perm filename ARPAUP.MAI[RFC,NET] blob sn#852588 filedate 1988-01-30 generic text, type C, neo UTF8
COMMENT ⊗   VALID 00016 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	∂29-Jan-88  1916	HA-NSC-LIST-RELAY@SRI-NIC.ARPA 	ARPANET PSN UPGRADE BULLETIN BOARD    
C00005 00003	28-Jan-88 04:32:47-PST,12833000000000001
C00008 00004
C00010 00005
C00012 00006
C00014 00007
C00016 00008
C00017 00009
C00019 00010
C00021 00011
C00022 00012
C00024 00013
C00025 00014
C00028 00015
C00029 00016
C00033 ENDMK
C⊗;
∂29-Jan-88  1916	HA-NSC-LIST-RELAY@SRI-NIC.ARPA 	ARPANET PSN UPGRADE BULLETIN BOARD    
Received: from SRI-NIC.ARPA by SAIL.Stanford.EDU with TCP; 29 Jan 88  19:16:09 PST
HA-NSC-Distribution
!DDN1.ARPA
abauman
!MC.LCS.MIT.EDU
arpanet-bboard
!A.ISI.EDU
bowers
!
Date: Fri, 29 Jan 88 16:59:56 PST
From: DDN Reference <NIC@λSRI-NIC.ARPAλ>
Subject: ARPANET PSN UPGRADE BULLETIN BOARD
To: HA-NSC-Distribution: ;
cc: abauman@λDDN1.ARPAλ,
    snively@λDDN1.ARPAλ,
    sccmgr@λDDN1.ARPAλ,
    bowers@λA.ISI.EDUλ,
    nic@λSRI-NIC.ARPAλ,
    arpanet-bboard@λMC.LCS.MIT.EDUλ
Message-ID: <12370586271.26.NIC@SRI-NIC.ARPA>

At the request of Code B613 of the DDN DCS, the NIC is establishing an
ARPANET PSN 7.0 Upgrade bulletin board.  The information sent to
this bboard will contain weekly status reports concerning problems and
solutions related to the upgrade of ARPANET PSNs to the new end-to-end
protocols.

If you choose to be on the distribution list for this bboard, you will
receive weekly status reports on upgrade developments, problems and
resolutions.  All status reports will be archived in a mail file on
SRI-NIC.ARPA.  Archived reports can either be FTPed using
login="anonymous" and password="guest" or requested through SERVICE,
the NIC's automatic mail service.  The pathname is
NETINFO:ARPA-UPGRADE.MAIL

If you want to be added to the distribution list, please send an
electronic mail request to NIC@SRI-NIC.ARPA.
-------

28-Jan-88 04:32:47-PST,12833;000000000001
Return-Path: <ABauman@ddn1.arpa>
Received: from ddn1.arpa by SRI-NIC.ARPA with TCP; Thu 28 Jan 88 04:31:52-PST
Date: 28 Jan 88 06:59 EST
From: ABauman @ DDN1.arpa
Subject: 28JAN88 PSN 7.0 status report

------------
Received:  from BBN.COM by DDN1.ARPA ; 27 Jan 88 16:47:52 EST
Received-2:  from bbn.com by BBN.COM id aa13946; 27 Jan 88 16:43 EST
Received-3:  from ccv.bbn.com by BBN.COM id aa13917; 27 Jan 88 16:43 EST
Date:      Wed, 27 Jan 88 16:30:45 EST
From: turkewit @ ccv.bbn.com
To: ARPAupgrade @ ccv.bbn.com, Bowers @ a.isi.edu, NCassidy @ ccv.bbn.com, DCAB610 @ DDN1.ARPA
cc: Turkewitz @ ccv.bbn.com
Subject:   ARPANET Weekly Status Report
Text: 

ARPANET PSN 7.0 Status Summary for NEE			      27 January 1988


This week's status report contains two new categories:

RESOLVED  PROBLEMS are problems that have been moved off the active list
because solutions exist for the user community, and there is  no  longer
development  work  being  done.   In  some cases, permanent SRB-approved
patches are not yet available, but temporary ones are deployed in  their
absence.

PATCHES DEPLOYED TO THE ARPANET is a list of patches that are currently
deployed to the network to solve PSN 7.0 problems.

	--Ken Turkewitz

1.  Hung VC

    Problem Description:

        This is a problem with an X.25 host on an IOP  call  that  would
	hang  the  original  VC that gets open to itself.  The X.25 host
	would not reply with a receiver ready (RR)  in  order  to  fully
	open  the connection and allow data to flow from the AHIP to the
	X.25 host.

   Current (26 Jan 88) Status:

	Currently, the interim patch (dev54) corrects interactions  with
	SUN  systems.  The interim patch has been deployed to the entire
	network.  The final patch is in the final  testing  process  and
	will be deployed 29 Jan and/or 30 Jan.

	IR  #  87-0055  is still open, until the final patch is in place
 	and accepted.

