Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> pooja_khobragade(at)persistent(dot)co(dot)in writes:
>> Error while connecting to server FATAL: shmemindex size is wrong
>> for datastructure xlog ctl : expected xxx actaul xxx
>
> That's an interesting one. The only theory that comes to mind is
> that you've got multiple versions of Postgres installed on your
> machine, and when the postmaster tries to launch a subprocess, the
> wrong copy of postgres.exe is being selected for some reason. Or
> possibly you updated in place and didn't restart the postmaster?
Just so the OP is aware, running pg_ctl with the restart option is
generally not adequate to prevent problems with an update, in place
or not. The postmaster must be stopped and then started. Some
service scripts may do that, others don't. So the safe thing,
unless you know exactly how your service script is written, is to
stop and then start again.
-Kevin