Re: pg_upgrade using appname to lock out other users

On Wed, Jun 15, 2011 at 09:14:16PM -0400, Stephen Frost wrote:
> Bruce,
>
> * Bruce Momjian (bruce(at)momjian(dot)us) wrote:
> > I have researched this and need feedback.
>
> In general, I like the whole idea of using random/special ports for the
> duration of the upgrade. I agree that we need to keep the ability to
> check the existing clusters. My gut feeling is this: keep the existing
> port options just as they are, so --check works just fine, etc. Use
> *only* long-options for the "ports to use during the actual upgrade" and
> discourage their use- we want people to let a random couple of ports be
> used during the upgrade to minimize the risk of someone connecting to
> one of the systems. Obvioulsy, there may be special cases where that's
> not an option, but I don't think we need to make it easy nor do I think
> we need to have a short option for it.

As an operations guy, the idea of an upgrade using a random,
non-repeatable port selection gives me the hebejeebees. Mr. Murphy will
com knocking, sooner or later, with the server picking a port that just
happens to be available right now, because it's service is restarting,
or is under inet control.

Ross
--
Ross Reedstrom, Ph.D. reedstrm(at)rice(dot)edu
Systems Engineer & Admin, Research Scientist phone: 713-348-6166
Connexions http://cnx.org fax: 713-348-3665
Rice University MS-375, Houston, TX 77005
GPG Key fingerprint = F023 82C8 9B0E 2CC6 0D8E F888 D3AE 810E 88F0 BEDE