    Impact on Service:

	none, temporary patch acceptable

    Cross-referencing:  IR: 87-0055
			PR:
			Solution or Patch No.:  PSN 7.0 S079-03
			dev/pat/diag: dev54



2.  4.3 BSD systems hangup during cutover to NEE

    Problem Description:

	This is a PSN problem with Berkeley 4.3 systems  that  sometimes
	occurs  during  cutover  to  NEE.   This  does not happen during
	normal operation.  Work arounds to this include looping the host
	or restarting the node.

   Current (26 Jan 88) Status:

	BBNCC  development  is  working  directly  with  Mike  Karels of
	Berkeley to isolate the  reason  for  systems  hangup.   Cutover
	testing  during  holiday 18 Jan 88 with N#78 personel.  BBNCC is
	working on the permanent fix to this problem, and should have it
	ready  in time for use for the MILNET.  The NCR for redeployment
	of Testnode #115 is being processed for testing of the PSN patch
	for permanent correction of this host problem in the PSN.

    Impact on Service:

	None,  Workaround  successful,  patch needed by cutover time for
	MILNET

    Cross-referencing:  IR: 88-0005
			PR:
			Solution or Patch No.:
			dev/pat/diag: 


3.  HDH Precedence Level

    Problem Description:

        This  problem occurs on HDH packet hosts.  Messages delivered at
	maximum precedence to the host can cause the host to  reset  its
	connection.   As  a  workaround  to this PSN problem, precedence
	levels for HDH packet hosts are all  being  set  to  zero  (low)
	precedence.   BBNCC  is working on an eventual patch to fix this
	problem.  In the meantime, there is no impact to the  users,  if
	the workaround is deployed.

    Current (26 Jan 88) Status:

	Workaround in place, permanent corrective patch is being tested.

    Impact on Service:

	none

    Cross-referencing:  IR: 88-0001
			PR:
			Solution or Patch No.:
			dev/pat/diag:
			
    Users that Have Reported the Problem:


4.  HDH Gateway resetting themselves once or twice a day

    Problem Description:

        We have an open problem with Gateways restarting once or twice a
	day.  BBNCC Development is actively trying  to  determine  where
	the  problem  lies. We have been working closely with MIT#6 host
	4.  Thinking Machines Inc. maintains the host  and  are  helping
	from  their end.  MIT# 6 host 4 and ISI 22 host 5 are hosts with
	this problem.

    Current (26 Jan 88) Status:

	BBNCC Development is actively  looking  into  the  problem,  but
	haven't  been  able  to  isolate the it yet. This is the highest
	priority problem and a team of Developers are working to  figure
	out  why  these hosts are being reset. We have contacted ACC for
	support on how their HDH board is being affected with the NEE.


    Impact on Service:

	Users are still having problems with resets once or twice a day,
	causing momentary interruption  (1  minute)  of  service.   Site
	check  indicates  minimal  impact  on  daily  operations.  BBNCC
	Development has this as their highest priority.

    Cross-referencing:  IR:
			PR:
			Solution or Patch No.:
			dev/pat/diag:
			

5.  Traffic not passing between two CMU hosts

    Problem Description:

        Jim  Ellis  (6 January) reported a problem involving traffic not
	being passed between two of their hosts.

    Current (27 Jan 88) Status:

        Problem went away.  We are continuing to  monitor  these  hosts.
	If no more problems appear by 5 February, this problem will move
	to the RESOLVED list.

    Impact on Service:

	none

    Cross-referencing:  IR:
			PR:
			Solution or Patch No.:
			dev/pat/diag:


6.  HDLC flow-control parameters

    Problem Description:

	This  problem  involves  flow-control  parameters for hosts when
	initially configured for New End-to-End.

    Current (26 Jan 88) Status:

        Workaround in place

	We are working on the fix, to be in place in time for the MILNET
	cutover.   Patch  has  been tested successfully will be reviewed
	shortly by the Software Review Board.

    Impact on Service:

	none

	When  the  ARPANET  was  first  cutover to NEE, this problem was
	worked around by modifying the parameters, so it is no longer  a
	problem  for  the  ARPANET.   Therefore,  users  do not see this
	problem.

    Cross-referencing:  IR: 87-0034
			PR:
			Solution or Patch No.:
			dev/pat/diag:
			
	This was a network wide problem.

7. Undetermined crashes

    Problem Description:

	Several   node   crashes   are   under  investigation  by  BBNCC
	Development.  This only includes outages that are related to the
	NEE.

    Current (26 Jan 88) Status:

	BBNCC  Development  is  looking  at  all  dumps to determine the
	reasons for the service disruptions.  Dumps  for  the  following
	nodes are currently under investigation:

		At Lincoln #10(2)
		At CSC #25
		AT UCLA #1

    Impact on Service:

	These  were  all  isolated  cases,  causing  temporary   service
	disruptions.

    Cross-referencing:  IR: Each crash/dump has a separate IR.
			PR:
			Solution or Patch No.:
			dev/pat/diag:
			
	These outages are all reported by the ARPANET Monitoring Center,
	not by users.

8. AHIP ready line problems with gateways using ECUs

    Problem Description:

        This appears to be a problem with the interaction of the PSN and
        the ECUs regarding the ready line.

