#1
  1. No Profile Picture
    Registered User
    Devshed Newbie (0 - 499 posts)

    Join Date
    Aug 2011
    Posts
    3
    Rep Power
    0

    Bind9 DNS on Solaris 10 x4270 & CPU usage


    I have configured a Bind9 DNS on a X4270 machine with Solaris10

    I am excuting some repformance tests with DNSPERF tool and maximun CPU usage is 23%. I have seen with
    prstat -L -p PID
    that named process usses only 2 of the 8 available CPU at the same time although threads for all CPUs exist.

    Any idea why this happens and how can I change it? I want to reach at least 70%.
  2. #2
  3. No Profile Picture
    Registered User
    Devshed Newbie (0 - 499 posts)

    Join Date
    Jul 2009
    Location
    Los Angeles
    Posts
    11
    Rep Power
    0
    Originally Posted by parisph
    I have configured a Bind9 DNS on a X4270 machine with Solaris10

    I am excuting some repformance tests with DNSPERF tool and maximun CPU usage is 23%. I have seen with
    prstat -L -p PID
    that named process usses only 2 of the 8 available CPU at the same time although threads for all CPUs exist.

    Any idea why this happens and how can I change it? I want to reach at least 70%.
    I recently upgraded our servers OS to solaris 10. Now the DNS process runs away and quits after using up all the swap space.
  4. #3
  5. No Profile Picture
    Registered User
    Devshed Newbie (0 - 499 posts)

    Join Date
    Sep 2011
    Posts
    7
    Rep Power
    0
    Hello, may you please attach your named.conf and your zone file, something is terribly wrong. It's insane usage of CPU.
  6. #4
  7. No Profile Picture
    Registered User
    Devshed Newbie (0 - 499 posts)

    Join Date
    Aug 2011
    Posts
    3
    Rep Power
    0
    Finally I found a work around. By entering logging option in named.conf to save querries in a new file and not using syslogd and /var/adm/messages problem was solved. Limitation was connected somehow with solaris 10 and syslog.
  8. #5
  9. No Profile Picture
    Registered User
    Devshed Newbie (0 - 499 posts)

    Join Date
    Jul 2009
    Location
    Los Angeles
    Posts
    11
    Rep Power
    0
    Originally Posted by parisph
    Finally I found a work around. By entering logging option in named.conf to save querries in a new file and not using syslogd and /var/adm/messages problem was solved. Limitation was connected somehow with solaris 10 and syslog.
    Can you post your fix? Thanks.
  10. #6
  11. No Profile Picture
    Registered User
    Devshed Newbie (0 - 499 posts)

    Join Date
    Aug 2011
    Posts
    3
    Rep Power
    0

    logging for DNS


    logging {
    channel query.log {
    file "query.log" size 200m;
    print-time yes;
    };
    category queries { query.log; };
    };
  12. #7
  13. No Profile Picture
    Registered User
    Devshed Newbie (0 - 499 posts)

    Join Date
    Jul 2009
    Location
    Los Angeles
    Posts
    11
    Rep Power
    0
    Originally Posted by parisph
    logging {
    channel query.log {
    file "query.log" size 200m;
    print-time yes;
    };
    category queries { query.log; };
    };
    Awesome. Thanks.

IMN logo majestic logo threadwatch logo seochat tools logo