>> According to Ahmon Dancy: >> > Has there been any discussion on caching the results of gethostbyname >> > (in Connection.cc) so that a nameserver isn't continuously hammered w/ >> > the same lookup over and over? >> >> I don't recall it being discussed before. Of course, if you set up a >> caching-only name server right on the machine that runs htdig, the >> nameserver requests would likely become a non-issue. That's probably >> a cleaner solution than putting caching right into htdig, although it >> might be easy enough to add an IP address field to the Server class and >> have htdig use that for connections.
I think the latter approach should be considered. It's unfortunate to recommend that folks do extra systems admin stuff to make htdig work more efficiently. In my case, 172 bytes of unnecessary network data transfer occurs due to DNS lookups. We're indexing 58,100 pages. That comes out to 9.5MB of useless data transfer that could be eliminated (seemingly) easily. _______________________________________________ htdig-general mailing list <[EMAIL PROTECTED]> To unsubscribe, send a message to <[EMAIL PROTECTED]> with a subject of unsubscribe FAQ: http://htdig.sourceforge.net/FAQ.html