LISTSERV  checks how  much  storage is  *actually  available* just  after
booting. For  backbone sites, it wants  to have 2.5M available,  based on
the assumption that up to about 0.5M  can be later blocked by data in the
various storage  caches LISTSERV maintains  and that 2M are  necessary to
process large  DISTRIBUTE jobs. What  happens is  that you must  have had
2.49M  available when  the message  was  issued and  2.51M on  subsequent
restarts, possibly due to  FST's of R/O disks. There is  no need to worry
about LISTSERV not having enough storage, however I would personally be a
little bit worried regarding what happened  to the other 1.5M (about 300k
are used by EXECLOAD  and maybe 150k by PREXX, but that's  still a lot of
memory not accounted for). On the other  hand if you have large R/O disks
with nonshared FST's, this can be normal.
 
  Eric