    Current (26 Jan 88) Status:

        Both  sites  have been provided a workaround by Charles Hedrick.
	BBNCC Development  is  working  with  Charles  to  isolate  this
	problem.

    Impact on Service:

	none, workaround in place

    Cross-referencing:  IR:
			PR:
			Solution or Patch No.:
			dev/pat/diag:

9.  X.25 host connection

    Problem Description:

        Bob  Albrighton  contacted the NOC 11 Jan that his X.25 host was
        having connectivity problems.  Bob believes  problem  lies  with
	his reconfigured host software.

    Current (27 Jan 88) Status:

	Still pending further informaton from the site.  He is trying to
	install new software.  We will continue to work with Bob.  If no
	more problems appear by 5 February, this problem  will  move  to
	the resolved list.

    Impact on Service:

	The host has never been up.

    Cross-referencing:  IR:
			PR:
			Solution or Patch No.:
			dev/pat/diag:   


10. Host can't ping itself; rejecting call accept packet

    Problem Description:

	This was originally a problem with Host Access  being  disabled.
	Now  John  Nelson's  host  can't make calls to itself.  The X.25
	diagnostic cause is 67.

    Current (27 Jan 88) Status:

	John is checking his host software.  We continue to  check  with
	John  to  make  sure we understand what the problem turns out to
	be.

    Impact on Service:

	Minimal--He can reach all hosts he needs on the Internet.

    Cross-referencing:  IR:
			PR:
			Solution or Patch No.:
			dev/pat/diag:


RESOLVED PROBLEMS:


1.  HDH Buffer management problem

    Problem Description:

        The  problem existed whereby the PSN would lose input buffers on
	the HDH hosts when the HDLC link was declared down.

    IR: 87-0054


2.  128-byte packet problem

    Problem Description:

        This is a PSN problem in which a 128-byte packet gets  delivered
	as  a 126-byte packet with the an exception bit (M, Q or D) set.
	SUN hosts are currently reporting this problem.  It was found to
	be an IOP or X.25 standard service problem.  

    IR: 88-0002


3.  Stuck X.25 VCs  (erroneously lasted last week as "Hung X.25 VCs")

    Problem Description:

	This was an X.25 Standard Service problem.  The  PSN  would  get
	into a state in which the delivery window would close.  This was
	most prevalent with Gateway sites that ping each other to  check
	if another gateway is up.

    IR: 87-0053


4.  HDLC and modem control

    Problem Description:

        Mike  Boring  reported  problems  since  8  Jan  with  his host.
	Initial trouble showed that there was no level 1 modem  signals.
	The problem was traced to a known problem with HDLC. 

    IR:88-0004


5.  AHIP-L NOP

    Problem Description:

        Martin Frost identified a problem with hosts  at  Stanford  that
	were  configured  to  receive  an  AHIP  type  NOP,  and the PSN
	responded with an AHIP-L NOP.  This is a configurable  parameter
	in the PSN for AHIP hosts.  The configuration change was made 18
	Jan 88.  We will check that the change corrected the problem.

    IR:  none

OTHER ARPANET STATUS INFORMATION:

	Minor problems with statistics are being worked  on  to  collect
	data  next  week.   This  data  will be gathered with the latest
	network analysis recomendations for buffer allocations in place.

PATCHES DEPLOYED TO THE ARPANET as of 22 Jan 88

								     DATE
DEV & DIAGS    IR NUMBER                SOLUTION #*     PAT #   P #  DEPLOYED
-----------    ---------                ----------      -----   --   --------
                                   
DEV458                             
	                              
DEV21          IR87-0038-ARPANET-SW     S040-01         pat11
	                           
DEV10          IR87-0040-ARPANET-SW     SO28-01
	                           
SPEC09         IR87-0020-MILNET-SW    
	                           
DIAG10         IR87-0038-ARPANET-SW   
                                   
DEV15          IR87-0036-ARPANET-SW     S033-02         pat17
	                           
DEV16          IR87-0043-ARPANET-SW     S034-01         pat21
	                           
DEV28          IR87-0044-ARPANET-SW     S051-01
	                           
DEV32          IR87-0045-ARPANET-SW     S055-01
	                           
DEV33          IR87-0046-ARPANET-SW     S056-01
	                           
DEV48          IR87-0050-ARPANET-SW     S074-01                      12/17/87
	                           
PAT4           IR87-0050-ARPANET-SW                     pat4
	                           
DEV22          IR87-0051-ARPANET-SW     S042-01         pat25
	                           
DEV52          IR87-0053-ARPANET-SW     S077-01         pat28
	                           
DIAG38/dev57   IR87-0054-ARPANET-SW       
	                           
DEV54          IR87-0055-ARPANET-SW     S079-03                      1/13/88
	                           
DEV59          IR88-0002-ARPANET-SW     S118-01                      1/12/88

pat20          IR in process                             pat20

diag39                                                               1/11/88**


**Note:  Not deployed to entire network, just nodes requested by development
	 for diagnostic purposes