DEVR - Device Resource Delays Report

The Device Resource Delays report (DEVR) shows the devices (volumes) and the jobs using or being delayed by them (as indicated on the Device Delays report).

On the DEVR report, the type of delay is listed under Pend Reasons as:

DB
Device busy delay
CMR
Initial command response time

How to request this report

To request the DEVR report, select 3 from the Primary menu, and then select 3 on the Resource Report Selection Menu (shown in Figure 1) or enter the following command: 
DEVR [volser]

Contents of the report

Figure 1. DEVR Report
                        RMF V2R4  Device Resource Delays        Line 1 of 374
Command ===>                                                  Scroll ===> HALF

Samples: 100     System: MVS1  Date: 05/11/19  Time: 10.03.20  Range: 100   Sec


 Volume S/   Act  Resp  ACT CON DSC PND %,  DEV/CU              Service  USG DLY
   /Num PAV  Rate Time   %   %   %  Reasons Type     Jobname  C Class     %   %

 160444 S    17.5  1.2    2   2   0 PND   0 33909    GPMSRVPM S GPMSERVE   0   1
  08803   1.2H                              2105     RMF      S SYSSTC     2   0
 SYSUSR S     0.3  0.4    0   0   0 PND   0 33903    *MASTER* S SYSTEM     1   0
  0D440   2*                                2105     CATALOG  S SYSTEM     1   0
 MVSTGT S     0.8  0.9    0   0   0 PND   0 33909    GPMSRVPM S GPMSERVE   1   0
  07790                                     2105
 SYSXCP S     1.4  0.7    1   1   0 PND   0 33903    XCFAS    S SYSTEM     1   0
  0D32A                                     2105

The first block of columns in the report contain information related to each volume.

The remaining columns contain information related to each job. RMF sorts the volumes in descending order according to the average number of delayed users (1 user delayed 100% is equivalent to 100 users each delayed 1% of the time), and the waiting jobs by descending delay percentages. If RMF is unable to obtain valid hardware data, it prints dashes (---) for the hardware measurements, while percentages normally appear.

The fields DLY DB % and DLY CU % contain data about I/O request delays caused by contention at the control unit and device level.

The graphic DEVR report shows the average number of active users for each device that were delayed, connected, disconnected or pending.

Field descriptions

Table 1. Fields in the DEVR Report
Field Heading Meaning
Volume/Num The name of an online volume and the device number where the volume is mounted. The first digit of the device number represents the ID of the subchannel to which the I/O device is physically configured.
S/PAV
S
An S in the first line of this column indicates that the device was generated during system generation as a shared device.
PAV
PAV count — A value in the second line of this column gives the number of parallel access volumes (base and alias) which were available at the end of the reporting.

If the number has been changed during the report interval, it is followed by an '*'.

If the device is a HyperPAV base device, the number is followed by an 'H', for example, 1.2H. The value is the average number of HyperPAV volumes (base and alias) for that range. 
                                Accumulated # of HPAV devices
 Average # of HPAV devices =  ---------------------------------
                                     Number of Samples
Act Rate The rate per second that I/O instructions (SSCH, RSCH, and HSCH) to a device completed successfully.
The calculation is: 
           # I/O Instructions
Act Rate = ------------------
               Range Time
Resp Time The average response time (in milliseconds) that the device required to complete an I/O request.
The calculation is: 
                Active Time
Resp Time = ------------------ + IOS Queue Time
            # I/O Instructions
ACT % The percentage of time during the report interval when the device was active. To derive this value, RMF computes the accumulated percent active time as follows: 
ACT % = PEND % + CON % + DSC %
PEND %
Percentage of time all I/O requests wait in the logical control unit queue (CU-HDR) before there is an available path. Pending time includes the time spent waiting for a channel, control unit, or head of string, or for the actual device (if it is a shared device that is reserved by another processor).
CON %
Percentage of time the device was connected to a channel path to actually transfer data between the device and storage.
DSC %
Percentage of time the device has an active channel program and is disconnected (not transferring data). Disconnect time includes seek time, normal rotation delay time, and extra rotation delay because the channel was busy when the device needed to reconnect.
CON % The percent connect time. See the description under % ACT. RMF calculates the value as follows: 
        Accumulated Connect Time
CON % = ------------------------ * 100
               Range Time
DSC % The percent disconnect time. See the description under %ACT. RMF calculates the value as follows: 
        Accumulated Disconnect Time
DSC % = --------------------------- * 100
              Range Time
Note:
  1. When comparing the ACT %, CON %, DSC %, or PND % fields with the USG % field in this report, you must be aware that ACT %, PND %, CON %, and DSC % are measured multi-state values, while USG % is a sampled single state value. If a single I/O request is very long (such as a long-running channel program), the PND %, CON %, and DSC % values might be too low because of timer overflow errors.
  2. The channel updates the data fields used to calculate CON %, DSC %, and PND % when the I/O operation completes. Therefore, some of the time from the previous report interval might be included in these values, while some of the time in the current report interval might be absent from these values. This discrepancy is noticeable on devices that have very long channel programs, such as paging devices.
PND % Reasons
The first entry is always the pending percentage (PND). See the description under % ACT. RMF calculates the value as follows: 
        Accumulated Pending Time
PND % = ------------------------ * 100
              Range Time

DLY DB % and DLY CU % are included in pending time.

