Package com.yahoo.search.cluster

Class MonitorConfiguration

  • public class MonitorConfiguration
extends java.lang.Object
    The configuration of a cluster monitor instance
    Author:
    bratseth

    • Method Summary

      All Methods Instance Methods Concrete Methods 
      Modifier and Type Method Description
      long getCheckInterval()
      Returns the interval between each ping of idle or failing nodes Default is 1000ms
      long getFailLimit()
      Returns the number of milliseconds a node is allowed to fail before we mark it as not working
      long getIdleLimit()
      Gets the number of ms a node (failing or working) is allowed to stay idle before it is pinged.
      long getRequestTimeout()
      Returns the number of milliseconds to attempt to service a request (at different nodes) before giving up.
      void setCheckInterval​(long intervalMs)
      Sets the interval between each ping of idle or failing nodes Default is 1000ms
      void setFailLimit​(long failLimit)
      Sets the number of milliseconds a node is allowed to fail before we mark it as not working
      void setFailQuarantineLimit​(int failQuarantineLimit)
      The number of times a node must fail in one hour to be placed in quarantine.
      void setIdleLimit​(int idleLimit)
      Sets the number of ms a node (failing or working) is allowed to stay idle before it is pinged.
      void setQuarantineTime​(long quarantineTime)
      The number of ms an unstable node is quarantined.
      void setResponseAfterFailLimit​(int responseAfterFailLimit)
      Sets the number of times a failed node must respond before it is put back in service.
      java.lang.String toString()  

      • Methods inherited from class java.lang.Object

        clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

    • Constructor Detail

      • MonitorConfiguration

        public MonitorConfiguration()

    • Method Detail

      • setCheckInterval

        public void setCheckInterval​(long intervalMs)
        Sets the interval between each ping of idle or failing nodes Default is 1000ms

      • getCheckInterval

        public long getCheckInterval()
        Returns the interval between each ping of idle or failing nodes Default is 1000ms

      • setResponseAfterFailLimit

        public void setResponseAfterFailLimit​(int responseAfterFailLimit)
        Sets the number of times a failed node must respond before it is put back in service. Default is 3.

      • setIdleLimit

        public void setIdleLimit​(int idleLimit)
        Sets the number of ms a node (failing or working) is allowed to stay idle before it is pinged. Default is 3000

      • getIdleLimit

        public long getIdleLimit()
        Gets the number of ms a node (failing or working) is allowed to stay idle before it is pinged. Default is 3000

      • getRequestTimeout

        public long getRequestTimeout()
        Returns the number of milliseconds to attempt to service a request (at different nodes) before giving up. Default is 5000 ms.

      • setFailLimit

        public void setFailLimit​(long failLimit)
        Sets the number of milliseconds a node is allowed to fail before we mark it as not working

      • getFailLimit

        public long getFailLimit()
        Returns the number of milliseconds a node is allowed to fail before we mark it as not working

      • setFailQuarantineLimit

        public void setFailQuarantineLimit​(int failQuarantineLimit)
        The number of times a node must fail in one hour to be placed in quarantine. Once in quarantine it won't be put back in productuion before quarantineTime has expired even if it is working. Default is 3

      • setQuarantineTime

        public void setQuarantineTime​(long quarantineTime)
        The number of ms an unstable node is quarantined. Default is 100*60*60

      • toString

        public java.lang.String toString()
        Overrides:
        toString in class java.lang.Object