On Fri, Feb 06, 2009 at 04:03:29PM +1100, Sarton O'Brien wrote: > Sarton O'Brien wrote: >> Is this one of those recompile scenarios or has something actually been >> lowered and is preventing pgsql from starting? I would guess two separate issues.. the SHM one may be to do with enforcing rlimits for mapping anons, of which there was some discussion recently. > After recompiling postgresql I was presented with 'invalid checksum in > control file', which is usually associated with moving to a different > arch. This may be the time_t size change, if it wrote a timestamp in the control file. > I'm guessing there was probably some libc changes or something equally > fundamental. I must have missed the heads up. > > All is well now after recompiling, restoring from a dump and modifying > the conf. Bummer, but good to know, I'll be prepared accordingly for my pg instances.
Attachment:
pgpXnhKl92U4n.pgp
Description: PGP signature