William, william@elan.net wrote:
Hi,
I'd like to find what is the best way to determine if particular inetnum was allocated directly by RIPE, this would include all blocks allocated to LIRs, all blocks allocated to large organizations and blocks allocated as small PI blocks.
From reminder the netblocks totally within another inetnum are discarded as well as "superblocks" listed under RIPE name. I'd like to find if you
Currently I have come with the following way to do it - first find all inetnum objects listed as "ALLOCATED PI" with network name starting as "EU-ZZ", this find all PI space; then discard all "ASSIGNED PA" blocks except those within range of previously found "ALLOCATED PI" blocks and also discard "ALLOCATED PI" and "LIR-PARTITIONED PA" blocks. think doing it this way would leave any blocks out or include too many larger block then direct RIPE allocations.
Doing it the way I described above, already produced some data available at http://www.completewhois.com/bogons/data/allocated-cidr-ripe_main.txt and non-allocated space is at http://www.completewhois.com/bogons/data/bogons-cidr-ripe_main.txt I'd like to verify validity of this data and the process I used, especially considering that I see number of errors, for example there are several blocks > /24 listed as allocated (but I did not think RIPE did any direct allocations or assignemsns of size > /24), for example: 193.58.0.64/26 193.164.232.192/26 193.201.147.0/26 193.201.148.64/26 193.243.183.0/26 193.254.0.64/26 194.1.127.128/26 194.9.166.0/26 194.149.71.128/26 194.153.156.0/26 194.180.159.128/26 194.180.226.192/26 194.246.39.0/26 194.246.39.128/26 195.35.104.64/26
Please have a look at: ftp://ftp.ripe.net/ripe/stats/issued The README explains the contents, and you can review the source code there if you want slightly different versions of this data. You can also look at: ftp://ftp.ripe.net/ripe/stats/ The two are slightly different views on the same data. The first comes from the Whois Database itself, and the second is produced from our internal registry database. There are some records in the issued files that are not in the stats files. We are in the planning phase of a project to account for this and other differences, including inter-RIR conflicts, based on the excellent work that Geoff Huston has been doing on projecting the exhaustion of IPv4. If you intend to filter based on this data, I recommend using the issued files, because it contains a superset, and you are less likely to block legitimate traffic. -- Shane Kerr RIPE NCC