Below PND % are the pend reasons that contribute to the total pending percentage. A value appears only when there is a non-zero delay percentage. Pend Reasons can be one of the following:

DB
Device busy delay, which is the percentage of time during the report interval when the channel subsystem measured I/O request delay because the device was busy. Device busy might mean that the volume is in use by another system, the device is reserved by another system, a head of string busy condition caused the contention, or some combination of these conditions has occurred. 
          Accumulated DB Delay Time
DLY DB% = -------------------------  * 100
                Range Time
CMR
Command response time delay, which is the percentage of time during the report interval when the first command of an I/O instruction of the channel program is sent to the device, until the device indicates it has accepted the command. 
           Accumulated CMR Delay Time
DLY CMR% = -------------------------- * 100
                  Range Time
Note: If either hardware data or volume-related percentages are not available, this field is blank.
DEV/CU Type The top number represents the device type. The bottom number represents the control unit model.
Jobname Name of a job using or being delayed by the DEV volume. The DEVR delay report does not summarize data by job groups; all jobs within a job group are reported individually. RMF lists all jobs for each device, by descending delay percentages.
C A one-character abbreviation for the job class as follows:
S
Started task
T
TSO
B
Batch
A
ASCH
O
OMVS
Service Class The name of the service class that a specified job has been running in.
USG % The percentage of time when the job has had a request accepted by the channel for the specified Volume, but the request is not yet complete.
DLY % Delay the waiting job (address space) is experiencing because of contention for a specific volume during the report interval, expressed as a percentage.

Monitor III Utility fields

Table 2 shows additional fields for the Device Resource Delay report.

You can use the Monitor III Utility to customize the DEVR report.

Table 2. Additional Fields in the DEVR Report
Field Heading Meaning
Percentage of pending time The percentage of time during the report interval when the device was pending.
IOS queue time The average number of milliseconds an I/O request must wait on an IOS queue before a SSCH instruction can be issued. Delay occurs when a previous request to the same subchannel is in progress.
Percentage of device busy delay The percentage of time during the report interval when the channel subsystem measured I/O request delay because the device was busy. Device busy might mean that the volume is in use by another
Percentage of control unit busy delay time The percentage of time during the report interval when there is I/O request delay because the control unit was busy.
Percentage of switch port busy delay time The percentage of time during the report interval when there is I/O request delay because the switch port was busy.

Report options

You can use the DEVR Report Options panel to select the volume to be included in the DEVR report, or all volumes, from a list of available volumes.

Figure 2. DEVR Report Options Panel
                             RMF DEVR Report Options              Line 1 of 17
Command ===>                                                   Scroll ===> HALF

 Change or verify parameters for the DEVR report. To exit press END.

      VOLSER ===> ALL         Volume to be reported or ALL

                               Available Volumes
BASECA     BASECB     BASEC0     BASEC1     BASEC2     BASEJM     BERDPK
BSS210     BSS999     CAT212     CAT84I     CB8480     CHKPTX     CHKPT1
CHKI01     CHKPT3     CHKPT4     CKSPL1     CKSPL2     CKSPL3     CLRPAK
C45I01     DASD03     DBLMN1     DBLMN2     DBLMN3     DBVL01     DBVL02
D60AF3     D60AF5     D60AF6     D60AF7     D60PAK     D602A2     D602A3
D602A4     D602B1     D602B2     D602B2     D602DC     D602D5     D602D6
D602D7     D6021B     D6021D     D6021D     D60215     D60217     D60411
D60412     D6044D     D60444     D60444     D60445     D60446     D606A8
D607AF     D607EC     D607E1     D60701     D60707     D608FO     D83CL3
D71CKP     D71CMN     D71SPP     D71WLD     D83CL1     D83CL2     D83D31
D83CMN     D83DB1     D83DB2     D83DMP     D83DRA     D83DRB     D83D31
D83D51     D83ILG     D83I80     D83I81     D83I90     D83I91     D83I92
D83I93     D83JCK     D83JC1     D83JC2     D83JD1     D83JD2     D83JFY
D83JS1     D83JS2     D83JVC     D83JV2     D83JV3     D83J22     D83J23
D83J31     D83J32     D83LOG     D83PAK     D83SPP     D83STA     D83STB
VOLSER
The volume serial number of the device that you want information about.
Enter
  • ALL for information about all devices that have jobs using it or being delayed by it in the system.
  • A name with an asterisk ('*') as a "wild card" character. For example: to request a report for all volumes starting with D8, specify 'D8*' for VOLSER.
    Note: You cannot use the wild card when calling the report, that is, when you use the command DEVR volser. Here, volser must be a complete volume serial number, an asterisk will be interpreted as part of the volser.
  • One of the volumes listed under Available Volumes.
  • The volume serial number of a device that will be in the system at a later time.

If the volume that you specify is not currently available, it will appear on the report when it is available.

Your selection is saved across sessions in the current option set.

Available Volumes
The list of the online volumes in the system.

If the volume you want is not listed, it was not online during the current report interval. If you specify the volume, it will appear on the report when it is online.