Thread: fmgr_info ERROR
Postrgresql 7.1.3 We are receiving this error in the log file: ... 2001-10-30 14:22:34 [17465] ERROR: fmgr_info: function 59668709: cache lookup failed ... several times And just above it in the log file is: 2001-10-30 14:22:26 [17355] NOTICE: InvalidateSharedInvalid: cache state reset several times in a row. Are they related? What could be causing this? How do I trace the number given (59668709) to the actual function or is thisnot relevant? Thanks, -- Laurette Cisneros (510) 420-3137 NextBus Information Systems, Inc. www.nextbus.com Passenger Information Everywhere
Laurette Cisneros <laurette@nextbus.com> writes: > We are receiving this error in the log file: > 2001-10-30 14:22:34 [17465] ERROR: fmgr_info: function 59668709: cache lookup failed > And just above it in the log file is: > 2001-10-30 14:22:26 [17355] NOTICE: InvalidateSharedInvalid: cache state reset > several times in a row. > Are they related? What could be causing this? Given that those are coming from different backends, I don't think they're related. The NOTICE can be ignored in any case. > How do I trace the number given (59668709) to the actual function or is this not relevant? select * from pg_proc where oid = 59668709 ... but I bet you'll find there's no matching entry. This seems to indicate that you've dropped a function entry that was still in use somewhere. What queries are drawing the error messages? (If you don't know, turn on query logging to find out.) regards, tom lane
Yes, it seems there was a hiccup (on our part) of dropping the function and recreating it but not dropping and recreatingthe trigger. Thanks for the quick insight. L. On Tue, 30 Oct 2001, Tom Lane wrote: > Laurette Cisneros <laurette@nextbus.com> writes: > > We are receiving this error in the log file: > > 2001-10-30 14:22:34 [17465] ERROR: fmgr_info: function 59668709: cache lookup failed > > And just above it in the log file is: > > 2001-10-30 14:22:26 [17355] NOTICE: InvalidateSharedInvalid: cache state reset > > several times in a row. > > > Are they related? What could be causing this? > > Given that those are coming from different backends, I don't think > they're related. The NOTICE can be ignored in any case. > > > How do I trace the number given (59668709) to the actual function or is this not relevant? > > select * from pg_proc where oid = 59668709 ... but I bet you'll find > there's no matching entry. This seems to indicate that you've dropped > a function entry that was still in use somewhere. > > What queries are drawing the error messages? (If you don't know, turn > on query logging to find out.) > > regards, tom lane > -- Laurette Cisneros (510) 420-3137 NextBus Information Systems, Inc. www.nextbus.com Passenger Information Everywhere