DirectSlave - solution to add and remove zones on slave without DirectAdmin

Hey Guys,
is there a proper way to make suggestions for this project?

i.e add SSL support for DirectSlave?
 
Yes, the maker of the project, Roman_m is a very nice guy and responsive to his emails. So you can make a suggestion in this thread but he will possibly see it sooner if you write a suggestion in an email to him.
 
Suddenly I got these errors:
rndc: connect failed: 127.0.0.1#953: connection refused
rndc: connect failed: 127.0.0.1#953: connection refused
flock() on closed filehandle FILE at /usr/local/directslave/bin/directslave line 191.
print() on closed filehandle FILE at /usr/local/directslave/bin/directslave line 192.
rndc: connect failed: 127.0.0.1#953: connection refused
rndc: connect failed: 127.0.0.1#953: connection refused
flock() on closed filehandle FILE at /usr/local/directslave/bin/directslave line 191.
print() on closed filehandle FILE at /usr/local/directslave/bin/directslave line 192.
rndc: connect failed: 127.0.0.1#953: connection refused
rndc: connect failed: 127.0.0.1#953: connection refused

Changes on existing zones are possible, put the zone file of new domains are not created.
Though the path to the zone file are defined in the named.conf
 
Suddenly I got these errors:

Changes on existing zones are possible, put the zone file of new domains are not created.
Though the path to the zone file are defined in the named.conf

You got this errors after updating your system and perl distribution to major version? What version of DirectSlave you use?
 
Now I think what caused the errors, I updated Bind.
I use DirectSlave 1.3.

Check permissions of zone files stored in 'named_workdir', if you use DirectSlave 1.3 in 'master' mode.
It's better to switch to 'slave' mode because new versions of bind (9.9.2 for example) use his own format to store secondary zones, fetched from primary servers and you doesn't need to rewrite zone .db configs any more.

Also, using rndc from DirectSlave is recommended to correctly reload your secondary nameserver to minimize the time needed for secondary response on new zones arrived.
 
Last edited:
I confirm that it work with multiple access perfectly.

Tested on CentOS 6.4 64Bit, gonna test on CentOS 5.9 64Bit this evening and report.

Regards
 
Check permissions of zone files stored in 'named_workdir', if you use DirectSlave 1.3 in 'master' mode.
It's better to switch to 'slave' mode because new versions of bind (9.9.2 for example) use his own format to store secondary zones, fetched from primary servers and you doesn't need to rewrite zone .db configs any more.

Also, using rndc from DirectSlave is recommended to correctly reload your secondary nameserver to minimize the time needed for secondary response on new zones arrived.

Can I just change "master" to "slave" and save and restart or are there more things to do?
 
Work fine on CentOS 5.9 too.

The only thing to do is be sure to install perl modules via CPAN, the one from yum are not up2date.

Regards
 
Can I just change "master" to "slave" and save and restart or are there more things to do?

No.

You need to set up allow AXFR on primary NS for secondary, then reconfigure DirectSlave to 'slave', then push your domain configs to DirectSlave server doing

Code:
 echo "action=rewrite&value=named" >> /usr/local/directadmin/data/task.queue
on each primary DirectAdmin servers.

Also, you can continue to use 'master' mode, but in this case you need to double-check 'named_workdir' permission, should be ug+rw
 
Last edited:
Hello Roman,

First of all, thanks for an outstanding good support of you.

I updated to DirectSlave 2.1.
When I run it in debug mode, I get an output of the test connection and of updates of zone files, but no output if I create a new domain.
The test connection says everything is fine and the --check option of directslave also says everything is fine.

No.

You need to set up allow AXFR on primary NS for secondary, then reconfigure DirectSlave to 'slave', then push your domain configs to DirectSlave server doing

Code:
 echo "action=rewrite&value=named" >> /usr/local/directadmin/data/task.queue
on each primary DirectAdmin servers.

Also, you can continue to use 'master' mode, but in this case you need to double-check 'named_workdir' permission, should be ug+rw
 
Just a question, does the zone get created or it doesnt?

If it does is just a missing echo, so, nothing to worry about, if it doesnt that's would be a problem cause should mean isnt working correctly.

I actually suppose is just an echo error, otherwise i did notice that also ;)

Regards
 
It's not creating the zone file.

Just a question, does the zone get created or it doesnt?

If it does is just a missing echo, so, nothing to worry about, if it doesnt that's would be a problem cause should mean isnt working correctly.

I actually suppose is just an echo error, otherwise i did notice that also ;)

Regards
 
Thats strange, i did just test it and in activity.log i can see the zone creation and also added to named correctly.

Are you sure DirectSlave is running? Have you done the multi-server test? Does it reply correctly?

Regards
 
DirectSlave/2.1-beta connection OK.
Trying to create/open file 'access_log' -> /usr/local/directslave/log/access.log...
OK.
Trying to create/open file 'action_log' -> /usr/local/directslave/log/action.log...
OK.
Trying to create/open file 'error_log' -> /usr/local/directslave/log/error.log...
OK.
Trying to create/open file 'named_conf' -> /var/named/secondary/named.conf...
OK.
Trying to create/open file 'pid' -> /usr/local/directslave/run/directslave.pid...
not found, creating... OK.

Access.log:
178.21.112.90 - - [05/Sep/2013:15:42:17 +0000] "GET /CMD_API_LOGIN_TEST HTTP/1.0" 200 34 "-" "-"
Action.log:
2013-09-05 15:42:17: Auth passed for admin from xxx.xxx.xxx.xxx

The named rewrite is not listed in the logs.
 
Back
Top