Sun, 16 Apr 1995 05:21:51 +1000 (from dcrocker@mordor.stanford.edu) Received: from [198.120.32.25] (arc-tac1-slip5.nsi.nasa.gov [198.120.32.25]) by Mordor.Stanford.EDU (8.6.11/8.6.6) with SMTP id MAA10873; Sat, 15 Apr 1995 12:20:33 -0700 Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Sat, 15 Apr 1995 12:21:04 -0700 To: yakov@watson.ibm.com From: Dave Crocker Subject: RFC 1787 - Routing in a Multi-provider Internet Cc: Big-Internet@munnari.OZ.AU Yakov, Your RFC opens the topic of global, inter-provider service and suggests a number of lines of inquiry. It does a good job of discussing a number of discrete technical issues, all of which I do believe need attention. In your summary at the end, you list follow-on work that pertains to development of several categories of solutions or pursuit of studies to understand some categories of problems. I'd like to suggest an additional activity. It is intended to cut through theoretical views and allow assessment of bottom-line issues simply and directly: Measure current and continuing inter-provider behaviors. Publish the results. This needs to be done on a global (and probably user-driven) basis. Providers are competent and well-intentioned, but they each have their own agendas, as your RFC notes. We need to develop an ongoing basis for assessing the performance qualities of the global Internet. Simple measures such as throughput, latency and reliability among randomly chosen sites are probably adequate. Does anyone else share this view? d/ -------------------- Dave Crocker Brandenburg Consulting +1 408 246 8253 675 Spruce Dr. fax: +1 408 249 6205 Sunnyvale, CA 94086 dcrocker@networking.stanford.edu From owner-Big-Internet@munnari.OZ.AU Sun Apr 16 06:59:38 1995 Received: from murtoa.cs.mu.OZ.AU by munnari.oz.au with SMTP (5.83--+1.3.1+0.50) id AA15213; Sun, 16 Apr 1995 06:59:38 +1000 (from owner-Big-Internet@munnari.OZ.AU) Return-Path: Received: from mailing-list by murtoa.cs.mu.OZ.AU (8.6.9/1.0) id GAA24867; Sun, 16 Apr 1995 06:58:14 +1000 Received: from munnari.oz.au by murtoa.cs.mu.OZ.AU (8.6.9/1.0) with SMTP id GAA24840; Sun, 16 Apr 1995 06:38:57 +1000 Precedence: list Received: from mulga.cs.mu.OZ.AU by munnari.oz.au with SMTP (5.83--+1.3.1+0.50) id AA14665; Sun, 16 Apr 1995 06:38:54 +1000 (from lixia@parc.xerox.com) Received: from alpha.Xerox.COM by mulga.cs.mu.OZ.AU with SMTP (5.83--+1.3.1+0.50); id AA14984 Sun, 16 Apr 1995 06:20:44 +1000 (from lixia@parc.xerox.com) Received: from redwing.parc.xerox.com ([13.2.116.19]) by alpha.xerox.com with SMTP id <14401(6)>; Sat, 15 Apr 1995 13:18:06 PDT Received: by redwing.parc.xerox.com id <177520>; Sat, 15 Apr 1995 13:18:00 -0700 Date: Sat, 15 Apr 1995 13:17:57 PDT Sender: Lixia Zhang From: Lixia Zhang To: Dave Crocker Cc: Big-Internet@munnari.OZ.AU Subject: Re: RFC 1787 - Routing in a Multi-provider Internet In-Reply-To: Your message of Sat, 15 Apr 1995 12:21:04 -0700 Message-Id: > This needs to be done on a global (and probably user-driven) basis. > Providers are competent and well-intentioned, but they each have their own > agendas, as your RFC notes. We need to develop an ongoing basis for > assessing the performance qualities of the global Internet. Simple > measures such as throughput, latency and reliability among randomly chosen > sites are probably adequate. > > Does anyone else share this view? Yes I do. But as you probably know already, Matt Mathis of PSC chaired a successful provider performance measurement BOF at Danvers IETF. As I heard the conclusion from the BOF is to continue the effort by setting up a new WG. Would you think this new WG to be a right place, and an adequate step, to address the issue in hand, or you believe additional efforts necessary? Lixia From owner-Big-Internet@munnari.OZ.AU Sun Apr 16 09:00:23 1995 Received: from murtoa.cs.mu.OZ.AU by munnari.oz.au with SMTP (5.83--+1.3.1+0.50) id AA18433; Sun, 16 Apr 1995 09:00:23 +1000 (from owner-Big-Internet@munnari.OZ.AU) Return-Path: Received: from mailing-list by murtoa.cs.mu.OZ.AU (8.6.9/1.0) id IAA24985; Sun, 16 Apr 1995 08:58:16 +1000 Received: from munnari.oz.au by murtoa.cs.mu.OZ.AU (8.6.9/1.0) with SMTP id IAA24958; Sun, 16 Apr 1995 08:39:06 +1000 Precedence: list Received: from mulga.cs.mu.OZ.AU by munnari.oz.au with SMTP (5.83--+1.3.1+0.50) id AA17749; Sun, 16 Apr 1995 08:38:53 +1000 (from dcrocker@mordor.stanford.edu) Received: from Mordor.Stanford.EDU by mulga.cs.mu.OZ.AU with SMTP (5.83--+1.3.1+0.50); id AA16267 Sun, 16 Apr 1995 08:24:31 +1000 (from dcrocker@mordor.stanford.edu) Received: from [198.120.32.25] (arc-tac2-slip2.nsi.nasa.gov [198.120.32.46]) by Mordor.Stanford.EDU (8.6.11/8.6.6) with SMTP id PAA11406; Sat, 15 Apr 1995 15:21:46 -0700 Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Sat, 15 Apr 1995 15:21:58 -0700 