Index Journeys
RFC 622 UNKNOWN INFRASTRUCTURE · OPERATIONS

Scheduling IMP/TIP down time

How the network planners notified sites about scheduled maintenance on switching nodes and terminal interfaces

On rfc-editor.org ↗
PUBLISHED
March 1974
AUTHORS
A.M. McKenzie
STREAM
Legacy
DOI
10.17487/RFC0622

RFC text

Original text · RFC Editor View on rfc-editor.org ↗ 
Network Working Group                                     Alex McKenzie
RFC # 622                                                 BBN-NET
NIC # 21995                                               March 13, 1974


                      Scheduling IMP/TIP Down Time


From time to time it is necessary for an IMP or TIP to be taken down
other than during its scheduled Preventive Maintenance period.  Examples
of such cases are retrofit programs and the off-line diagnosis and
repair of intermittent problems. In the past we have scheduled these
downs with our "primary site contact" at the affected site; in our view
this individual was responsible for representing all of the hosts (and
TIP users) connected to that IMP/TIP. Although this policy has worked
well in the past, the recent proliferation of Very Distant Hosts,
together with complaints from representatives of other (local or
distant) Hosts, has caused us to review our procedures.

This note announces a modification of the previous policy. We will
continue to schedule IMP/TIP downs with our "primary site contact."  We
will continue to consider this individual a representative of all Hosts
connected to that IMP/TIP, with the following exception. Any Host
organization wishing to be notified by telephone that down time has been
scheduled for their IMP/TIP may nominate some individual as a "secondary
contact."  The name of the secondary contact, together with a telephone
number where he/she can be reached, should be submitted to me by any of
the following methods:

    US mail

        Network Control Center
        Bolt, Beranek and Newman Inc.
        50 Moulton Street
        Cambridge, Mass. 02138
        ATTN: Alexander McKenzie

    Journal system

        Ident = AAM

    Network mail

        MCKENZIE at BBN-TENEX

The NCC operators will notify the secondary contact by telephone after
scheduling down time with the primary contact.

Following is the current list of "primary site contacts" (and backup
"primary" contacts):

    UCLA - Anita Coley (Dave Crocker, Lou Nelson)
    SRI - Jim White (Don Wallace)
    UCSB - Ron Stoughton (Bob Ploger)
    Utah - Dennis Ting
    BBN (IMP) - Steve Chipman
    MIT-MAC - Al Vezza (Abhay Bhushan)
    RAND - Roy Yashimura
    SDC - Ken Brandon (Doug Pintar)
    Harvard - Bob Nickerson
    Lincoln - John Laynor
    Stanford - Ted Panofsky (Lester Earnest)
    Illinois - Jack Bouknight (Gary Grossman)
    Case - Dave Reide (Alan Rosenfeld)
    Carnegie - Bill Broadley
    AMES IMP - John McConnell (Paul Zima, Dick Brown)
    AMES TIP  - Bob Linebarger (Toby Gonzales, Glen Holtzer)
    MITRE -   Jean Iseli (Susan Poh)
    RADC - Tom Lawrence (Grant Strength)
    NBS - Tom Pyke (Bob Rosenthal)
    ETAC - Capt. Petregal (John Smith)
    LLL - Robert Abbott (Bob Sherwood)
    ISI - Tom Boynton (Jerry Pipes)
    USC - James Pepin
    GWC - Capt. Goldsmith (Peggy Irving)
    DOCB - Dave Lillie (Sky Stevenson)
    SDAC - Buz Owen (Vernon Bruffey, Rick Perez, Bill Whyte)
    Belvoir - Ken Gibson (Larry Wright)
    ARPA - Steve Crocker (Pam Cutler)
    Aberdeen - Melvin Wrublewski
    CCA - Phil Peterson (Hal Murray)
    Xerox - Robert Taylor (Bob Metcalfe)
    FNWC -  R. Mottencamper (Merrill Peterson)
    LBL - Bob Fink (Jose Alvarez)
    UCSD - Jim Madden (Grant Rostad)
    Hawaii - John Davidson
    RML - Mike Young
    NORSAR - Per Tveitane (Oddnund Hansen)
    London - Peter Kirstein (Hugh Gamble)
    Tymshare - Mike Marrah (Robert Martinez)
    MIT-IPG - Leo J. Ryan (Mike Padlipsky)
    Moffett - Ed Mortenson (Gene Lichner)
    Rutgers - Tom Webb (Jeff Broido)
    Wright-Patterson - Capt. Woodward (Fred Pitts)

Please let me know if the "primary" contact for your site should be
changed.










       [ This RFC was put into machine readable form for entry ]
       [ into the online RFC archives by Alex McKenzie with    ]
       [ support from GTE, formerly BBN Corp.            10/99 ]

AI ELI5

On this page
PROBLEMThe network needed to take down IMPs (Interface Message Processors) and TIPs (Terminal Interface Processors) for maintenance and repairs, but the previous notification policy did not reach all affected organizations.
MECHANISMThe RFC modifies the policy to allow any host organization to nominate a secondary contact who will be notified by telephone when their IMP or TIP is scheduled for downtime, in addition to the primary site contact.
IMPROVED SITE NOTIFICATIONThis gave individual organizations better visibility into planned maintenance that affected their operations, adapting the notification system to the growing network.

§1The Notification Problem

From time to time, network switching nodes called IMPs and TIPs needed to go offline for maintenance: retrofit programs, repairs of intermittent problems, and scheduled preventive maintenance. The Network Control Center had previously coordinated these downtimes by contacting only the primary site contact at each installation, trusting them to represent all the hosts and users connected there.

As the network grew with many distant hosts spread across different organizations at single sites, this approach left many users and administrators in the dark about upcoming maintenance.

When a network box needed work, the bosses only told one main contact person at that site. But if many different groups used that same network box, the others might not find out.

§2The New Secondary Contact System

The RFC modifies the policy to allow any host organization to nominate a secondary contact. This person would receive a telephone notification after the Network Control Center schedules the downtime with the primary contact.

Organizations could submit the secondary contact's name and telephone number through three methods: US mail to the Network Control Center at BBN in Cambridge, the ARPANET journal system (ident AAM), or network mail to MCKENZIE at BBN-TENEX.

Now any group at a site could say: please also call this person when there's maintenance scheduled. The network would call them on the phone.

§3Impact and Historical Context

The RFC includes a comprehensive list of the primary site contacts at 34 network locations, from UCLA and Stanford to international sites like London and NORSAR. This mechanism represented an early attempt to scale notification procedures as ARPANET grew from a few dozen nodes to a larger, more distributed system with multiple autonomous organizations at single sites.

The policy reflected the operational challenges of coordinating maintenance across a growing network where human communication and notification were still the primary mechanisms for coordination.

The list showed all the main people they called about maintenance. This change helped as the network grew bigger and had more organizations sharing the same equipment.