Dialog Processing & Lock Management

After defining a lock object in the data dictionary, a developer asks where to write the lock/unlock code. What do you tell them?

Your lock vanishes the moment the dialog COMMITs, before the update task runs. Which _SCOPE value did you (wrongly) use, and which fixes it?

Which OS command tests the message-server connection and lists available instances and logon groups?

A colleague says dynamic work processes spin up whenever load rises and shut down when it drops. What's the actual trigger?

A trace shows one multi-screen transaction touched three different dialog work processes. Is something broken?

You need to audit which locks are currently held. Why can't you just SELECT from a lock table, and where do you actually look?

What do the four ACID properties guarantee for an SAP LUW?

On a current system you hit lock-table-full and reach for enque/table_size. Why is that the wrong knob now?

A developer added logging inside the generated DEQUEUE_ module for a lock object and it worked in test. Why will it silently stop working, and how do you recognize these objects?

Priority-based runtime caps abort tasks past their limit. Which request type maps to high, normal, and low priority?

A developer assumes user context stays loaded in the same work process between dialog steps, so reloads are free. Why is that assumption wrong?

In SM50 some work processes show status 'standby'. Should you count them toward your usable capacity?

enque/process_location defaults to REMOTESA. When troubleshooting locks, which instance does that tell you to look at?

Your dispatcher process dies but the work processes are still up. Is the instance still usable?

After an enqueue server restart, which locks come back and which are gone for good?

Users report lock failures but you can't tell if the enqueue service itself is alive — maybe GUI access is the problem. What's the next-best check?

To save memory on a small instance, can you run a single dialog work process?

The enqueue server crashes for a moment. Is this a harmless blip users can retry through?

An app server is overloaded mid-day; will the message server's load balancing migrate its logged-in users to a quieter server?

Before forcibly deleting a user's lock, you confirm they've logged off. Why isn't that enough, and what else must you check?

Interactive users complain of slowness exactly when a big batch run kicks off. How do logon groups prevent this, and where do you set them up?

A lock has sat in SM12 for a while and a user wants it cleared. Why is deleting it not a routine cleanup action?

An inherited profile still sets rdisp/max_wprun_time. On S/4HANA 2020+, why must it go, and what does it sabotage if left?

A read-then-display screen and a posting transaction both need the same record. Which lock mode does each take, and where do people get it backwards?

The enqueue server is a classic single point of failure whose loss can cause data loss. On the modern ABAP platform, which components eliminate that SPOF and how are locks protected?

Dialog work processes can spawn other dialog work processes and deadlock when all are held. What two-tier work-process concept does S/4HANA use to break out of that, and what can stand-by processes NOT do?

An interactive SAP GUI dialog step keeps running and eventually gets aborted after about 11 minutes. Which profile parameter enforces that limit on the modern S/4HANA scheduler, and what does it imply about where long-running work belongs?

You are tempted to size the enqueue work processes with rdisp/wp_no_enq. Why does SAP tell you to leave that parameter alone, and what should provide the lock server instead?