Lock Management (Enqueue)

On a classic single-instance ABAP system with no HA requirement, should you deploy a standalone enqueue server for the throughput boost?

Lock entries normally clear themselves — so why do stale entries pile up and block users, and what routine catches them?

You've found an old orphaned lock in SM12 and want to delete it — what two things must you verify first before removing it?

An instance has no ENQ work process — can you conclude it has no lock table either?

Under heavy load you want several enqueue work processes — why can't you spread them across different instances to balance the load?

SAP recommends running the enqueue server and message server on the same instance — what problem does splitting them onto different instances create?

A classic single-instance system (no (A)SCS) loses its enqueue server — what happens to in-flight business transactions, not just the locks?

Both E and X are exclusive lock modes — what can the lock owner do with E that it cannot do with X?

An optimistic (O) lock lets several users hold it at once, just like a shared lock — so what makes it different from S?

In the lock table, what do the Name and Argument fields identify, and what does it mean when the Argument contains wildcards?

A lock table Owner ID is more than just a user — what is it built from, and what second value travels with it?

The classic enqueue server barely scales past three CPUs — where is the bottleneck, and how does the standalone enqueue server get around it?

A single lock object covers several tables — does one lock mode apply to the whole object, or can it differ?

A developer sets a lock with _SCOPE=1 and assumes COMMIT WORK will clear it — what actually removes that dialog lock?

In SM12 Extras > Statistics, which counter tells you lock contention is happening, and what does a rising value mean?

What are the four SAP lock modes, and which one allows multiple concurrent holders?

Why must the enqueue replication server (ERS) sit on a different host than the enqueue server for HA to actually work?

You are looking for the instance running the enqueue service and an instance has no ENQ work processes — does that rule it out?

A lock request hangs and the message server is down — why can locking be affected even though the ENQ process is healthy?

The lock table is overflowing under heavy locking — which enqueue parameter do you raise, and which similarly-named ones would NOT help?

In SM12 you see both blue and black lock entries — which color is the dialog owner's lock, and why does the distinction matter before you delete one?

An admin assumes that after a lock-server restart every lock is restored — where do locks actually live, and which ones are silently lost?

On a pure central instance, is an ENQ work process strictly required to access the lock table — and when does that answer flip?

How does an SAP (enqueue) lock's lifetime differ from a database lock's, and which property lets it survive an enqueue server failure?

A developer says the lock object is what gets created and removed each time a user edits a record — what are they confusing, and which one is design-time?

Why must the standalone enqueue server (Kernel 6.40+) run on its own free instance number rather than share one?

On a current AS ABAP / S/4HANA system, what does SMENQ do that SM12 does not, and is SM12 gone?

You lock a record in dialog with ENQUEUE then post the change via update task — which _SCOPE value keeps the lock alive through COMMIT WORK, and what breaks with the wrong one?

SM12 shows Peak Util near the lock table maximum — what does Peak Util mean, and what must you rule out before simply enlarging the table?