portsnap can't access portsnap[124].freebsd.org

[Scott Bennett]

     On Sat, 30 May 2009 21:05:44 +1000 John Marshall
<john.marshall at riverwillow.com.au> wrote:
>On Sat, 30 May 2009, 05:39 -0500, Scott Bennett wrote:
>>      Thanks for that much, then, Glen.  So we still don't know what is wr=
>ong.
>> I did try it again around midnight CDT, and it still failed the same way.
>>      In the meantime, my frustration and impatience got the better of me,=
> and
>> I touched up a copy of ports-supfile and csupped it.  I also used portins=
>tall,
>> which was there because I had selected it as a package during the 7.2-REL=
>EASE
>> installation, to install portmaster, which is now running the builds.
>>      Nevertheless, I'd still rather switch back to portsnap ASAP after th=
>is,
>> so I'm still hoping someone has an idea what's wrong with portsnap or the
>> systems at freebsd.org.
>
>Have you excluded local factors (proxy servers, firewalls)?  I have not
>seen any issues at all with portsnap.  I have done a few fetches today
>and haven't seen any problems at all.  This one a few minutes ago
>happened to hit portsnap2.  I noticed that one of the earlier ones today
>was from portsnap1.
>
  [portsnap session omitted  --SB]
     There was no proxy.  However, I think you may have hit upon the problem.
It depends upon what TCP port number portsnap uses.  If it connects to the
HTTP port, i.e., port 80, then I know what happened.  While running portmaster,
I soon had to deal with a problem where all of the fetches for a package
failed, but immediately, not after lengthy waits for timeouts.  It quickly
dawned on me that the http_proxy environment variable had been set to connect
things like fetch(1) and wget(1) to privoxy, which I haven't reinstalled yet
since installing 7.2-RELEASE.  This gets set in /root/.cshrc.extensions, a file
that I source from /root/.cshrc to keep most of my changes separate from stuff
that could get replaced accidentally in a mergemaster run between a buildworld
and an installworld or, as in this case, a full new OS installation.  I
unsetenved that, and various ports' Makefile fetch targets were happy again.
I never knew what port portsnap used, but maybe that's it.
     Once portmaster finishes rebuilding the relatively small set of already
installed packages and ports, I'll give portsnap another shot at it to see
what happens.
     Thanks much for the idea that connected the two problems for me!


                                  Scott Bennett, Comm. ASMELG, CFIAG
**********************************************************************
* Internet:       bennett at cs.niu.edu                              *
*--------------------------------------------------------------------*
* "A well regulated and disciplined militia, is at all times a good  *
* objection to the introduction of that bane of all free governments *
* -- a standing army."                                               *
*    -- Gov. John Hancock, New York Journal, 28 January 1790         *
**********************************************************************