bad_newsgroup - xref slave

bill davidsen davidsen at tmr.com
Tue Feb 18 23:06:30 UTC 2003


In article <ylk7g1w8z4.fsf at windlord.stanford.edu>,
Russ Allbery  <rra at stanford.edu> wrote:
| Miquel van Smoorenburg <list-inn-workers at news.cistron.nl> writes:
| > Russ Allbery  <rra at stanford.edu> wrote:
| 
| >> I'm not sure what would be a better way of handling this.
| 
| > Just ignore the non-existing group? Falling over and dying is a
| > bit drastic.
| 
| If the non-existing group is ignored, though, you no longer have a
| complete slave (and the Xref information in the article is also wrong for
| the slave, but that's less of an issue).  As soon as you accept the
| article, you can't ever (easily) recover full slave state; if you throttle
| and the admin goes and fixes the newsgroup list, you then recover full
| slave status as soon as the feed catches up.
| 
| I was toying with the idea of deferring it, in the hope that it's a race
| condition, but if it's not just a race, one can defer the article for
| forever.

There are two reasonable things to do, neither of which is to hang
solid.

1 - if you really want a full slave, then create any newsgroup listed in
the Xref line (NOT the Newsgroups line!) as it's on the master.

2 - if you only want to slave the groups you have, say big8, then simply
ignore any groups you lack.

If we had an "autocreate wildmat" feature we could do either of these.
Possible new feature for some future release.

I can't think of any real world case when I want the slave to just stop.
Like benchmarks, you can come up with an unlikely case, but I can't
think of one I have ever actually used, and I've been running lots of
slaves since 1995.


-- 
bill davidsen <davidsen at tmr.com>
  CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.


More information about the inn-workers mailing list