>For various reasons  not worth going into here,  the RICEVM1<->UHOU link
>is a real bottleneck.
 
No problem, LISTSERV is designed to be able to do that. And with EARN I'm
used to cases where there isn't the  slightest hint of the beginning of a
possibility of a technical reason :-), so no problem here.
 
>1. Does changing my local copy of LINKSWT affect only distributions that
>   start at RICEVM1?
 
Right, that  was the only  way to cleanly solve  the problem of  server X
thinking that  server Y  is closer while  Y thinks X  is closer.  It also
greatly minimizes the  impact on the network of a  particular node having
an outdated  file (for any reason,  like their node no  longer recognizes
Netdata files because of that infamous HPO5 bug).
 
The proper method when you have a problem  like that is to send me a note
explaining what's going on and what you would like to see changed. I then
send  a global  change to  LINKSWT FILE,  either immediately  if it  is a
serious problem, or after a couple days/weeks for less important changes.
Putting a  high weight on the  guilty link is usually  enough, unless you
have a ring  (like A <--> B  <--> C <--> A  - if you weight  A-B, you may
also need  to weight A-C). Please  let me know  if this is the  case, I'm
already adding an entry for RICEVM1-UHOU  and will send the update when I
hear from you.
 
  Eric