FYI ONLY >> Non-Operational (at least directly)

Robert Mathews-ICICX mathews at hawaii.edu
Tue Apr 13 18:42:49 UTC 1999



On Tue, 13 Apr 1999, Sean Donelan wrote:

> Date: Tue, 13 Apr 1999 02:53:04 -1000
> From: Sean Donelan <SEAN at SDG.DRA.COM>
> To: nanog at merit.edu
> Subject: Re: FYI ONLY >> Non-Operational (at least directly)
> 
> randy at psg.COM (Randy Bush) writes:
> >> University of Hawaii' external data services were effected (Primary and
> >> Secondary DNS were not functional).
> >
> >hmmmmm.  wonder how that could hapen.
>
>				[ ... ] 
>
> It appears Hawaii.EDU at least has arranged for an off-island backup of
> its DNS.  There may have been some other configuration problem, but at
> first glance Hawaii.EDU and Hawaii.NET had a chance of keeping its
> DNS functioning.  Hawaii.GOV's NS records have a slight problem, which
> may have killed it when connectivity was lost.  And things like HI.US would
> have been completely dead.  The US-domain sucks in many states, so that's
> not too unusual.  DRA made a standing offer to the US-domain administrator
> to act as one of the primary or secondary servers for any LIB.xx.US.

For those wish to know.. The University of Hawaii (UH) has had a
long-standing relationship with NASA, providing the University with
Secondary DNS since Hawaii's connectivity to the net.  That arrangement
is not new. The physical dispersion of UH campuses and community colleges
-- has over the years complicated network topology (I am in no way
responsible for it, NOR do I answer to complications within it...:) ).
Correspondingly, DNS as a function, does experience difficulty
with respect to stability -- within the UH system.

I am not going to go there...  when it comes to HI.US
 
> AT&T may handle the physical circuits, but the Internet service for the
> University of Hawaii is handled by UUNET.  So there are lots of different
> people you can ask why the redundancy failed, or if it was ever present.

I did not want to be that complete Sean..  :)  yes, however, if the bits
are experiencing a problem crossing a bridge -- no matter where the bridge
is (on land.. over a stream and between islands), what you ultimately end
up with is 'byte-jam' or as they say here in Hawaii, 'byte-jam-up'.

It does not ease the situation WHEN your bits transit one provider's
backbone (whom you chose to assure redundancy) does ultimately receive
transit (at times such transit is not for long distances..) from a yet
another whose infrastructure is now effected. I am not suggesting however,
that is the case here..  AT&T is investigating.  :)

I am going to jump out of this conversation all, and get back to work.
Thought that it would be interesting to report this to all who were
interested.

Best.





More information about the NANOG mailing list