Previous  |  Next  >  
Product: Cluster Server Guides   
Manual: Cluster Server 4.1 Agent Developer's Guide   

MonitorInterval

Duration (in seconds) between two consecutive monitor calls for an ONLINE or transitioning resource. Default is 60 seconds. The MonitorInterval attribute value can be overridden.

MonitorStatsParam

Refer to the VERITAS Cluster Server User's Guide for details about the MonitorStatsParam attribute, and the MonitorTimeStats attribute that is updated by VCS. See also ComputeStats.

MonitorStatsParam is a resource type-level attribute, which stores the required parameter values for calculating monitor time statistics.


static str MonitorStatsParam = { Frequency = 10, ExpectedValue =
  3000, ValueThreshold = 100, AvgThreshold = 40 }   

  • Frequency: Defines the number of monitor cycles after which the average monitor cycle time should be computed and sent to the engine. The value for this attribute is must be between 1 and 30 and is set to 0 by default.
  • ExpectedValue: The expected monitor time in milliseconds for all resources of this type. Default=3000.
  • ValueThreshold: The acceptable percentage difference between the expected monitor cycle time (ExpectedValue) and the actual monitor cycle time. Default=100.
  • AvgThreshold: The acceptable percentage difference between the benchmark average and the moving average of monitor cycle times. Default=40.

The MonitorStatsParam attribute values can be overridden.

MonitorTimeout

Maximum time (in seconds) within which the monitor entry point must complete or else be terminated. Default is 60 seconds. The MonitorTimeout attribute value can be overridden.

The determination of the value of the MonitorTimeout attribute can be assisted by the use of the MonitorStatsParam attribute.

NumThreads

Number of threads used within the agent process for managing resources. This number does not include the three threads used for other internal purposes. Default is 10 threads. The NumThreads attribute value cannot be overridden.

OfflineMonitorInterval

Duration (in seconds) between two consecutive monitor calls for an OFFLINE resource. If set to 0, OFFLINE resources are not monitored. Default is 300 seconds. The OfflineMonitorInterval attribute value can be overridden.


Note   Note    In previous releases, agents monitored offline resources once every minute by default. To reduce monitoring overhead, this frequency was changed to once every five minutes. This interval may be adjusted to meet specific configuration requirements.

OfflineTimeout

Maximum time (in seconds) within which the offline entry point must complete or else be terminated. Default is 300 seconds. The OfflineTimeout attribute value can be overridden.

OnlineRetryLimit

Number of times to retry online, if the attempt to online a resource is unsuccessful. This attribute is meaningful only if clean is implemented. Default is 0. The OnlineRetryLimit attribute value can be overridden.

OnlineTimeout

Maximum time (in seconds) within which the online entry point must complete or else be terminated. Default is 300 seconds. The OnlineTimeout attribute value can be overridden.

OnlineWaitLimit

Number of monitor intervals to wait after completing the online procedure, and before the resource becomes online. If the resource is not brought online after the designated monitor intervals, the online attempt is considered ineffective. This attribute is meaningful only if the clean entry point is implemented.

If clean is not implemented, the agent continues to periodically run monitor until the resource comes online.

If clean is implemented, when the agent reaches the maximum number of monitor intervals it assumes that the online procedure was ineffective and runs clean. The agent then notifies the engine that the online failed, or retries the procedure, depending on whether or not the OnlineRetryLimit is reached. Default is 2. The OnlineWaitLimit attribute value can be overridden.

OpenTimeout

Maximum time (in seconds) within which the open entry point must complete or else be terminated. Default is 60 seconds. The OpenTimeout attribute value can be overridden.

Operations

Indicates the valid operations of resources of the resource type. The values are OnOff (can be onlined and offlined), OnOnly (can be onlined only), and None (cannot be onlined or offlined). Default is OnOff. The Operations attribute value cannot be overridden.

ResourceInfo

The ResourceInfo (association-string) is a temporary attribute, the scope of which is set by the engine to be global for failover groups or local for parallel groups. Because ResourceInfo is a temporary attribute, its values are never dumped to the main.cf file.

The values of the ResourceInfo attribute are expressed in three mandatory keys: State, Msg, and TS.For State, the possible values are "Valid," (the default), "Invalid," and "Stale"). Msg (default is "", a null string) contains the output from the entry point. TS contains the time at which the attribute is has been updated or modified. These mandatory keys are updated only by the agent framework, not the entry point. The entry point can define and add other keys (name-value pairs) and update them.

