You are here: Home > Participate > Join a Discussion > Mailman Archives
<<< Chronological >>> Author Index    Subject Index <<< Threads >>>

Re: RIPE DNS Hostcount October 1992

  • To: (Marten Terpstra)
  • From: Tony Bates < >
  • Date: Tue, 3 Nov 1992 10:06:26 +0000 (GMT)
  • Cc:

Marten Terpstra writes:
>  bob@localhost writes:
>   * Marten,
>   * 
>   * > The UK and DE have been counted by ULCC and DE-NIC,
>   * > and the output collected at the NCC. As you will see, this has increased
>   * > counts for both countries quite significantly, especially the UK.
>   * 
>   * The increase in UK figures you saw may not be due to the method of
>   * collection, but rather a real effect. Our own figures showed a substantial
>   * increase this month -- about 4,200 "real" hosts. (Our accounting of
>   * "real" hosts is always a bit more conservative than yours -- we identified
>   * 40636 this month, rather than the 43301 you saw. This is possibly because
>   * we have a better (local) understanding of what constitutes a duplicate?)
> I don't get it. Tony uses *exactly* the same tools for counting as I do, so
> it is quite impossible that the definition of duplicates is different,
> because the tools decides what is a duplicate and what is not.
> Also, from private mails between Tony and me, his count has always been
> higher than mine, and Tony has been using the same tools right from the
> start. If you are not talking about the figures Tony regularly sends around
> in the UK, what other hostcount measure do you take ?
> I understand that some of the increase is real increase, but the rest is
> increase due to to better connectivity.
Marten and Bob are right. We had a big jump with a lot of delegations this mont
but unfortunately, the count Bob got 
(my monthly auto-script) actually didn't get as much as the local count I 
did for Marten. We lost a couple of zones for some reason that had a lot of 
duplicates. As Marten says I do use the same tools so it is not a local 
understanding of duplicates thing. I hope this clears this up.

> As for the duplicates, our program defines a host with more than one hostname
> for the same IP address as a duplicate host. So for instance;
>      86400   IN      A
>  86400   IN      A
> is counted as one (1) duplicate host, because it is the same machine with
> more than one name. These duplicates are checked inside the domain it is
> counting; in case of the hostcount this is the top-level domain. Also of
> course, hosts with more than one IP address are only counted once, but these
> do not show up in the duplicates.
>   * We still have a considerable number of JIPS sites that have not yet
>   * delegated so I guess we'll carry on growing for a while yet...
> I guess so. You should be able to get very near Germany soon ...

This is true. We have a lot of zone still not delegated.
> -Marten

  • Post To The List:
<<< Chronological >>> Author    Subject <<< Threads >>>