ResourceAdvisorConfig.properties file
The parameters in this file affect the following dynamic workload broker server settings:
- Heartbeat signal from the agents
- Time interval for job allocation to resources
- Time interval for notifications on resources
- Polling time when checking the status of remote engine workstations
- Maximum number of results when allocating jobs to global resources
You can modify the following parameters in the ResourceAdvisorConfig.properties file:
- DatabaseCheckInterval
- Specifies the time interval within which the dynamic workload broker server checks the availability of the database. The default value is 180 seconds.
- ResourceAdvisorURL
- Specifies the URL of the Resource Advisor.
- RaaHeartBeatInterval
- Specifies the time interval within which the Resource Advisor expects a heartbeat signal from the dynamic agent. The default value is 200 seconds. After the maximum number of retries (specified in the MissedHeartBeatCount parameter) is exceeded, the Resource Advisor reports the related computer as unavailable. In a slow network, you might want to set this parameter to a higher value. However, defining a higher value might delay the updates on the availability status of computer systems. If, instead, you decrease this value together with the value defined for the NotifyToResourceAdvisorPeriodSeconds parameter, this might generate network traffic and increase CPU usage when updating cached data. The value defined in this parameter must be consistent with the NotifyToResourceAdvisorPeriodSeconds parameter defined in the JobManager.ini file, which defines the time interval for each dynamic agent to send the heartbeat signal to the Resource Advisor.
- MissedHeartBeatCount
- Specifies the number of missed heartbeat signals after which the computer is listed as not available. The default value is 2. In a slow network, you might want to set this parameter to a higher value.
- MaxWaitingTime
- Specifies the maximum time interval that a job must wait for a resource to become available. If the interval expires before a resource becomes available, the job status changes to Resource Allocation Failure. The default value is 600 seconds. You can override this value for each specific job by using the Maximum Resource Waiting Time parameter defined in the Job Brokering Definition Console. For more information about the Maximum Resource Waiting Time parameter, see the Job Brokering Definition Console online help. If you set this parameter to -1, no waiting interval is applied for the jobs. If you set this parameter to 0, the Resource Advisor tries once to find the matching resources and, if it does not find any resource, the job changes to the ALLOCATION FAILED status. If you increase this value, all submitted jobs remain in WAITING status for a longer time and the Resource Advisor tries to find matching resources according to the value defined for the CheckInterval parameter.
- CheckInterval
- Specifies how long the Resource Advisor waits before retrying to find matching resources for a job that did not find any resource in the previous time slot. The default value is 60 seconds.
- TimeSlotLength
- Specifies the time slot interval during which the Resource Advisor allocates resources to each job. Jobs submitted after this interval has expired are considered in a new time slot. The default value is 15 seconds. The default value is adequate for most environments and should not be modified. Setting this parameter to a higher value, causes the Resource Advisor to assign resources to higher priority jobs rather than to lower priority jobs when all jobs are trying to obtain the same resource. It might also, however, cause the job resource matching processing to take longer and the resource state updates from agents to be slowed down. Setting this parameter to a lower value, causes the Resource Advisor to process the resource matching faster and, if you have a high number of agents with frequent updates, to update the resource repository immediately. If job requirements match many resources, lower values ensure a better load balancing. If most jobs use resource allocation, do not lower this value because the allocation evaluation requires many processing resources.
- NotifyTimeInterval
- Specifies the interval within which the Resource Advisor retries to send notifications on the job status to the Job Dispatcher after a notification failed. The default value is 15 seconds. The default value is adequate for most environments and should not be modified.
- MaxNotificationCount
- Specifies the maximum number of attempts for the Resource Advisor to send notifications to the Job Dispatcher. The default value is 100. The default value is adequate for most environments and should not be modified.
- ServersCacheRefreshInterval
-
Specifies with what frequency (in seconds) the Resource Advisor checks the list of active and backup dynamic workload broker servers for updates. This list is initially created when the master domain manager is installed, and after that it is updated every time a new backup master is installed and connected to the master domain manager database (the master domain manager and every backup master include also a dynamic workload broker server). When the Resource Advisor agents send their data about the resources discovered in each computer, they are able to automatically switch between the servers of this list, so that the dynamic workload broker server that is currently active can store this data in its Resource Repository. For this reason, the Resource Advisor agents must know at all times the list of all dynamic workload broker servers. The possible values range between 300 (5 minutes) and 43200 (12 hours). The default value is 600 seconds.
- StatusCheckInterval
- Specifies the time interval in seconds the Resource Advisor waits before polling for the status of a resource. For example this timeout applies when checking the status of a remote engine. The default value is 120 seconds.