The table shows every prefix that originated from the target AS in the last week.
The table can be sorted on any column by clicking on the headers. The size
is provided as a separate column to ease sorting.
As the current instance of the RIS database started around April 2009, any prefix
with a first seen time before 2009-04-01 may be a lot older than 2009. However,
the current RIS database can not see that.
Note that this also means that only prefixes seen after April 2009 are included
in the graph.
The registry listed is based on IANA's
IPv4 Address Space
document.
The prefix table changes background color for IANA reserved and RFC1918 space, and unusually
large or small prefixes. An IPv4 prefix is unusually small when smaller than a /24 and
unusually large when bigger than a /7. An IPv6 prefix is unusually small when smaller than /48 and
unusually large when larger than /10.
The number of peers is the number of RIS peers sending this prefix now. This can be used as an
indication for the spread of the prefix.
Explanation
This list and diagram show the 'transits' seen for the target AS.
Whether or not these actually have a business relation to the target AS can not be
determined from routing data. Therefore, the software simply looks at all peerings
seen for this AS in the last three months and counts all ASes to the left of the target AS in an AS path.
The software ignores any peerings seen from a direct RIS peer. So if RIS would see
'123 456 789 666', the 123-456 peering would not be counted, the 456-789 and 789-666 peering would.
Therefore, all ASes listed here are peers that pass the routes sent by the target AS
on to at least some other ASes.
ASes with less than 5% are counted under 'other'.
The ratio between the different 'transits' is based on the number of routes seen.
Therefore, this is not related to the amount of traffic and changes in the RIS
peerings may cause changes in the ratios.
Explanation
The diagram shows the average length of all AS paths seen in the last week originating
from the target AS.
A second diagram can be shown where prepending is stripped from AS paths before calculating.
Note that this is not necessarily related to geographical distance - if the collector
in Japan (DIX-IE) would have only one peer and that would happen to be a transit of the
target AS, the average distance for DIX-IE would be shown as two (originating AS and its transit).
The absolute minimal distance is one, but most paths will be two ASes at least (without
prepending).
When comparing plots, you may sometimes see that although the AS path gets longer, the line
get shorter. This is because the line length is calculated on a ratio between the AS paths,
not the absolute numbers.
Explanation
