Re: Horology & stats tests fail in 7.4.3/RH 9

On Mon, Jul 19, 2004 at 12:33:40AM -0400, Tom Lane wrote:
> Chuck Bearden <cbearden@hal-pc.org> writes:
> > I am building from a freshly downloaded tar.gz of 7.4.3.  The machine is
> > a 4CPU Dell with an unsupported RH 9 install.
>
> > When I ran the regression tests, two failed: horology and stats.  I'll
> > append the content of the regression.diffs file to the posting.
>
> > Is this something I should be worried about?
>
> Your horology results appear to match the no-DST-before-1970 variant,
> which would be unsurprising on some variants of Unix but it is a tad
> surprising on a recent Linux platform.  Still, unless you are planning
> to do stuff with pre-1970 timestamps it's not worth getting excited
> about.

It doesn't sound like something I need to worry about for this project.

> The stats failure shows that the statistics collector is not starting.
> The only cause of that I've seen in the past is broken or overly
> restrictive configuration of local-loopback IP connections.  Look in the
> postmaster's log for more details.

That makes sense.  I need to review the iptables rules--they weren't my
creation, and I'm not sure the guy who set them up opened up traffic on
lo completely.  As I said, I don't have root unless I'm practically
sitting in the admin's lap.

Thanks very much for your responses.  They were quite helpful.

Chuck