Uploaded image for project: 'Jenkins'
  1. Jenkins
  2. JENKINS-10160

When the network is broken,then jenkins will query dns info in loop.The log file will grow up very fast.

    • Icon: Bug Bug
    • Resolution: Duplicate
    • Icon: Major Major
    • core
    • None
    • Windows XP Pro|SunJDK1.6.0_21|Tomcat 7.0.14|Jenkins1.418
      Debian 6|OpenJDK1.6.0_18|Tomcat7.0.14|Jenkins1.413
      Server 2008 R2 x64|Sun JDK 1.6.0_25 Jenkins 1.411 and Jenkins 1.413

      The network in my company is often broken.When it was broken,the log file of jenkins would grow up very fast until there is no disk space for it.
      Here is the log in my Windows.

      	question:      [DNSQuestion@9657493 type: TYPE_IGNORE index 0, class: CLASS_UNKNOWN index 0, name: ]
      	question:      [DNSQuestion@29482661 type: TYPE_IGNORE index 0, class: CLASS_UNKNOWN index 0, name: ]
      	question:      [DNSQuestion@16787101 type: TYPE_IGNORE index 0, class: CLASS_UNKNOWN index 0, name: ]
      	question:      [DNSQuestion@12328574 type: TYPE_IGNORE index 0, class: CLASS_UNKNOWN index 0, name: ]
      	question:      [DNSQuestion@7738582 type: TYPE_IGNORE index 0, class: CLASS_UNKNOWN index 0, name: ]
      	question:      [DNSQuestion@776128 type: TYPE_IGNORE index 0, class: CLASS_UNKNOWN index 0, name: ]
      	question:      [DNSQuestion@25267887 type: TYPE_IGNORE index 0, class: CLASS_UNKNOWN index 0, name: ]
      	question:      [DNSQuestion@8281222 type: TYPE_IGNORE index 0, class: CLASS_UNKNOWN index 0, name: ]
      	question:      [DNSQuestion@32275542 type: TYPE_IGNORE index 0, class: CLASS_UNKNOWN index 0, name: ]
      	question:      [DNSQuestion@32262868 type: TYPE_IGNORE index 0, class: CLASS_UNKNOWN index 0, name: ]
      

          [JENKINS-10160] When the network is broken,then jenkins will query dns info in loop.The log file will grow up very fast.

          Patrick Rose added a comment -

          recampbell - this one, there's a compilation error I need to look into but I've not had the chance to yet.

          Patrick Rose added a comment - recampbell - this one , there's a compilation error I need to look into but I've not had the chance to yet.

          Furai added a comment -

          This bug have again brought my machine down. Is there any fix to it or a workaround?

          Furai added a comment - This bug have again brought my machine down. Is there any fix to it or a workaround?

          Oleg Nenashev added a comment -

          As discussed in https://github.com/jenkinsci/jmdns/pull/2, my proposal is to replace the custom library fork by the official release from https://github.com/jmdns/jmdns . All custom patches were before 2010, and I doubt they are really needed anymore

          danielbeck jglick would be nice to get your feedback here

           

          Oleg Nenashev added a comment - As discussed in https://github.com/jenkinsci/jmdns/pull/2 , my proposal is to replace the custom library fork by the official release from  https://github.com/jmdns/jmdns . All custom patches were before 2010, and I doubt they are really needed anymore danielbeck jglick would be nice to get your feedback here  

          Jesse Glick added a comment -

          Just do JENKINS-33596 already so that it becomes someone else’s problem: whoever decides to maintain that plugin.

          Jesse Glick added a comment - Just do JENKINS-33596 already so that it becomes someone else’s problem: whoever decides to maintain that plugin.

          Furai added a comment -

          But if I understood right - people who were running their war directly with

          -Dhudson.DNSMultiCast.disabled=true 
          

          and

          -Dhudson.udp=-1
          

          Were still getting the issue as per this link.  Is this somehow related? Or are this options disabling completely different things?

          Furai added a comment - But if I understood right - people who were running their war directly with -Dhudson.DNSMultiCast.disabled= true and -Dhudson.udp=-1 Were still getting the issue as per this link .  Is this somehow related? Or are this options disabling completely different things?

          The reason why the workaround doesn't work for people, is because of the bad ordering of the parameters.

          The parameters

           -Dhudson.DNSMultiCast.disabled=true -Dhudson.udp=-1
          

          are parameters to java, not jenkins, and therefore should be specified after java command but before 

          -jar /usr/share/jenkins/jenkins.war
          

          All parameters after jenkins.war are parameters to jenkins.

          So, check the process command with 

          pgrep -a java
          

          and you should see something similar to

          6916 /usr/bin/java -Djava.awt.headless=true -Dhudson.DNSMultiCast.disabled=true -Dhudson.udp=-1 -jar /usr/share/jenkins/jenkins.war --webroot=/var/cache/jenkins/war --httpPort=8080 --ajp13Port=-1 --httpListenAddress=192.168.255.1
          

          Hope that helps.

           

           

           

          Lior Goikhburg added a comment - The reason why the workaround doesn't work for people, is because of the bad ordering of the parameters . The parameters  -Dhudson.DNSMultiCast.disabled= true -Dhudson.udp=-1 are parameters to java, not jenkins , and therefore should be specified after java command but before  -jar /usr/share/jenkins/jenkins.war All parameters after jenkins.war are parameters to jenkins . So, check the process command with  pgrep -a java and you should see something similar to 6916 /usr/bin/java -Djava.awt.headless= true -Dhudson.DNSMultiCast.disabled= true -Dhudson.udp=-1 -jar /usr/share/jenkins/jenkins.war --webroot=/ var /cache/jenkins/war --httpPort=8080 --ajp13Port=-1 --httpListenAddress=192.168.255.1 Hope that helps.      

          Furai added a comment -

          What's the status on the fix to that? Recently it brought my machine down yet again. Maybe I should just containerise jenkins instance.

          Might actually add those java parameters mentioned above to prevent it from happening.

          Furai added a comment - What's the status on the fix to that? Recently it brought my machine down yet again. Maybe I should just containerise jenkins instance. Might actually add those java parameters mentioned above to prevent it from happening.

          Patrick Rose added a comment - - edited

          Might actually add those java parameters mentioned above to prevent it from happening.

          We're running with those java args, and it still happens to us (albeit infrequently)

          EDIT: Which makes me wonder if JENKINS-33596 would actually fix the issue? If we're turning off on our end, I'm not sure why we keep getting the issue once every 6 months or so.

          Patrick Rose added a comment - - edited Might actually add those java parameters mentioned above to prevent it from happening. We're running with those java args, and it still happens to us (albeit infrequently) EDIT : Which makes me wonder if JENKINS-33596 would actually fix the issue? If we're turning off on our end, I'm not sure why we keep getting the issue once every 6 months or so.

          Patrick Rose added a comment -

          Thanks to fcojfernandez, this should now be fixed in 2.178!

          Patrick Rose added a comment - Thanks to fcojfernandez , this should now be fixed in 2.178!

          Oleg Nenashev added a comment -

          We believe that JENKINS-25369 fixes it as well. Please reopen if you experience any issues

          Oleg Nenashev added a comment - We believe that  JENKINS-25369 fixes it as well. Please reopen if you experience any issues

            Unassigned Unassigned
            hellyguo Helly Guo
            Votes:
            32 Vote for this issue
            Watchers:
            46 Start watching this issue

              Created:
              Updated:
              Resolved: