At about 3pm Thrusday afternoon, John Hammond of UConn and Jim
Owen of Yale reported to me some problems they
were experiencing with files originating at Marist.  Since LISTSERV
seemed to be a large portion of the problem files, I took LISTSERV
offline at about that time.  It would accept commands, but not
distribute files.  Approximately 2 hours later we were able to
identify the problem and correct it.  We have identified the
problem as having been a human error in attempting to deal with
network problems.  We have identified the problem as having
started at 11:46am EST on 29MAR94 and ending on 15:52 EDT on
07APR94.  The best pattern I can use describe the problem is
random.  Files that were queued  at 15 minute intervals were at
risk.  We have been able to identify that 3488 files were lost,
but we are unable to identify what they contained.  We know some
of them were routing tables, and we regenerated them, and they have
all been reshipped.  Others would be mail for lists with peers
at Marist.  Below is a table showing the number of files lost
on each day.  We apologize for this inconveinance, and have
taken steps to prevent it from happening again in the future.
If there are any questions, please feel free to direct them to me.
 
/ahw
 
 
                                   Cumulative  Cumulative
        DATE   Frequency   Percent   Frequency    Percent
     ----------------------------------------------------
     940329        178       5.1         178        5.1
     940330         22       0.6         200        5.7
     940331         70       2.0         270        7.7
     940401        391      11.2         661       19.0
     940402         22       0.6         683       19.6
     940403          2       0.1         685       19.6
     940404         45       1.3         730       20.9
     940405        257       7.4         987       28.3
     940406       1049      30.1        2036       58.4
     940407       1452      41.6        3488      100.0