Re: BUG #5660: Can't start db service if specify effective_io_concurrency

tkbysh2000@yahoo.co.jp wrote:
> Hi,
> 
> Thank you very much for your quick support.
> I've understood. Please close this bug ticket.
> 
> I hope add comment e.g"(unix only)" onto effective_io_concurrency line
> in postgresql.conf if possible.

We currently have in our docs:

         Asynchronous I/O depends on an effective posix_fadvise</>
         function, which some operating systems lack.  If the function is not
         present then setting this parameter to anything but zero will result
         in an error.  On some operating systems (e.g., Solaris), the function
         is present but does not actually do anything.

It is not Unix-only, but only some versions of Unix.

---------------------------------------------------------------------------


> 
> Regards.
> 
> -- 
>  
> 
> 
> On Thu, 16 Sep 2010 10:36:19 -0400
> Tom Lane  wrote:
> 
> > "Mikio"  writes:
> > > I'm using postgresql 9.0 rc1, and I specified 10 for
> > > effective_io_concurrency in postgresql.conf.
> > > I restarted postgresql windows service, but the service didn't start.
> > 
> > This is unsurprising: you can only set effective_io_concurrency on
> > platforms that have posix_fadvise(), which I rather doubt Windows does.
> > 
> > > I looked windows event viewer, I found an event from PostgreSQL, and it was
> > > below.(The characters in the message were broken.)
> > 
> > >> FATAL:  p[^"effective_io_concurrency"?????> 
> > Hm, what I get when I try that on a machine without posix_fadvise()
> > is
> > 
> > FATAL:  parameter "effective_io_concurrency" cannot be changed
> > 
> > I think you have some other configuration problem that's messing up
> > your log entries.
> > 
> > 			regards, tom lane
> 
> 
> 
> -- 
> Sent via pgsql-bugs mailing list (pgsql-bugs@postgresql.org)
> To make changes to your subscription:
> http://www.postgresql.org/mailpref/pgsql-bugs

-- 
  Bruce Momjian          http://momjian.us
  EnterpriseDB                             http://enterprisedb.com

  + It's impossible for everything to be true. +