partial/no transfer of RR to BIND9 secondary
pmacias at lucent.com
pmacias at lucent.com
Mon Jun 11 20:02:34 UTC 2001
All,
I have a peculiar problem that I cannot yet find solution to by
experimentation or from info in the archives.
On a Linux 2.2.13 box I can run BIND8 with this configuration and
correctly load RRs from "corp-1.com", "corp-2.com" and "corp-3.com":
-------------------------------------------
options {
directory "/etc/domain/";
forward first;
forwarders {
199.118.54.11;
199.118.35.146;
199.118.34.237;
};
};
zone "0.0.127.IN-ADDR.ARPA" {
type master;
file "named.local";
};
zone "." {
type hint;
file "root.cache";
};
zone "corp-1.com" {
type slave;
file "db.corp-1";
masters {
199.118.54.11;
199.118.35.146;
199.118.34.237;
};
};
zone "corp-3.com" {
type slave;
file "db.corp-3";
masters {
199.118.54.11;
199.118.35.146;
199.118.34.237;
};
};
zone "corp-2.com" {
type slave;
file "db.corp-2.com";
masters {
135.140.109.254;
135.109.16.4;
};
};
---------------------------------------------------
On the same box when I run BIND 9.1.2 and use the same named.conf as
above I receive all records from "corp-2.com, partial records from
"corp-3.com" and no records from "corp-1.com.
named.run shows:
Jun 11 15:23:00.960 starting BIND 9.1.2 -d1
Jun 11 15:23:00.962 using 1 CPU
Jun 11 15:23:00.968 loading configuration from '/etc/named.conf'
Jun 11 15:23:00.984 the default for the 'auth-nxdomain' option is now 'no'
Jun 11 15:23:00.985 no IPv6 interfaces found
Jun 11 15:23:00.986 listening on IPv4 interface lo, 127.0.0.1#53
Jun 11 15:23:00.988 listening on IPv4 interface eth0, 135.140.109.156#53
Jun 11 15:23:00.992 now using logging configuration from config file
Jun 11 15:23:00.993 load_configuration: success
Jun 11 15:23:00.993 dns_zone_load: zone 0.0.127.IN-ADDR.ARPA/IN: start
Jun 11 15:23:00.993 dns_master_load: named.local:8: using RFC 1035 TTL semantics
Jun 11 15:23:00.994 dns_zone_load: zone 0.0.127.IN-ADDR.ARPA/IN: loaded
Jun 11 15:23:00.995 dns_zone_load: zone 0.0.127.IN-ADDR.ARPA/IN: dns_journal_rollforward: no journal
...
Jun 11 15:23:12.373 refresh_callback: zone corp-1.com/IN: Serial: new 1999073060, old 0
Jun 11 15:23:12.839 transfer of 'corp-1.com' from 199.118.54.11#53: receiving responses: multiple RRs of singleton type
Jun 11 15:23:12.841 transfer of 'corp-1.com' from 199.118.54.11#53: end of transfer
/var/log/messages gives:
Jun 11 15:23:00 quail named[21095]: starting BIND 9.1.2
Jun 11 15:23:00 quail named[21095]: using 1 CPU
Jun 11 15:23:00 quail named[21097]: loading configuration from '/etc/named.conf'
Jun 11 15:23:00 quail named[21097]: no IPv6 interfaces found
Jun 11 15:23:00 quail named[21097]: listening on IPv4 interface lo, 127.0.0.1#53
Jun 11 15:23:00 quail named[21097]: listening on IPv4 interface eth0, 135.140.109.156#53
Jun 11 15:23:00 quail named[21097]: running
Jun 11 15:23:12 quail named[21097]: transfer of 'corp-1.com' from 199.118.54.11#53: end of transfer
Jun 11 15:24:01 quail named[21097]: transfer of 'corp-3.com' from 199.118.54.11#53: end of transfer
--------------------------------------------------------------------
I only receive 5798 bytes of "db.corp-3" out of a possible 12452 bytes.
I know that the "receiving responses: multiple RRs of singleton type"
indicates multiple CNAME pointers on the master server for "corp-1". I
have tried setting "multiple-cnames yes:" in my named.conf, but that
is not yet implemented in BIND9.
Questions:
1. Why can't I receive any records from "corp-1.com"?
2. Why do I only receive partial records from corp-3.com
3. Why can I recieve all records from "corp-2.com"?
As far as I know all the master servers are running BIND8.
Thanx,
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Phil Macias * pmacias at okapi.kenan.com * 609 639 2172
More information about the bind-users
mailing list