JANET Midlands Regional User Group
Minutes of the 50th Meeting
14:30 on 17th October 2001 at Warwick
University
Present:
|
|
Chris Bayliss (CB) |
The University of
Birmingham |
|
|
Alan Benson (AB) |
South Birmingham College |
|
|
Max Caines (MC) |
Wolvehampton University |
|
|
James Hutton (JH) |
UKERNA |
|
|
Andy Jack (Chair) (AJ) |
University of Nottingham |
|
|
Maria MacCallum (MM) |
Warwick University |
|
|
Andy Morris (AM) |
West Midlands Regional
Support Centre |
|
|
Roy Pearce (RP) |
Birmingham University |
|
|
David Roberts (DR) |
University of Leicester |
|
|
Martin Whinnery(MW) |
South Birmingham College |
Neil Faver, Anil Bakrania and
Paul Maddock
Roy Pearce is based at The University of
Birmingham, not Aston University. Otherwise accepted.
None.
No additional agenda items.
A brief discussion was had regarding the
election of a chair, held over from the last meeting. In the absence of the
existing chair, Tim Clark, the meeting felt it was unfair make a change to the
position. The meeting aqreed to carry the item over to the next meeting.
The meeting was only aware
of the report covering the period April-June. JH reported that the Autumn report
covering July-Sept 2001 was being finalised. JH reported two significant items
in this report
i)
60% of UKERNA customers
were now covered in Netsight. AJ reported that the all the East Midlands sites
were being monitored except for the traffic levels of the University of Derby. JH asked that changes, additions or errors were
notified to nrg@ukerna.ac.uk. AJ thanked
Kevin Hoadley and Wally Jackson in their absence for their efforts. This was
echoed by the meeting. JH reported that the WAP interface to Netsight was now
considered part of the service and that documentation and details for the
service were available from Netsight itself, nottingham.netsight.ja.net. JH distributed a briefing paper on the Netsight
service.
ii)
The Managed Router
Service was rolling out and the Primary DNS Service was planned for roll out
shortly. Both services are aimed primarily at FE institutions.
JH
reported that use of the X400 service was falling, but that there were no plans
to discontinue the service.
CB
asked if the X29 service was still running. JH reported that it was. AJ noted
that he had a request from a user for access to the service. JH thought that
user accounts were still being allocated when they were requested.
AJ asked
if the Netsight service replaced the MAP downtime reports that were produced by
some MANs. JH said that it didn’t but that sites should contact Kevin Hoadley
for confirmation. MC asked whether the Netsight service was useful for
investigating outages. JH explained that the service held only information on
when the outages occurred and for how long. No data was held within the system
as to why the outage occurred, and therefore had no value in the investigation
of the causes of outages.
AJ
applauded item 2.9.6 in the report. The meeting felt that the benefit of
charging was in allowing sites to monitor their usage of the transatlantic
pipes, and more specifically the traffic patterns therein. The meeting felt
that implementing a system to provide similar information about the SJ4
backbone for an institution would be useful
JH agreed to take this suggestion back to Kevin Hoadley.
AJ
reported that Tim Clark had made available a set of perl scripts for processing
and graphing the data from bill.ja.net. He had found these especially useful
during the recent Code Red and Nimda
attacks. JH offered to feed this information back to UKERNA. AJ offered
to forward the URL to MidJUG@jiscmail.ac.uk
AJ also said that Chris Cheney at Cambridge had produced a similar package in
the form of a PC application for sites' itemised traffic data.
DR
asked if the outreach service mentioned in 5.4 of the Summer report was going
ahead in 2002. JH confirmed this, but said that there had been many volunteers
to pilot the service. Sites to take part in the pilot would be contacted in due
course. JH informed the meeting that a contract with BT had been signed in
connection with the pilot service.
AJ
noted that David Loames at Bradford was the webmaster of the JNUG site. He
informed the meeting that he would contact David Loames with a view to moving
the MidJUG site to the main JNUG site.
AJ
expressed a desire to know the planned new charging regime. The meeting echoed
this. JH reported that no decision on the topic had yet been made. JH also
noted that any charges incurred by FE institutions would continue to be paid by
LSC in 2002/2.
The
meeting had a discussion around the inconsistencies of bandwidth provision to
FE colleges, with most colleges having 2MB but some having as much as 6MB, or
2MB for each of their campuses. MM questioned the quoted price of
approximately £10K pa for an additional
2MB. JH reported that the discrepancies in costs were due to individual circumstances
and that the JCN had recommended reducing the cost of new connections and
additional bandwidth. JH acknowledged the need to document the current
situation. He added that UKERNA were less concerned about the speed that the
institution was connected, than the amount of traffic it generated.
