nanog mailing list archives
IETF BOF on Network Configuration (netconf)
From: Sean Donelan <sean () donelan com>
Date: Thu, 13 Mar 2003 16:20:39 -0500 (EST)
On Thu, 13 Mar 2003, Iljitsch van Beijnum wrote:
Too bad at least several of our collective favorite vendors don't seem to agree, as they don't provide very reasonable methods to update the router configuration in an automated way. Sure, there are ways to make it work but they are too complex to be useful in small networks. Iljitsch (Still waiting for vendors to support automatic filter retrieval from an LDAP server by routers...)
Will you be attending the Network Configuration BOF on Monday at the IETF in San Francisco? Configuration of networking devices has become a critical requirement for operators in today's highly interoperable networks. Operators from large to small have developed or used vendor specific mechanisms to transfer configuration data to and from a device and for examining device state information which may impact the configuration. Each of these mechanisms may be different in various aspects, such as session establishment, user authentication, configuration data exchange, and error responses. Utilities built upon tools such as Perl and "Expect" are used to control devices via the CLI, but are prone to failure due to the instability and lack of uniformity inherent in a CLI. Investigations conducted within the IETF, at OPS area meetings and in an IAB workshop over the past two years have identified operator requirements for a standard configuration protocol that: - Provides a clear separation of configuration data from non-configuration data - Is extensible enough that vendors will provide access to all configuration data on the box from a single protocol - Has a programmatic interface (avoids screen scraping and formatting-related changes between releases) - Uses a data representation that is easily manipulated using non-specialized text manipulation tools (perl, awk, etc.) - Supports integration with existing user authentication methods, such as RADIUS - Can be easily integrated with existing configuration database systems, such as RANCID - Provides support for multi-box configuration transactions (with locking and rollback capability) This BOF will focus on discussion of a protocol for the management of network device configuration that meets many of the operator requirements identified through these efforts. A draft that may serve as a useful starting point for this work can be found at http://www.ietf.org/internet-drafts/draft-enns-xmlconf-spec-00.txt.
Current thread:
- Re: route filtering in large networks, (continued)
- Re: route filtering in large networks Randy Bush (Mar 12)
- Re: route filtering in large networks E.B. Dreger (Mar 13)
- Re: route filtering in large networks Lars Erik Gullerud (Mar 13)
- Re: route filtering in large networks Alan Hannan (Mar 12)
- Re: route filtering in large networks Randy Bush (Mar 12)
- Re: route filtering in large networks Andy Dills (Mar 12)
- Re: route filtering in large networks Randy Bush (Mar 13)
- Re: route filtering in large networks Dorian Kim (Mar 13)
- Re: route filtering in large networks Stephen Sprunk (Mar 13)
- Re: route filtering in large networks Iljitsch van Beijnum (Mar 13)
- IETF BOF on Network Configuration (netconf) Sean Donelan (Mar 13)
- RE: Put part of Google on 69/8 (was Re: 69/8...this sucks) Vivien M. (Mar 12)
- Re: Put part of Google on 69/8 (was Re: 69/8...this sucks) Jack Bates (Mar 12)
- RE: Put part of Google on 69/8 (was Re: 69/8...this sucks) Vivien M. (Mar 12)
- Re: Put part of Google on 69/8 (was Re: 69/8...this sucks) Andy Dills (Mar 12)
- Re: Put part of Google on 69/8 (was Re: 69/8...this sucks) Valdis . Kletnieks (Mar 12)
- Re: Put part of Google on 69/8 (was Re: 69/8...this sucks) Owen DeLong (Mar 13)
- Re: Put part of Google on 69/8 (was Re: 69/8...this sucks) Owen DeLong (Mar 13)
- Re: Put part of Google on 69/8 (was Re: 69/8...this sucks) Stephen Sprunk (Mar 12)
- Re: Put part of Google on 69/8 (was Re: 69/8...this sucks) Tim Thorne (Mar 12)
- Re: 69/8...this sucks Valdis . Kletnieks (Mar 12)