[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Yesterday's latencies

Hi Harry, Art, and Daryl,

I spent some time this morning trying to explain the high latencies that
you saw yesterday afternoon.  At that time, traceroutes did show
relatively slow paths along Abilene, but they were not horrendous.

This morning I checked some Abilene performance pages, looking for
network explanations.  Although I found "hills" (as opposed to "peaks")
during that time period, I didn't see anything that would clearly
explain those latencies.  

The pages that I used the most were: 

        http://hydra.uits.iu.edu/~abilene/traffic/  
                (Art's page.  Click on the nodes to get more info)
        http://monon.uits.iupui.edu/abilene
                (Click on nodes to get more info)
        http://waco.uits.iu.edu/abilene/today.cgi
                (to get info about the UCAR subnet)

Daryl reported actually seeing latencies on motherlode, but I did not
see that nor could I find evidence of it in the log.

At this point, I will simply wait to see if this occurs again.  If you
see evidence of this please let me know.  

And, if you ever see motherlode running particularly slowly I'd be
interested in knowing that as well.  You all have machines that are
'allow'ed on motherlode, so you can run 'notifyme' from those hosts and
compare the current time with the product ingest time.  (You all
probably already know this already.) 

Thanks!

Anne
-- 
***************************************************
Anne Wilson                     UCAR Unidata Program            
address@hidden                 P.O. Box 3000
                                  Boulder, CO  80307
----------------------------------------------------
Unidata WWW server       http://www.unidata.ucar.edu/
****************************************************