Sean Donelan SEAN at SDG.DRA.COM
Fri May 21 20:13:46 UTC 1999

alan at globalcenter.NET (Alan Hannan) writes:
>  However, for those who have subscribed to nanog for more than 3
>  years, nanog has traditionally been useful to discuss operational
>  issues, where issue is defined as a concept or problem.  
>  Real time issues are generally helped little by nanog discussions.

I agree 100%.  Unfortunately, unlike Alan, I've found nanog discussions
are very good at fixing 'real time issues' and not so good at fixing
long-term operational issues.  

Carriers should be the ones who first notify their customers about problems.
In theory, there should be no way I could know and post about a problem
before a carrier monitoring its own facilities.  Likewise, reporting a
routing loop problem to a carrier should result in its repair without the
need to post about it on a public mailing list.  After all, it affects
them whether reported by a customer or a non-customer.

Most of the reasons why I post problems have their root-causes in some
very old operational problems.  Heck, the routing loop Alan mentioned seems
to have been aggravated by a Cisco IOS bug (or at least similar to a bug)
dating back to the days when Alan was still a network tech in Nebraska.

How about this as a long-term operational issue, defined as a concept
or problem: why are we still discussing the same issues three years
later?  What makes them so difficult to fix?
Sean Donelan, Data Research Associates, Inc, St. Louis, MO
  Affiliation given for identification not representation

More information about the NANOG mailing list