WLM service class resources history attributes
These attributes represent the Workload Manager (WLM) service or report class information for the selected service or report class period over the recent historical period. There is one row of data per an RMF Monitor III reporting interval or group of RMF Monitor III intervals.
Active Time Active Time in Seconds to Microsecond precision.
Actual Velocity/Percentile (RMF metric) Service or Report Class Period Actual Velocity (If SUMEVG returned). If Goal Type is Velocity this is the velocity percentile value achieved. Otherwise this column is blank.
Average Execution Time (RMF metric) Service or Report Class Period Average Execution Time. For CICS transactions, this includes execution time in AOR and following regions.
For IMS transactions, this includes execution time within the MPR.
For Batch, TSO, and so on, this is the average time that transactions spent in execution.
In the Average Response Time columns, zeros will show up for server classes in most cases, because theirtransactions are address spaces, and response times are available only for ended transactions. So there are only numbers, when one of the address spaces in that service or report class ends, or is RESET via operator command.
Average Response Time (RMF metric) Service or Report Class Period Actual Response Time in milliseconds to 1 decimal place (tenths of milliseconds).
Avg Res Time (secs) Average Response Time in Seconds to Microsecond precision.
- Queued - Average time a job was delayed for reasons other than the ones mentioned below. This field therefore basically includes the time a job was delayed for initiation. For TSO users, this can be a portion of LOGON processing. For APPC, this is the time the transaction spent on an APPC queue.
- R/S Affinity (Resource affinity scheduling delay) - Average time the job was delayed due to resource or system affinity scheduling. This means that resource(s) required for the job to run were not available at some point while the job was queued to JES2.
- Ineligible (Operational or JES scheduling delay) - Average time a job was delayed due to
operational delays or JES scheduling delays, examples are:
- Job held by operator
- Job class or job queue held
- Duplicate jobname serialization
- Job class execution limits
- Conversion (JCL conversion delay) - Average time a job was delayed for JCL conversion. Jobs held during conversion (due to affinity, HSM recall, or enqueue contention) contribute only to conversion time, not to ineligible or R/S affinity times. Conversion time is not part of the total response time.
The time a job was delayed due to TYPRUN=HOLD or TYPRUN=JCLHOLD is NOT included in any of the transaction times.
In all other cases, this is the average time that transactions spent waiting on a JES or APPC queue. Also note that queue time may not always be meaningful, depending on how the customer schedules work. For example, if a customer submits jobs in hold status and leaves them until they are ready to be run, all of the held time counts as queued time. That time may or may not represent a delay to the job.
In the Average Response Time columns, zeros will show up for server classes in most cases, because theirtransactions are address spaces, and response times are available only for ended transactions. So there are only numbers, when one of the address spaces in that service or report class ends, or is RESET via operator command.
- Unavailable=' '
- Report=R
- Service=S
- Unavailable=0xFF
- S=0x10
- C=0x20
- SC=0x40
- Unavailable=0xFF
- SU/sec=0x10
- %LPAR=0x20
- #CPs=0x40
- MSU=0x80
Duration (RMF metric) Service or Report Class Period Goal Duration. Service or report class period duration in unweighted CPU service units (that means, not multiplied with the service coefficients) per second.
A duration is required in all but the last service or report class period. For single periods or for the last period of multiple periods this value is always zero.
Goal Importance (RMF metric) Service or Report Class Period Importance.
- 1 - Highest
- 2 - High
- 3 - Medium
- 4 - Low
- 5 - Lowest
- D - Discretionary
Goal Percentile (RMF metric) Service or Report Class Period Percentile Goal If Goal Type is Percentile Response Time this is the percent of transactions that should meet the Response Goal value desired. Otherwise this column is zero.
- Unavailable
- Highest
- High
- Medium
- Low
- Lowest
Goal Type (RMF metric) Service or Report Class Period Goal Type. The type of goal for the service or report class period. Valid values:
- System
- Discretionary
- Velocity
- Average Response Time
- Percentile Response Time
Goal Value (RMF metric) (RMF metric) Service or Report Class Period Response Time Goal. If Goal Type is either Average Response Time or Percentile Response Time this is the response time value desired. Otherwise this column is blank.
Goal Value Response time goal Value in Seconds to Microsecond precision.
