High recursive client counts

[Mark Andrews]

In message <CAHk3O4q96L8Cc4CK=058ALe4wH_NdqVNd3qEUczpO8MkoUYm7w at mail.gmail.com>, Scott Bertilson writes:
> 
> This got me to take a look at "rndc recursing" on one of our servers.
> 
> It is disappointing that queries for the same FQDN/type/class from the same
> client (different source port and query ID though) are handled individually
> rather than being merged somehow.  Is this because of the ID or the source
> port, both, or something else?

What makes you think they are not merged?  Named detects duplicates.
It has self tuning limits on the number of simultaneous clients for
the FQDN/type/class tuple dropping clients if there are too many.
It also drops duplicates where the source port and address are
duplicated.

Named still has to reply to all the clients which is why they are
on the recursing list.

Mark
-- 
Mark Andrews, ISC
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: marka at isc.org