Bind problems with Debian Squeeze

Discussion in 'Linux Networking' started by Rob van der Putten, Jun 26, 2011.

  1. Hi there


    Sometimes RNDC doesn't work, making stopping bind using
    /etc/init.d/bind9 impossible (rndc: connect failed : connection
    refused.). Some suggest that /etc/bind/rndc.key should be owned
    root:bind instead of bind:bind, but that doesn't help.
    I put a retry counter in the /etc/init.d/bind9 stop section. If it
    expires it bluntly kills Bind. I hope that helps

    Bind starts too late. Some services which need Bind won't start.
    I blame insserv; NFS is stared before Bind, but what's to use of NFS
    when you don't have a running nameserver?
    LprNg won't start either.
    The weird thing is that I have all the required information in
    /etc/hosts and a order hosts,bind.
    Can I run Squeeze without insserv? I think that this will solve the problem.


    Regards,
    Rob
     
    Rob van der Putten, Jun 26, 2011
    #1
    1. Advertisements

  2. Hi there

    This works.
    Creating a (empty) file '.legacy-bootordering' in '/etc/init.d/' will
    make insserv mimic the 'old fashioned' boot behaviour. That is, without
    concurrency. This will start Bind earlier, preventing the LprNg problems.
    The NFS problems were due to a 'stale' directory 'mountnfs' in
    '/var/run/network/'.


    Regards,
    Rob
     
    Rob van der Putten, Jun 27, 2011
    #2
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.