Workload Analyzer subreports: Job Count, Size, Queue, and Throughput
The Workload Analyzer reports are based on time, queue, size, and throughput. The following reports are a part of the Workload Analyzer report:
: This report is based on the number of jobs running at a given period of time. The purpose of this report is to display the data that is based on time and load. You can determine the time for the job count on the basis of the start and end time of the job. Data for load is determined based on how many jobs are active during a given period of time. Calculations are done to determine the number of jobs that are active during a given period of time.
For example, the following figure displays the job counts for a given period of time. (A) shows the jobs that start before the time interval and end after the time interval. (B) shows the jobs that start before the time interval and end during the time interval. (C) shows the jobs that start and end during the time interval. (D) shows the jobs that start during the time interval and end after the time interval.
: This report is based on the amount of data that is transferred during a given period of time. The calculation is based on the total amount of data that is backed up by the jobs divided by the time period.
For example, in the following figure there are four jobs that have different durations across a four hour time period. The left-hand side displays the amount of data that's transferred during the four hour time period. The right-hand side of the report displays the start time and end time for each job, amount of data transferred, and duration. Job A is active for a period of four hours, it starts at 1:20 P.M. and ends at 4:20 P.M. During that period the total data that is transferred is 80,000 MB. You can calculate data transfer rate, multiply data transfer rate by the number of minutes during each hour when there is job activity. Job activity in the first hour is 20 minutes, in the second and third hour it is 60 minutes, and fourth hour it is 20 minutes. Jobs B, C, and D follow a similar pattern. Load is the sum of loads of all jobs during that hour. The yellow slot shows maximum activity and the data that is transferred is 115960 KB.
: This report is based on the period for which, jobs are in a queue state before data back up begins. Queue time is calculated through the use of timestamps such as time taken to initiate the job, time taken to start and end the job. Jobs that have a high queue time are identified to reduce scenarios where many jobs with high queue time are initiated at the same time.
The queuing is calculated at individual job level and then aggregated to the level where the report is generated.
: This report is based on the rate in Mbytes\Sec at which the data is transferred This calculation provides an important indicator to understand the performance. The throughput report is based on cross-hour calculation. The report displays accurate analysis of the data. Compared to the other workload reports, the totals for the rows of this report are averaged.
The Start Write time ensures that the speed is only measured against the time in which, the actual data transfer takes place. This calculation ensures that the queue time is not considered. The queue report calculates the queue time. In the following figure, the throughput is calculated across four jobs that span for a period of three hours. One job belongs to Policy B and three jobs belong to Policy A. The green boxes indicate the time in which the write job occurs.