example:
INSERT INTO coredns_records (zone, name, ttl, content, record_type) VALUES
('example.org.', '@', 30, '{"ip": "1.1.1.1"}', 'A'),
('example.org.', '*', '60', '{"ip": "1.1.1.0"}', 'A');

example.org A 1.1.1.1

abc.example.org A 1.1.1.0
123.example.org A 1.1.1.0
ccc.example.org A 1.1.1.0

When you use an @ symbol in name field of an SOA or NS record its not found. You have to leave the name field blank. Just a bit confusing when converting from text zone files.

Hi there,

first: Thanks for the plugin. Really nice and easy to get to work.

But I think I found a bug: Currently it is impossible to create records for a zone itself. Only for subdomains.

Based on your examples I am using foo.example.org as zone name.

When a request for "test.foo.example.org" is received, the completed name ist "test.foo.example.org."
"." + Zonename (".foo.example.org.") is stripped out. Thats right.
So the query is:

SELECT name, zone, ttl, record_type, content FROM coredns_records WHERE zone = 'foo.example.org.' AND name = 'test' AND record_type IN ('A')

But when you do a request for "foo.example.org" also ".foo.example.org." is tried to stripped out. This fails of course. The final query is:

SELECT name, zone, ttl, record_type, content FROM coredns_records WHERE zone = 'foo.example.org.' AND name = 'foo.example.org.' AND record_type IN ('A')

This does not find any records. If I try to use the data like in the query given the result will be for "foo.example.org.foo.example.org" which also causes an error.

I think the easiest solution is to set the name empty if the name matches the zone.

Thanks and greetings from Germany
Lukas

2023-10-24 12:11:47 [INFO] 172.17.0.1:39435 - 31667 "A IN foo.mydomain.com. udp 39 false 512" NOERROR qr,aa,rd 76 0.040718436s
2023-10-24 12:12:00 [INFO] 172.17.0.1:41790 - 11714 "A IN boo.mydomain.com. udp 39 false 512" NOERROR qr,aa,rd 140 0.109066749s

Foo exists while Boo doesn't. Same RCODE response. Should the Boo query failing give a different RCODE? One like NXDOMAIN or NOTZONE? Without a different code the "alternative" plugin will not work. It uses the RCODE to forward failed queries.

conf/Corefile

. {
    mysql 'admin':Admin_123@tcp(127.0.0.1:3306)/gva?param=value
}
. {
    bind 0.0.0.0
    forward . 8.8.8.8
    cache 3600 .
    log
}

Error
plugin/mysql: Error 1045: Access denied for user ''@'172.17.0.1' (using password: NO)

Hi, I would like to ask which version of the coredns source code this plugin is based on? After compiling coredns-1.11 successfully, I can connect to mysql, and the debug "records" have extracted data from mysql, but it still timeout and the log error is NOERROR qr,aa,rd.

Hi guys,
How can I add a PTR record in MySQL DB?

I have spent over a week trying to understand "go" and how this plugin works. I can't quite figure out how to fix it.

Its not showing the authoritative name servers when doing an SOA dig.

If I do a dig @0.0.0.0 SOA example.org with the "file" plugin I get the ANSWER SECTION and the AUTHORITY SECTION:

;; ANSWER SECTION:
example.org.            3600    IN      SOA     ns1.example.org. email.example.org. 2023022400 600 1800 2419200 600

;; AUTHORITY SECTION:
example.org.            3600    IN      NS      ns1.example.org.
example.org.            3600    IN      NS      ns2.example.org.