To: Lixia Zhang From: Dave Crocker Subject: Re: RFC 1787 - Routing in a Multi-provider Internet Cc: Big-Internet@munnari.OZ.AU At 1:17 PM 4/15/95, Lixia Zhang wrote: >Would you think this new WG to be a right place, and an adequate step, >to address the issue in hand, or you believe additional efforts >necessary? my apologies. I completed missed the fact of that BOF. (Managed to get distracted by far less technical issues at Danvers.) It certainly sounds like the wg should be just the right place to pursue this. many thanks for mentioning it. d/ -------------------- Dave Crocker Brandenburg Consulting +1 408 246 8253 675 Spruce Dr. fax: +1 408 249 6205 Sunnyvale, CA 94086 dcrocker@networking.stanford.edu From owner-Big-Internet@munnari.OZ.AU Mon Apr 17 00:18:53 1995 Received: from murtoa.cs.mu.OZ.AU by munnari.oz.au with SMTP (5.83--+1.3.1+0.50) id AA12766; Mon, 17 Apr 1995 00:18:53 +1000 (from owner-Big-Internet@munnari.OZ.AU) Return-Path: Received: from mailing-list by murtoa.cs.mu.OZ.AU (8.6.9/1.0) id AAA25773; Mon, 17 Apr 1995 00:18:34 +1000 Received: from munnari.oz.au by murtoa.cs.mu.OZ.AU (8.6.9/1.0) with SMTP id AAA25757; Mon, 17 Apr 1995 00:01:06 +1000 Precedence: list Received: from clark.net by munnari.oz.au with SMTP (5.83--+1.3.1+0.50) id AA12086; Mon, 17 Apr 1995 00:01:01 +1000 (from hcb@clark.net) Received: (hcb@localhost) by clark.net (8.6.12/8.6.5) id KAA13996; Sun, 16 Apr 1995 10:00:50 -0400 From: Howard Berkowitz Message-Id: <199504161400.KAA13996@clark.net> Subject: Re: RFC 1787 - Routing in a Multi-provider Internet To: dcrocker@networking.stanford.edu (Dave Crocker) Date: Sun, 16 Apr 1995 10:00:50 -0400 (EDT) Cc: yakov@watson.ibm.com, Big-Internet@munnari.OZ.AU In-Reply-To: from "Dave Crocker" at Apr 15, 95 12:21:04 pm X-Mailer: ELM [version 2.4 PL24alpha3] Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Content-Length: 2991 Dave Crocker suggests to Yakov: > > Yakov, > > Your RFC opens the topic of global, inter-provider service and > suggests a number of lines of inquiry. It does a good job of discussing a > number of discrete technical issues, all of which I do believe need > attention. > > In your summary at the end, you list follow-on work that pertains > to development of several categories of solutions or pursuit of studies to > understand some categories of problems. > > I'd like to suggest an additional activity. It is intended to cut > through theoretical views and allow assessment of bottom-line issues simply > and directly: Measure current and continuing inter-provider behaviors. > Publish the results. > > This needs to be done on a global (and probably user-driven) basis. > Providers are competent and well-intentioned, but they each have their own > agendas, as your RFC notes. We need to develop an ongoing basis for > assessing the performance qualities of the global Internet. Simple > measures such as throughput, latency and reliability among randomly chosen > sites are probably adequate. > > Does anyone else share this view? Dave, I very much would be interested in seeing performance characterization of different providers. My question/observation here is the set of parameters used for this characterization. Do remember that I am a recovering internationalstandardsaholic, so I sometimes slip back to my origins. There are, however, several standards that actually might be useful here: ANS X3.102 and X3.141, and their internationalizations into CCITT X.140. This family of standards is intended to describe user-oriented performance, rather than the component-oriented metrics used by Scott Bradner. They were, however, intended to allow comparison among data service providers with different underlying technology. They are sufficiently complex that I don't want to go through the algorithms here. I've written tutorials on them, which I will work on making available online. The basic model, however, defines a technology-independent set of "reference events" such as Start of Access, Access Failure due to System Blocking, User Information Transfer Attempt, etc., and defines the performance parameters on these abstract events. A specific set of mappings to the Reference Events is then defined to protocol-specific Interface Events (vaguely like porting the UNIX kernel). Early versions of the X3.102 parameters were used to evaluate ARPANET and X.25 provider performance. I have never seen them used in the Internet, and wonder if this is just a matter of their not being noticed? Unless there are technical reasons to the contrary, I suggest they be considered. Traditional data carriers (e.g., X.25) are probably already familiar with them. Howard Berkowitz PSC International Slight disclaimer--I am an official coauthor of X3.141, and of the predecessor document to X3.102, INT-FED-STD-1033.