Execution state summary

The DEXAN wait reasons fall into seven major categories.

  • Using CPU
  • Swap reason waits
  • Miscellaneous waits
  • Logical channel waits
  • Control unit
  • Device waits
  • Enqueues

Each of these categories is discussed in turn. The codes shown are for the PCT displays.

Using CPU
Using CPU for a job is always shown as the first item in the display.
Swap reason waits

Table 1 lists the SRM swap reasons.

Table 1. List of SRM swap reasons
Reason Code
Term-Out TOU
Term-In TIN
Long-Wat LON
Aux-Strg AUX
Real-Str RST
Det-Wait DET
Request REQ
Enq-Xchg EEX
Exchange EXC
Unilatrl UNI
Transwap TSW
Cent-Str CSS
Sys-Page SPS
Too-Long OLS
APPC-Wt APS

Advanced Program-to-Program Communication (APPC) enables applications to use SNA LU6.2 to communicate between transaction programs (TPs) running in an SNA environment, either on the same system or on different systems. APPC transaction programs can use z/OS® services, such as dataspaces, in their conversations.

Miscellaneous waits

Table 2 describes the miscellaneous wait reasons.

Table 2. List of miscellaneous wait reasons
Reason Description Code
HSM Backup Dataset The address space is waiting for HSM to execute a data set backup request. BKP
JES Job Cancel The address space has issued a cancel request to JES2. CAN
HSM Read JES3 C/I The address space is waiting for HSM to execute a C/I locate request. CIL
Common Page-In Wait The address space is waiting for a PLPA or Common page-in. COM
Waiting for CPU The address space is waiting on the active CPU dispatching queue. High percentages here may mean the address space is low priority, or perhaps a shortage of CPU cycles. CPW
JES Delete Request The address space has issued a job delete request to JES2. DEL
HSM Delete Dataset The address space is waiting for HSM to delete a data set. DLD
Disk Mount Pending The address space is waiting for a disk to be mounted by the operator. DMP
ECB Wait The address space has issued a voluntary wait. Examples are an IMS message region waiting for work, a CICS® region waiting for work, or OMEGAMON waiting between screen refreshes. High percentages of ECB waits may indicate a program error or normal voluntary waiting as previously discussed. ECB
ECB Wait with STIMER The address space has issued a voluntary wait. Examples are an IMS message region waiting for work, a CICS region waiting for work, or OMEGAMON waiting between screen refreshes. An STIMER is outstanding in the address space along with the voluntary wait. ECS
HSM TSO HLIST The address space is waiting for HSM to execute a TSO HLIST command. HLS
JES Status Request The address space has issued a job status request to JES2. JST
Waiting for MVS Lock The address space is waiting to acquire a local or global z/OS lock. LCK
HSM Migrate Dataset The address space is waiting for HSM to migrate a data set. MIG
Waiting for Staging The address space is waiting for a mass storage volume to be staged. MSS
Private Page-In Wait The address space is waiting for a page-in operation. PAG
JES PROCESS SYSOUT The address space issued a process SYSOUT request to JES2. PSO
HSM Recall Dataset The address space is waiting for HSM to recall a data set. RCL
HSM Read Control Record The address space is waiting for HSM to read a control data set record. RCR
HSM Recover Dataset The address space is waiting for HSM to recover a backup data set. RCV
SRM Delay (MPL) The address space is swapped out and is ready to be swapped back in, but the SRM has not yet permitted the swap-in. This can occur when a domain is at the target MPL and the SRM has to either increase the target or swap out another address space in the domain. When SRM has scheduled a swap-in via ASM, the work shows up in SWAP PAGE-IN WAIT or in SWI state. .) RDY
JES2 Requeue Request The address space has issued a job requeue request to JES2. RQE
SRM Delay (RTO) SRM delayed the TSO transaction due to a RTO specification in the IPS. RTO
STIMER Wait The address space issued a STIMER and is voluntarily waiting for it to end. This is typically not considered degradation because it is a voluntary wait. STI
Swap Page-In Wait This is an ASM condition. The address space has been given to the ASM queue and is waiting for z/OS to swap it in to storage. An excessive value for this wait could be caused by a slow I/O device or DASD contention. SWI
Tape Mount Pending The address space is waiting for a tape to be mounted by the operator. TMP
Swapped with outstanding WTOR The address space has been swapped out due to Detected Wait, and the address space also has an outstanding WTOR. WTO
I/O device waits

Table 3 describes the control unit and device waits:

Table 3. List of device wait reasons
Reason Code
DISK volser aaa ACTIVE aaa
TAPE volser aaa ACTIVE aaa
CONTROL UNIT aaX QUEUED QaaX
DISK volser aaa QUEUED Qaaa
TAPE volser aaa QUEUED Qaaa
DISK volser aaa RESERVED Raaa

where aaa is the address of the device.

Active
An I/O is in progress to the device.
Queued
An I/O is queued waiting for the device to become available for use by this address space.
Reserved
An I/O is waiting because of a reserve or active I/O to this device from another CPU, or because of head-of-string contention from this CPU or another.

Logical Channel Waits apply only to z/OS systems.

Logical channel waits are abbreviated Lnn. The long form is LOGICAL CHANNEL nn QUEUED.

Enqueues

Table 4 lists the enqueues.

Table 4. List of enqueues
Reason Code
SYSDSN enqueue *DSN
SYSIEA01 enqueue *IEA
SYSIEFSD enqueue *IEF
SYSIGGV1 enqueue *IG1
SYSIGGV2 enqueue *IG2
SYSIKJBC enqueue *IKJ
SYSSMF01 enqueue *SMF
SYSIEWLP enqueue *IEW
SYSVSAM enqueue *VSA
SYSVTOC enqueue *VTO
SYSZVARY enqueue *ZVA
Misc SYS enqueue *SYS
Misc USR enqueue *USR

For definitions of these enqueues, see Volume 1 of the IBM® MVS/Extended Architecture Debugging Handbook.

The CENQ command controls enqueue analysis.

Note: By default, the enqueue analysis is turned off to conserve CPU resources, which may amount to as much as 5%. When it is turned off, enqueue waits are attributed to ECB waits. Thus, the ECB percentages increase.

Using the CENQnn format you can reduce the overhead of enqueue analysis and still capture data. Note that the number of cycles required for a significant count is also increased (e.g., 40 samples take 200 cycles with CENQ05). The DEXAN counts are also adjusted to account for the delayed sampling.