Hello All,

I posted this query some time ago, then got swamped with other problems. I
have not yet found a satisfactory explanation for this occurrence so here
goes.
Back in May this year, I upgraded our listserv backbone machine to use
Listserv 1.8d and LSMTP 1.1b. After doing this we noticed that the stats on
delivery had dropped down to roughly one third of what they had been before
the upgrade. Admittedly some modifications  had to be made to the accounting
script since the format of the accounts file had changes, however when
tested with a small sample file everything appeared to be accounted for
properly.
Now having run the CTR query to the listserv machine for the months of
April, May and June I have found that the only significant change seems to
be in the CPU usage column. In April the CPU usage was 0 microseconds. In
May it had gone up to 4.5e6 micro-seconds and in June was up to 22e6
micro-seconds (The May rise was lower than June since the upgrade took place
on the 25th of May). The CPU usage has remained at 22e6 micro-seconds since.

Can anybody out there explain this behaviour, I am stumped. Am I missing
something obvious. According to the number of recipients in the outbound CTR
stats the figure is relatively stable. The bandwidth usage register #2 has
dropped markedly from 70554 in February down to only 9563 in July but it
picked up again to 38867 in August, does this info. help at all ?

Thanks for any help offered.

Yours ever hopefully
Duncan Russell

----------------------------------------------------------------------------
---------
Mr Duncan Russell
Information Technology Department
RAL Computing Services Support Group            Phone (44) (0) 1235 445542
Atlas Centre  RAL               Fax   (44) (0) 1235 446626
Chilton DIDCOT Oxon             E-mail [log in to unmask]
OX11 0QX    UK

RAL Computer Services Support Group WWW Page:
http://www.itd.clrc.ac.uk/Group/ITDSERRCS
----------------------------------------------------------------------------
---------