;; Query time: 1 msec
;; SERVER: 127.0.0.1#53(0.0.0.0)
;; WHEN: Fri Feb 24 16:07:47 CST 2023
;; MSG SIZE  rcvd: 199

If I do a dig @0.0.0.0 SOA example.org with the "mysql" plugin I get the ANSWER SECTION but NOT the AUTHORITY SECTION:

;; ANSWER SECTION:
example.org.            30      IN      SOA     ns1.example.org. email.example.org. 1677276345 600 1800 2419200 600

;; Query time: 3 msec
;; SERVER: 127.0.0.1#53(0.0.0.0)
;; WHEN: Fri Feb 24 16:05:45 CST 2023
;; MSG SIZE  rcvd: 119

Here is my Corefile (I am commenting out the plugin that I am not using to test) :

. {
    forward . 8.8.8.8 9.9.9.9
    log
    errors
    cache
    debug
    reload
}

example.org {
    file corednsconf/db.example.org
    # mysql {
    #     dsn root:password@tcp(192.168.40.198:3306)/coredns
    # }
    log
    errors
    debug
}

Attached are the zone file and the sql file that I am using to test with.
coredns_2023-02-24.sql.zip
db.example.org.zip

Hello,

Is this how I would server more than one domain from the same database?

This is my current Corefile...

. {
    forward . 8.8.8.8 9.9.9.9
    log
    errors
    debug
    cache
    reload
}

example.org {
    mysql {
        dsn root:XXXXXXXXXX@tcp(coredns.us-east-1.rds.amazonaws.com:3306)/coredns
    }
    log
    errors
    debug
}

example.com {
    mysql {
        dsn root:XXXXXXXXXX@tcp(coredns.us-east-1.rds.amazonaws.com:3306)/coredns
    }
    log
    errors
    debug
}

I'll move the dsn to an environment variable later, I am just trying to do a POC.

Thanks for any help.

Chris

Hello,

thanks for Working on this nice Plugin.
I try to setup it up for Prometheus DNS service discovery.
When I try to query a SRV Record the request just runs into an timeout.
MariaDB [dns]> select * from coredns_records;

When I try to query the SRV Record I get an Timeout on every DNS Client I tested. The A Record works without an Problem.
In the Log I just see three Requests:

[INFO] [IPV6_ADDRESS]:52152 - 30764 "SRV IN nodeexporter._tcp.mon.local. udp 68 false 4096" NOERROR qr,aa,rd 136 0.001510555s
[INFO] [IPV6_ADDRESS]:52152 - 30764 "SRV IN nodeexporter._tcp.mon.local. udp 68 false 4096" NOERROR qr,aa,rd 136 0.001565826s
[INFO] [IPV6_ADDRESS]:52152 - 30764 "SRV IN nodeexporter._tcp.mon.local. udp 68 false 4096" NOERROR qr,aa,rd 136 0.002095277s

Corefile:

.:53 {
    forward . 149.112.112.112 2620:fe::fe 185.95.218.42 2a05:fc84::43 5.1.66.255  2001:678:ed0:f000::
    prometheus [::]:9253
    errors
    log
    debug
}
mon.local.:53 {
mysql {
    dsn coredns:coredns@tcp(127.0.0.1:3306)/dns
    }
    errors
    log
    debug
}

Could you please look into this Problem?

IF I have a NS record that isn't FQDN with a trailing period I get an error.

504 cloud.com. 3600 {"host":"infr-pdnsn-401"} NS

dig @localhost NS cloud.com

; <<>> DiG 9.10.6 <<>> @localhost NS cloud.com
; (2 servers found)
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: SERVFAIL, id: 30356
;; flags: qr rd; QUERY: 1, ANSWER: 0, AUTHORITY: 0, ADDITIONAL: 1
;; WARNING: recursion requested but not available

;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 4096
;; QUESTION SECTION:
;cloud.com.             IN      NS

;; Query time: 27 msec
;; SERVER: ::1#53(::1)
;; WHEN: Fri Feb 10 09:32:05 CST 2023
;; MSG SIZE  rcvd: 46

If I add the FQDN it works as expected.

504 cloud.com. 3600 {"host":"infr-pdnsn-401.cloud.com."} NS

Requesting any record type for domain.tld doesn't work, for 2 reasons:

1. The database request is sent for zone equal domain.tld AND name equal domain.tld. name should be an empty string.
2. The response (FQDN) is incorrectly formatted, it is returned for .domain.tld, where it should have been returned for domain.tld.

Hello ,

Are there any plans for wildcard support soon ?
It is the only feature missing that prevents me from picking this up :)

It feels like it could be similar to etcd, which implements the ServiceBackend interface.Is there a problem?