- Unknown==0xFE
- Unavailable=0xFF
- Yes=0x10
- No=0x20
- Unavailable=0xFF
- Yes=0x10
- No=0x20
Ineligible Queue Time Ineligible Queue Time in Seconds to Microsecond precision.
JCL Conversion Time JCL Conversion Time in seconds to Microsecond precision.
Managed System The name of the managed system from which this data is collected. The name has the format <plexname>:<smfid>:MVSSYS, where <plexname> is the Sysplex this LPAR resides in and <smfid> is the unique identifier for the system.
Max Memory Limit GB Maximum Memory Limit GB .
Max Service Rate Capacity Maximum Service Rate Capacity.
Memory Used Actual Memory Used Actual.
Min Service Rate Capacity Minimum Service Rate Capacity.
- Execution velocity goal
For velocity goal thePerf Index = Goal % / Actual %
Actual %
is:
Using Samples is the number of samples categorized as using CPU or active in I/O. All other sample categories like waiting for CPU or waiting for I/O are considered Delay Samples.Actual(%)= (# Using Samples / (# Using Samples + # Delay Samples)) * 100
- Average or Percentile response time
goal
For average response time goal the Actual (sec) value is the average response time over all transactions in the sampling period.Perf Index = Actual (sec) / Goal (sec)
For the percentile goal there are 14 buckets maintained by WLM that are based on the goal response time. The buckets are:Bucket Percentage of Goal Example for 2 second goal 1 <50% <1.0 2 60% >1.0 to 1.2 sec 3 70% >1.2 to 1.4 sec 4 80% >1.4 to 1.6 sec 5 90% >1.6 to 1.8 sec 6 100% >1.8 to 2.0 sec 7 110% >2.0 to 2.2 sec 8 120% >2.2 to 2.4 sec 9 130% >2.4 to 2.6 sec 10 140% >2.6 to 2.8 sec 11 150% >2.8 to 3.0 sec 12 200% >3.0 to 4.0 sec 13 400% >4.0 to 8.0 sec 14 >400% >8.0 sec
The target response time for each bucket is then calculated based on the goal response time. As transactions complete they are counted in a bucket based on their actual response time. The Actual (sec) value for the Performance Index calculation is derived by multiplying the goal percent times the actual total transaction count to get the target transaction count and then summing all the transaction in the buckets from bucket 1 upward until the total is just greater than the target transaction count. The response time associated with this bucket is then declared the Actual (sec) value.
SUMTYP
is
in the range of 1 - 8).. Type of group being reported on:- W Workload group
- S Service class
- R Report class
- n Service/report class period
Queued Time Queued Time in Seconds to Microsecond precision.
Recording End Time Time at the end of the RMF reporting interval for the current row of near-term historical data. The valid format is yy/mm/dd hh:mm:ss.
Recording Start Time Time at the start of the RMF reporting interval for the current row of near-term historical data. The valid format is yy/mm/dd hh:mm:ss.
Recording Read Time The time interval required to read the interval data in the current row. The valid format is yy/mm/dd hh:mm:ss.
Resource Group Name (RMF metric) The resource group name associated with this service or report class. You can assign only one resource group per service or report class in a service policy. You can override the resource group assigned to a service or report class in each service policy.
Response Time Actual Total Response Time Actual Total in Seconds to Microsecond precision.
RMF Status Status indications for availability of RMF and lock data.
R/S Affinity Time R/S Affinity Time in Seconds to Microsecond precision.
Sample Period Sample period in seconds.
Service Class The service or report class name. Valid value is a string, with a maximum of eight characters. This field displays unavailable if the system is in compatibility mode. If a job changes its service or report class during the report interval, RMF displays eight asterisks (********) instead of the service or report class name. If the service or report class is not available, RMF displays eight dashes (--------).
Service Class Description (RMF metric) An area of 32 characters describing the service or report class. The descriptive text is available to performance monitors for reporting.
Service Class SUs/Second (RMF metric) Transactions per second that ended during the interval.
- Unavailable=0xFF
- No=0x00
- Yes=0x01
- Unavailable=0xFF
- Yes=0x10
- No=0x20
Transaction Rate (RMF metric) The transactions per second that ended during the current interval.
Wait Time Transaction wait time in Seconds to Microsecond precision.
Workload Name The name of the workload that contains the service or report class. Valid value is a simple text string of from 1 through 8 characters; for example, BATAVGWTIMECH.