JH
reported that, although costs currently incurred by colleges on their primary
connection were being paid for until 2002, LSC had announced no decision beyond
this point as to whether this would continue.
The
meeting felt that there were two issues with regard to multiply connected FE
sites, bandwidth and resilience. MM enquired as to UKERNA’s view of college
consortia pooling resources to improve bandwidth, resilience or both. JH
replied that UKERNA had no problem with this in principle, and stressed that
the fewer difficulties any connection created for UKERNA or ULCC the better.
UKERNA would encourage such cases to be resolved within the local MAN. JH
suggested that any ideas or proposals regarding FE connections should be
directed to Steven Percival.
AJ
made the group aware of the Usability Survey mentioned in the JNUG meeting
minutes under item 16.
RP
asked for clarification whether funding for the caching service was to continue
beyond next year. JH replied that no decision had been taken, but that this was
within the usual budgeting timescales, and that no decision was timetabled
until 2002. AJ questioned whether there were plans for caching to occur at MAN
level, and highlighted that the current charging scheme discouraged this
practice. JH was not aware of any plans and pointed out that the bottlenecks
that were a driving force behind caching no longer existed. The possible change
in charging methods may be more likely to encourage MAN level caching. The
group was concerned that without this financial motive driving caching, sites
may be less enthusiastic about the practice.
JH
reported that traffic based charging in line with the current method was unlikely to continue beyond July 2002
due to the technology not being available to maintain a suitable granularity of
sampling with such high volumes of traffic. Some bandwidth based charging
method was the currently preferred replacement.
AJ proposed
dropping this item from the MidJUG agenda as there had been no minutes
available for some time. DR reported that he had received the minutes by email
and that the UCISA-NG site had no webmaster.
DR
distributed proposals from UCISA-NG regarding the JANET SLA and asked that
comments should be directed to UCISA-NG. JH pointed out that the Netsight boxes
were in the current SLA.
AJ
reminded the group of the recent sobering experience that were the Code Red and
Nimda attacks.
The
group had a brief discussion around the problems, pitfalls and solutions
regarding networking Halls of Residence. The University of Warwick had opted
for a solution farmed out to an ISP, while Nottingham University had supplied
the service themselves. Birmingham University
had previously outsourced their residences networking which now seemed
to be working adequately. Who the students paid for the network service and
issues surrounding access to services which impose restrictions according to
source IP address were given particular attention.
There
were no issues reported.
The
group supported AJ’s positive comments on the quality of service provided by
SJ4.
There
were no issues to be reported.
Report from the Midlands Janet User Group.
MIDJUG met on Wednesday
17th October 2001 at Warwick University. The Midlands Networking Forum (MNF)
met in the morning with Edmund Sutcliffe giving a presentation on Wireless
LANs.
Some items from the UKERNA Summer
2001 Report were discussed (the Autumn report was not generally available at
the time). In particular, the introduction of the Netsight network monitors was
welcomed. Kevin Hoadley and Wally Jackson should be congratulated on their
efforts. It was hoped that these may go some way to replacing the MAP reports
although it is recognised that Netsight can never record the reasons for
particular outages.
The progress of the pilot
JANET ADSL service was also welcomed as several sites would probably take it up
when it becomes a formal service.
It was hoped that a
forthcoming meeting of the JISC assessors might lead
to an announcement on
charging soon. In line with the second paragraph on Item 2.9.6 of the Spring
UKERNA report, the transatlantic accounting and itemisation services had been
extremely useful in spotting and resolving problems over the summer with worms
such as Code Red and Denial of Service attacks. These figures were especially
useful in conjunction with a web page written by Tim Clark (http://www.warwick.ac.uk/T.Clark/jig.html)
which produced graphs for specified date ranges and protocols.
The future of the National
Web Cache Service after next year was discussed. The possibility of caching at
a MAN level rather than nationally was mentioned. This had been discussed in
the East Midlands some years ago but never implemented as the transatlantic
charging scheme would have made the allocation of costs difficult. Perhaps the
demise of charging based on transatlantic usage will mean that it is time to
revisit it. On the other hand, it may be that increased bandwidth within
SuperJANET and across the Atlantic has diminished the need for caching.
It was felt that the rules
and charging rates for FE colleges who wished to procure more than their
allotted 2 Mbps should be written down and publicised by UKERNA.
There was general approval
for the speed and reliability of SuperJANET4 since its introduction earlier in
the year.
Andy Jack
Acting Chair, MIDJUG
There
was no other business.
The
next meeting was provisionally booked for 24th January 2002 at De
Monfort University, Leicester Campus, starting at 2:30pm.
The
topic for the next MNF meeting to take place 10:30am in the same location was
set as Web Traffic Filtering.