Hello!
I've recently begun seeing an unusually high number of error messages
from my database, of this type:
NOTICE: InvalidateSharedInvalid: cache state reset
I noticed in the archives that people have asked about this issue before,
but my case seems a bit different, in that my errors do not seem related
to any database vacuuming. Tom Lane noted in one of his archived replies
that these messages can sometimes be caused by applications with very long
database access times. I'm running a series of cron'ed Perl tools, each
of which opens a connection to Postgres for up to 5 minutes. Could this
be the problem? More generally: what is a cache state reset, and what
might trigger one?
Thank you very much in advance.
Sincerely,
David Lobron