provide Display SHCDS command with results that go to SYSLOG for automation.

We discussed this further and these reports are not suitable for the operlog.

Alternatively here are the operator commands can be issued with shunted commands and urids for outstanding recovery.
https://www.ibm.com/docs/en/zos/2.5.0?topic=command-displaying-storage-management-subsystem-information-sms
These commands can be used for that purpose: [,SHUNTED,{SPHERE(sphere)|UR({urid|ALL}}{,WTOR} ] and [,URID({urid|ALL}){,WTOR} ]

RLS also recently closed apar OA62895 which lists enables users of Transactional VSAM (TVS) to know the 5 data sets with the greatest number of record locks held for a specific unit of recovery when message IGW10074I is issued.

Description:
Message (IGW10074I) enhancements for
improved OLTP monitoring.

This enhancement enables users of Transactional VSAM (TVS) to
know the 5 data sets with the greatest number of record locks
held for a specific unit of recovery when message IGW10074I is
issued due to the MAXLOCKS threshold being reached. MAXLOCKS is
an optional PARMLIB keyword in IGDSMSxx and is intended to be
used to ensure efficient resource utilization and to prevent
excessive lock contention for RLS/TVS requests. Previously
customers would see an IGW10074I message indicating the unit of
recovery that went above the threshold and how many record
locks were held but no reference to which data sets the unit of
recovery was accessing. With this enhancement the IGW10074I
message is enhanced to display the 5 data sets with the highest
number of records locks held for the unit of recovery,
the number of record locks and how many data sets in total
were not displayed. This will allow customers to better monitor
and identify application anomalies and issues.

As we understand it, this RFE is a request to ?mask? the output of the SHCDS LISTSUBSYS(ALL) command, or something equivalent, to only display subsystems with lost or retained locks.

It is a potential roadmap item but we don't have a release/date to provide at this time

.As we understand it, this RFE is a request is to ?mask? the output of the SHCDS LISTSUBSYS(ALL) command, or something equivalent, to only display subsystems with lost or retained locks.

It is a potential roadmap item but we don't have a release/date to provide at this time

Based on documentation:

RLS and TVS Recovery Guidelines Terri Menendez 02/28/2024

Part 5. Transactional Recovery

1. A. Retained Locks

Retained locks occur when active locks are held by recoverable subsystems (i.e. CICS or a TVS instance) for recoverable data sets (i.e. LOG(UNDO/ALL)) and one of the following occurs:

1. Recoverable data set is closed without ending all active transactions (i.e. cancel CICS).

2. SMSVSAM terminates with active locks.

3. System failure with active locks.

4. Backout fails (shunted) for one or more transactions holding active locks.

When a data set has retained locks, record management requests for the same record locks by RLS sharers will fail with RC=8 RPLERRCD=24(x’18’).

To determine which subsystem have retained locks issue:

SHCDS LISTSUBSYS(ALL)

SHCDS LISTSUBSYSDS(subsys)

SHCDS LISTDS(baseclustername)

Based on this RFE description:
SHCDS commands  work both in batch and in the TSO/E foreground, but the output isn't available in SYSLOG.
Automation around SMSVSAM/RLS issues would be easier if we could drive standard z/os commands such as /D SHCDS xxx and have output go to SYSLOG.

my immediate comment is that you could add automation in your applications for the SHCDS message (IGW6xxx) from the SHCDS commands.  However, like you said, "Running a batch job isn't very efficient when trying to automate based on messages." The main problem with SHCDS commands going to operlog is that they are generally voluminous and not suited for the console.  Could you tell us what specific SHCDS command outputs you are looking for, so maybe we can come up with a more specific solution.