CNAME performance hits, named statistics recording

[Barry Margolin]

In article <4.3.2.7.2.20000601104110.00d665a0 at 216.67.12.69>,
Forrest Aldrich  <forrie at forrie.com> wrote:
>What performance hits, if any, are there to utilizing CNAME records on big 
>production machines.  In particular, would this eventually be reflected in 
>evaluation figures provided by third parties, like Inverse.

The performance impact should be minimal, assuming you have reasonable TTLs
on all the records involved.  After the first lookup, both the CNAME and
the target will be cached on the local DNS server, and then there's minimal
overhead (probably less than a millisecond per lookup) for the
indirection.  If the CNAME records are between domains hosted on the same
authoritative DNS servers, there also won't be any more overhead during the
initial lookup, since it will return both of them in one response and
they'll both be cached.

-- 
Barry Margolin, barmar at genuity.net
Genuity, Burlington, MA
*** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups.
Please DON'T copy followups to me -- I'll assume it wasn't posted to the group.