Configuring different TTLs in multiple RRs for the same domain name, TYPE, and CLASS
carl at byington.org
Wed Mar 30 20:29:25 UTC 2016
-----BEGIN PGP SIGNED MESSAGE-----
On Fri, 2016-03-25 at 22:15 -0400, Barry Margolin wrote:
> If you're running a resolver for a small organization, the cache isn't
> going to get huge in the first place. How many different names will 50
> users access in a day?
Looking at 6 such small resolvers for organizations ranging from 20 to
200 users, the memory usage of bind/named is between 200MB and 900MB.
On the system with the largest memory usage:
rndc dumpdb -all
egrep -v '^( |'$'\t''|;)' named_dump.db | wc -l
shows 83K lines, many of them for an internal rpz zone. Removing that,
we have 25K lines left. Removing duplicate names (where we have multiple
record types), we have 21K names. Note that those may have accumulated
over much more than a day.
The longest TTLs are around 600K seconds, mostly from names in
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.14 (GNU/Linux)
-----END PGP SIGNATURE-----
More information about the bind-users