@device[x2700]
@make[report]
@flushleft[@B{EDINBURGH UNIVERSITY COMPUTING SERVICE}]
@flushleft[For the Attention of the E.C.U.C.]
@majorheading[Report on Communications]
@section[General]
@section[CPSE]
For the past 3 months
@begin[verbatim]
                   S/W       H/W                   S/W & H/W        Total
                 Crashes   Crashes  Environmental    MTBF            MTBF

KB  GEC 4190        1         1           1         46 days        46 days

AT  GEC 4190        0         2           2         46 days        27 days

AT  GEC 4160        0         1           1         92 days        46 days
@end[verbatim]
Although the GEC switches are now more reliable than they used to be there is
still an unfortunate number of incidents on these switches.

Over the past month the KB switch has been taken out of service at lunchtime 5
times and the AT switch 4 times (twice for H/W).  This was done to find and identify a fault which
could bring down the main trunk between those two switches.  This fault has now
been fixed and lunchtime slots will only be taken for emergencies during the
term.
@Subsection[TelePAC Switches]
For the past 3 months
@begin[verbatim]
                   S/W       H/W                    S/W & H/W        Total
                 Crashes   Crashes  Environmental     MTBF            MTBF

TelePAC  1 (KB)    0         0            0          > 92 days    > 92 days

         2 (KB)    0         0            0          >127 days    > 92 days

         3 (Bush)  0         0            1          >127 days      92 days
@end[verbatim]
There have been no hardware or software problems with the TelePAC switches.
@section[PADs]
The MTBF of the PADs has continued to improve, the overall figure for the past
3 months is 35 days between incident which is a welcome continuation of the
short term trend reported at the last meeting.

The main causes of PAD crashes are now either from mains failure (i.e. after a
lightening strike in the area) or from problems with devices being switched on
or off in close proximity to the PADs.  A cure for the latter problem is being
urgently sought to enable us to further improve the reliability.

The Netcomm Super PAD has now been approved for use within the Academic
Community.  We have been running one on a trial basis for the past 6 months and
now intend to purchase a number with the aim of replacing the most heavily
loaded Camtec PADs.
@section[FEPs]
The MTBF of the EMAS-A Feps for the past 3 months in 26 days.  This corresponds
to 7 crashes on FE1 and no crashes on FE2.  The main cause of these crashes
hace been FTP faults most of which have been fixed.
@section[INFO]
Heavier use of INFO has shown up a number of faults in the Torch X.25 software
which caused line failures when a large number of users logged on, for example,
just after a crash on EMAS-A.  This situation has now been stabilised and we
are pursuing Torch for a more reliable version.
@Section[XCALL]
This machine has been switched off as there is no longer any need for
conversion between ITP and XXX.
@section[PSSE]
Service has been satisfactory.
@flushright[Brian Gilmore
11th September 1987.]