Chapter 28 Quiz: Data Sharing and Parallel Sysplex

Instructions

Answer all 20 questions. Each question has one correct answer unless stated otherwise. Review the explanations after completing the quiz to reinforce your understanding.

Question 1

What is the maximum number of members in a DB2 data sharing group?

A) 8 B) 16 C) 32 D) 64

Question 2

Which component provides hardware-assisted lock management and shared cache for data sharing?

A) Sysplex Timer B) Coupling Facility C) VTAM D) RACF

Question 3

In a data sharing group, each member has its own:

A) DB2 catalog (DSNDB06) B) User tablespaces C) Active log datasets D) Group buffer pool structures

Question 4

What is the purpose of the Shared Communications Area (SCA)?

A) To cache data pages shared between members B) To manage global locks across the group C) To store group-level control information such as the database exception table D) To route TCP/IP connections to data sharing members

Question 5

A tablespace becomes GBP-dependent when:

A) It is first created in a data sharing environment B) Multiple members are concurrently accessing it for read/write operations C) A REORG utility is run against it D) The GBP exceeds 50% utilization

Question 6

What is the recommended directory-to-data ratio for group buffer pools?

A) 1:1 B) 2:1 C) 5:1 D) 10:1

Question 7

Physical locks (P-locks) in data sharing are used to:

A) Enforce transaction isolation levels B) Control cached data coherency between members C) Lock coupling facility structures during maintenance D) Prevent unauthorized access to tablespaces

Question 8

What happens when a DB2 member fails while holding locks on behalf of in-flight transactions?

A) All locks are immediately released B) The locks become retained locks in the coupling facility C) Other members inherit the locks and continue the transactions D) The affected tablespaces are placed in STOP status

Question 9

Peer recovery in data sharing means:

A) A failed member recovers itself when restarted B) A surviving member performs recovery on behalf of the failed member C) All members collectively rebuild the coupling facility D) The coupling facility replays log records automatically

Question 10

False contention in the lock structure occurs when:

A) Two transactions on the same member deadlock B) A lock request times out C) Two different resources hash to the same lock table entry D) The lock structure runs out of space

Question 11

Which component synchronizes clocks across all z/OS systems in the sysplex?

A) XCF (Cross-System Coupling Facility) B) Sysplex Timer / STP (Server Time Protocol) C) Coupling Facility D) WLM (Workload Manager)

Question 12

The castout process writes:

A) Active log records to archive datasets B) Buffer pool pages from local memory to DASD C) Dirty pages from the GBP to DASD D) Lock entries from the lock structure to the SCA

Question 13

Cross-invalidation (XI) signals are sent by the coupling facility to:

A) Notify a member that its locally cached page copy is no longer valid B) Inform WLM that a member has exceeded its CPU allocation C) Alert operations that the GBP is full D) Request a member to release retained locks

Question 14

The Sysplex Distributor provides:

A) Coupling facility structure allocation B) Network-level load balancing of TCP/IP connections across DB2 members C) Global deadlock detection D) Log record distribution across members

Question 15

During rolling maintenance in a data sharing group:

A) All members must be at the same PTF level at all times B) One member at a time is stopped, patched, and restarted while others continue C) The group must be quiesced before any maintenance D) Rolling maintenance is not supported; all members must be stopped simultaneously

Question 16

Version coexistence in data sharing allows:

A) Running different operating system versions on different members B) Running different DB2 versions simultaneously during a migration C) Having different CFRM policies for each member D) Mixing 31-bit and 64-bit address spaces

Question 17

When a GBP-dependent pageset has a low GBP hit ratio, the most likely cause is:

A) The pageset is rarely accessed B) The GBP is too small, causing pages to be reclaimed before re-use C) The lock structure is undersized D) Too few members are in the group

Question 18

Which XES macro is used by DB2 to perform lock operations in the coupling facility?

A) IXLCACHE B) IXLLIST C) IXLLOCK D) IXLCONN

Question 19

In Meridian Bank's data sharing design, batch processing is assigned to a dedicated member (MBK3) primarily to:

A) Comply with regulatory requirements B) Minimize GBP overhead by reducing cross-member data sharing for batch tables C) Ensure batch jobs complete faster than online transactions D) Avoid using the coupling facility for batch operations

Question 20

If both coupling facilities in a duplexed configuration fail simultaneously, the result is:

A) Data sharing continues using local buffer pools only B) A group restart is required when a CF is restored C) Individual members continue independently with no shared state D) Retained locks are automatically released

Answer Key

Question	Answer	Explanation
1	C	The architectural maximum for a DB2 data sharing group is 32 members.
2	B	The Coupling Facility provides hardware-assisted shared memory, lock management, and cache structures.
3	C	Each member writes its own active log datasets. The catalog, user tablespaces, and GBP structures are shared.
4	C	The SCA stores group-level control information including the DBET (database exception table), function registration, and castout ownership.
5	B	A tablespace becomes GBP-dependent when multiple members have concurrent read/write interest, triggering P-lock negotiations.
6	C	IBM recommends a 5:1 directory-to-data ratio to retain directory entries for cast-out pages, enabling efficient cross-invalidation.
7	B	P-locks manage cached data coherency. L-locks (logical locks) enforce transaction isolation levels.
8	B	Locks held by a failed member become retained locks, preserving data integrity until peer recovery rolls back uncommitted transactions.
9	B	Peer recovery allows a surviving member to read the failed member's log and undo uncommitted transactions, releasing retained locks.
10	C	False contention occurs due to hash collisions in the lock table — different resources mapping to the same entry.
11	B	The Sysplex Timer (STP) ensures clock synchronization within one microsecond across all systems.
12	C	Castout writes dirty pages from the Group Buffer Pool (GBP) in the coupling facility to DASD (disk).
13	A	XI signals tell a member that another member has updated a page, invalidating the local cached copy.
14	B	The Sysplex Distributor is a TCP/IP feature that load-balances connections using WLM health information.
15	B	Rolling maintenance stops, patches, and restarts one member at a time while the group continues to serve traffic.
16	B	Version coexistence allows members at different DB2 versions to run simultaneously during migration (e.g., DB2 12 and DB2 13).
17	B	A low GBP hit ratio typically means pages are being reclaimed (evicted) before they are re-read — the GBP needs more space.
18	C	IXLLOCK is the XES macro for lock operations. IXLCACHE is for cache (GBP) operations. IXLLIST is for list (SCA) operations.
19	B	Isolating batch to a dedicated member reduces GBP-dependent activity because batch pagesets are accessed primarily by one member.
20	B	Dual CF failure loses all shared state (locks, GBP, SCA). A group restart is required to rebuild structures and recover the database.