|
Measures reported by P6EPPMJbSrvcTest
The Job Service process run as standalone for better performance and scalability. The Job Service can either run on the same server or a separate server or cluster. The Job Service is responsible for executing scheduled and real time application Jobs. Some of the key jobs processed by Job Service are Summarizer, scheduler, leveler, publications, updating baselines, adding/creating baselines, export-import project etc.
The Job service is capable of processing large number of projects, activities and resource assignments. The number of concurrent jobs affect the CPU and memory requirement of the Job Server and the load on database server. That's the reason it is important for administrators to be aware of job status, and server performance while the jobs are running. With these insights administrators can take necessary action to manage the jobs and increase server capacity if required.
This test monitors the Job Server to measusre key paramaters like memory utilization, number of jobs executing etc., to understand the performance of job execution and load on the server.
Outputs of the test: One set of results for P6 EPPM Server
The measures made by this test are as follows:
| Measurement |
Description |
Measurement Unit |
Interpretation |
| usedMemory |
Indicates the amount of memory that is being used by Job Service. |
MB |
If the memory usage is really high over multiple measurements, it could eventually lead to memory full resulting in Job Failure.
|
| maxMemory |
Indicates the maximum memory available on the server for Job Service. |
MB |
If memory usage is repeatedly reaching max memory, the administrator may need to kill some jobs to relieve the load on the system.
|
| jobsRunningThisServer |
Indicates the number of jobs running on the server. |
Number |
You can check if some jobs are running for too long, they may be in stalled state, and may need to be killed to make resources available for waiting jobs to execute.
|
| jobsPendingSystemWide |
Indicates the number of jobs pending for Job Service thread. |
Number |
|
| jobsExecutions |
Indicates the number of times a job has executed successfully. |
Number |
If the job has executed successfully a few times before, the probability of it running successfully will be high.
|
| jobsExecutedInLastHour |
Indicates the number of jobs finished in last hour. |
Number |
Signifies the rate at which jobs are executing. If the number is too low, the jobs may need to be broken down.
|
| jobsFailedInLastHour |
Indicates the number of jobs failed in last hour. |
Number |
If there are too many job failures, there may be an issues in system performance.
|
| pctJobsFailedLastHour |
Indicates the percentage of jobs failing in last hour against total number of jobs started. |
Percentage |
|
| avgExecutionTime |
Indicates the average time taken by a job to complete. |
Seconds |
You should look for trend in this data. If the average job execution time is increasing over a range of measurements, it might be an indicator of deteriorating performance of the server.
|
|