emv@mailrus.cc.umich.edu (Edward Vielmetti) (06/17/88)
There's a piece of mail on a.cc.umich.edu (35.1.33.90) destined for
euler.berkeley.edu that's been sitting around for a
few days. If mail can't go from Michigan to Berkeley in
a day, it's usually an indicator that something's quite wrong,
so I decided to snoop around some.
I called the Merit NOC and found this out:
It's possible to FTP from the 35.1.1 network to 128.32.142.1
just fine, with no unusual delays. The traffic is routed through
the proteon link to Cleveland (the one that was a source of problems
earlier this week).
It's impossible to FTP from the 35.1.33 network to 128.32.142.1,
where 35.1.33 is connected to 35.1.1 with two proteons and a T1
in the middle of them.
35.1.1 - proteon - T1 - proteon - 35.1.33
As far as my limited ability to check into routing tables goes,
everything looks OK. The only guess that I have at this point
is that the packets take too many hops and are dropped somewhere.
I can get to 128.32.140 and 128.32.137 just fine. Not knowing
what the particular network looks like at Berkeley, it's hard
to give any more information than this.
Even if there are mailer problems at euler (could very well be) it's
still an odd problem.
--Ed
Edward Vielmetti, U of Michigan mail group