Delay alarms.

Dear test-box hosts,

Last week, I wrote:

> As announced during the last RIPE meeting, we have installed a network
> delay alarm on some of the test-boxes.  [...]

> We plan to keep this setup running for a week or so, in order to make
> sure that the number of alarms generated is indeed small and related to
> real network events. If this is successful, the alarm will be installed
> on all test-boxes.

As you have probably noticed by now, some of the test-boxes have started
to generate alarms when delays from/to the box are above the expected
values.  No big or unexpected problems were found in the code, so we just
installed the code on all boxes.

We are logging all alarms generated by the boxes and are continiously
checking if they can be traced back to a "real" effects in the data.  
(For example: a change in the routing vector, number of hops, etc, as
opposed to a random statistical fluke.)  So-far, this appears to be the
case.  A more detailed study will be done as soon as we have collected
more statistics.

Also note that there are some startup effects.  The code has to figure out
what the expected delay for a connection actually is. During this period,
it might generate a false alarm or two.

Please do not hesitate to contact me if you have any further questions,

Henk

------------------------------------------------------------------------------
Henk Uijterwaal                    Email: henk.uijterwaal at ripe.net
RIPE Network Coordination Centre     WWW: http://www.ripe.net/home/henk
Singel 258                         Phone: +31.20.535-4414,  Fax -4445
1016 AB Amsterdam                   Home: +31.20.4195305
The Netherlands                   Mobile: +31.6.55861746  
------------------------------------------------------------------------------

The Committee (...) was unable to reach a consensus that substantial merit was
lacking. Thus, the appeal was deemed meritorious.          (Orlando NABC #19).