Named service won't start - seems to be caching errors

Question

When I try to restart / start the named service, it keeps failing.

service named restart

Results in:-

Stopping named:                                            [  OK  ]
Starting named:
Error in named configuration:
/etc/named.conf:2: missing ';' before '}'
                                                           [FAILED]

service named start

Results in:-

Starting named:
Error in named configuration:
/etc/named.conf:2: missing ';' before '}'
                                                           [FAILED]

The thing is, even if named.conf contains no syntax errors or even if the file doesn't exist at all, I'm still getting this error preventing me from starting the service.

Currently trying with a named.conf containing the following:-

options {
listen-on port 53 { 127.0.0.1; };
};

And checking the syntax of named.conf using

named-checkconf /etc/named.conf

Results in no errors. Yet when starting the service, I still get the above error.

I can't see any other errors in the logs relating to this and don't really know where to get to the bottom of this. Any ideas?

(Even reverting back to the named.conf file that was in place before I stopped the service returns the same error now and this was previously running fine).

Contents on init.d/named can be seen at this pastebin.

CentOS 6.4 - Linux version 2.6.32-358.6.2.el6.x86_64 ([email protected]) (gcc version 4.4.7 20120313 (Red Hat 4.4.7-3) (GCC) ) #1 SMP Thu May 16 20:59:36 UTC 2013

Update

bash -x /etc/init.d/named start

Snippet from results possibly of relevance?

+ echo 'Error in named configuration:'
Error in named configuration:
+ echo '/etc/named.conf:2: missing '\'';'\'' before '\''}'\'''
/etc/named.conf:2: missing ';' before '}'
+ failure
+ local rc=0
+ '[' color '!=' verbose -a -z '' ']'
+ echo_failure
+ '[' color = color ']'
+ echo -en '\033[60G'
                                                           + echo -n '['
[+ '[' color = color ']'
+ echo -en '\033[0;31m'
+ echo -n FAILED
FAILED+ '[' color = color ']'
+ echo -en '\033[0;39m'
+ echo -n ']'
]+ echo -ne '\r'
+ return 1
+ '[' -x /bin/plymouth ']'
+ /bin/plymouth --details
+ return 0
+ echo

Solution

Okay, so I managed to resolve this by changing one line in /etc/init.d/named:-

Line 29 in /etc/init.d/named changed from...

named='named'

To...

named='named -u named'

Then service would start back up again without issue. Nothing like a misleading error message!

Sigh.

Comments

[S edwards]

have you watch the script that probably exists in /etc/init.d/named ? the errors probably come from there

[zigojacko]

Even though it specifically states /etc/named.conf file? (I've run a named-checkconfig on /etc/init.d/named as well but nothing flagged that would prevent service from starting). Thanks.

[Anthon]

@GeoffJackson-zigojacko yes, the service command does not execute /etc/named.conf it probably executes /etc/init.d/named, and that you cannot check with named-checkconf as it is not a named config file.

[Anthon]

@GeoffJackson-zigojacko what is your distribution? (Add a tag to the question).

[zigojacko]

I see, thanks @Anthon (distro CentOS 6.4 - full version added above). What would be the process to check for issues in /etc/init.d/named please?

[S edwards]

@GeoffJackson-zigojacko copyt paste the content in your question or just open it and see where the error can happen to identify the problem, most of the time this is just a bash script. you can also tries bash -x /etc/init.d/named start to debug the script

[zigojacko]

Thanks @Kiwy - will try the bash debug. Content can be seen at this pastebin too.

[Anthon]

@GeoffJackson-zigojacko And if the debug doesn't work, you can always insert some print statements in the scripts to narrow down where in the /etc/init.d/named script the message is generated.

[zigojacko]

Managed to resolve by fluke (see my answer). Not entirely sure why this fixed it /was causing the error though. Thanks for both of your help.

[S edwards]

it's nice that you find a solution.