Global Resource Directory (GRD) in Oracle 19c RAC and ASM
Which three statements are true about the Global Resource Directory (GRD) in Oracle 19c RAC database and in Clustered ASM? (Choose three.)
Community Insight
The exam tests your understanding of GRD resource distribution and memory placement; the common trap is confusing the 'large pool' with the 'shared pool' for enqueue metadata or misunderstanding the multiplicity of shadow resources.
This question examines the architecture of the Global Resource Directory (GRD) in Oracle 19c RAC and Clustered ASM, specifically regarding shadow resources and memory allocation. It establishes that shadow GRD resources can exist across multiple instances, ASM distributes GRD resources among all instances, and enqueue metadata resides in the shared pool.
Candidates often select Option D instead of E, incorrectly believing enqueue metadata is stored in the large pool. Others may miss Option B, failing to recognize that shadow resources are not limited to a single instance.
Community Discussion (3 comments)
Comments & Corrections
No comments yet — spotted an error or have a note? Share it below.
Expert Analysis
Why the Answer Is Correct
The Global Resource Directory (GRD) manages global resources in RAC. Statement B is true because shadow GRD resources can indeed exist in more than one database instance for a given resource to ensure availability and performance. Statement C is true because in a Clustered ASM configuration, GRD resources are distributed among all ASM instances to manage disk group metadata globally. Statement E is true because the master metadata for GRD Enqueue Resources is held in the shared pool of the specific instance that owns the master role, not the large pool.Why the Other Options Are Wrong
Option A is incorrect because there can be multiple GRD shadow resources for an Enqueue Resource, not just one; this redundancy supports load balancing and failover. Option D is incorrect because the master metadata for enqueue resources is stored in the shared pool, which is optimized for library cache and dictionary caching, whereas the large pool is typically used for backup and recovery operations or parallel execution servers.Community Comment Notes
Community consensus strongly supports options B, C, and G (where G corresponds to the shared pool concept, mapped here to E). As lcoleandro noted, "Shadow GRD resources may exist in multiple database instances" and "GRD resources in Clustered ASM are distributed among all ASM instances." Another user confirmed that "BCG is correct," aligning with the standard Oracle documentation on GRD architecture. Comments consistently refute the idea that only one shadow resource exists or that the large pool holds enqueue metadata.Exam Strategy
When studying RAC architecture, focus on the distinction between the Shared Pool and Large Pool usage for different types of internal structures like the GRD and Library Cache. Memorize that GRD resources are distributed across instances in both RAC databases and Clustered ASM to ensure high availability.
Frequently Asked Questions
Why is enqueue metadata in the shared pool, not large pool?
Enqueue resources require frequent access to library cache-like structures, which are optimized in the shared pool. The large pool is reserved for large contiguous allocations like backup buffers.
Can multiple shadow GRD resources exist for one enqueue?
Yes, multiple shadow resources can exist to provide redundancy and allow other instances to quickly access lock information without contacting the master instance immediately.