The value of the ResourceInfo attribute can be displayed using the hares command. The output of hares -display shows the first 20 characters of the current value; the output of hares -value resource ResourceInfo shows all name-value pairs in the keylist.

The resource for which the info entry point is invoked must be online. When a resource goes offline or faults, the State key is marked "Stale" because the information is not current. If the info entry point exits abnormally, the State key is marked "Invalid" because not all of the information is known to be valid. The other key data, including Msg and TS keys, are not touched. The values of the ResourceInfo attribute can be manually cleared using the hares -flushinfo command. This command deletes any optional keys for the ResourceInfo attribute and sets the three mandatory keys to their default values.

See the hares manual page.

RestartLimit

Affects how the agent responds to a resource fault (see FaultOnMonitorTimeouts and ToleranceLimit). A non-zero RestartLimit causes VCS to invoke the online entry point instead of failing over the service group to another system. VCS attempts to restart the resource according to the number set in RestartLimit before it gives up and fails over. However, if the resource remains online for the interval designated in ConfInterval, earlier attempts to restart are not counted against RestartLimit. Default is 0. The RestartLimit attribute value can be overridden.


Note   Note    The agent will not restart a faulted resource if the clean entry point is not implemented. Therefore, the value of the RestartLimit attribute applies only if clean is implemented.

RegList

RegList is a type-level attribute of the type keylist that can be used to store, or register, a list of certain resource level attributes. The agent calls the attr_changed entry point for a resource when the value of an attribute listed in RegList is modified. The RegList attribute is useful where a change in the values of important attributes require specific actions that can be executed from the attr_changed entry point.

By default, the attribute RegList is not included in a resource's type definition, but it can be added using either of the two methods shown below.

Assume the RegList attribute is added to the FileOnOff resource type definition and its value is defined as PathName. Thereafter, when the value of the PathName attribute for a FileOnOff resource is modified, the attr_changed entry point is called.

  • Method one is to modify the types definition file (types.cf, for example) to include the RegList attribute. Add a line in the definition of a resource type that resembles:

  •   static keylist RegList = { attribute1_name, attribute2_name, ...}
    For example, if the type definition is for the FileOnOff resource and the name of the attribute to register is PathName, the modified type definition would resemble:

      .
      .
      .
      type FileOnOff (
        str PathName
        static keylist RegList = { PathName }
        static str ArgList[] = { PathName }
      )
      .
      .
  • Method two is to use the haattr command to add the RegList attribute to a resource type definition and then modify the value of the type's RegList attribute using the hatype command; the commands are:

  •   haattr -add -static resource_type RegList -keylist
      hatype -modify resource_type RegList attribute_name
    For example:

      # haattr -add -static FileOnOff RegList -keylist
      # hatype -modify FileOnOff RegList PathName

The RegList attribute cannot be overridden.

SupportedActions

The SupportedActions (string-keylist) attribute lists all possible actions defined for an agent, including those defined by the agent developer. The engine validates the action_token value specified in the hares -action resource action_token command against the SupportedActions attribute. It is the responsibility of the agent developer to initialize the SupportedActions attribute in the resource type definition and update the definition for each new action added to the action entry point code or script. See action. This attribute serves as a reference for users of the command line or the graphical user interface.

An example definition of a resource type may resemble:


Type DBResource (
  static str ArgList[] = { Sid, Owner, Home, User, Pwork, StartOpt, 
   ShutOpt }
  static keylist SupportedActions = { VRTS_GetRunningServices,
    DBRestrict, DBUndoRestrict, DBSuspend, DBResume }
  str Sid
  str Owner
  str Home
  str User
  str Pword
  str StartOpt
  str ShutOpt

In the SupportedActions attribute definition above, VRTS_GetRunningServices is a VERITAS predefined action, and the actions following it are defined by the developer. The SupportedActions attribute value cannot be overridden.

ToleranceLimit

A non-zero ToleranceLimit allows the monitor entry point to return OFFLINE several times before the ONLINE resource is declared FAULTED. If the monitor entry point reports OFFLINE more times than the number set in ToleranceLimit, the resource is declared FAULTED. However, if the resource remains online for the interval designated in ConfInterval, any earlier reports of OFFLINE are not counted against ToleranceLimit. Default is 0. The ToleranceLimit attribute value can be overridden.

 ^ Return to Top Previous  |  Next  >  
Product: Cluster Server Guides  
Manual: Cluster Server 4.1 Agent Developer's Guide  
VERITAS Software Corporation
www.veritas.com