Download Full Version HPE7-J01 HP Exam

HPE Peer Persistence for the Alletra 6000 (and Nimble) provides synchronous replication and automatic transparent failover (ATF) using the Automatic Switchover (ASO) feature. This process is governed by a Quorum Witness that monitors the health of both arrays in a group.

A fundamental design principle of HPE high-availability storage is to prevent "flapping"—a condition where a workload repeatedly bounces between two sites due to an unstable connection. Consequently, the ASO process is designed to be unidirectional and sticky. When a primary site failure occurs and the Quorum Witness authorizes an automatic switchover, the secondary array becomes the "Upstream" (active) array and begins serving I/O. Once the original primary site is restored and the replication link is re-established, the arrays will automatically begin a resynchronization to ensure data consistency. However, even after the data is fully synced, the system will not automatically move the workload back to the original array. The administrator must perform a manual handover to return the volume collection to its preferred primary site.

Regarding the other options:

Option B: If the witness is unreachable, ASO is automatically disabled to prevent a "split-brain" scenario where both arrays might try to become active simultaneously.

Option C: A single controller failure within an Alletra 6000 does not trigger an ASO. Instead, the array performs an internal controller failover (ALUA) which is transparent to the Peer Persistence relationship.

Option D: In a synchronous relationship, if the target becomes unavailable, the source array typically transitions to an "Out of Sync" state but continues to serve host I/O in order to maintain application availability, rather than rejecting it.

HPE Cloud Bank Storage is an extension of HPE StoreOnce Catalyst that enables the movement of deduplicated data to public, private, or hybrid cloud object storage. The Cloud Bank Detach feature is a critical lifecycle management capability designed for long-term retention and disaster recovery scenarios.

According to the HPE StoreOnce User Guide, the "Detach" operation is a specific administrative action that essentially "unplugs" the Catalyst store from the local StoreOnce appliance while leaving the data intact in the cloud bucket (e.g., AWS S3 or Azure Blob). For an administrator to initiate the detach process, the Cloud Bank store must currently be in a Read-Write (RW) state on the StoreOnce system. If a store is currently connected as Read-Only (often the case after a disaster recovery sync), it cannot be detached until it is promoted or was originally connected in a Read-Write capacity.

Once the detach operation is executed using the required Detach Capacity LTU (License to Use), the store enters a "Detached" state. In this state, the data in the cloud becomes immutable and the store is removed from the local StoreOnce system's active management. It is important to note that once detached, the store can only be reconnected to a StoreOnce system (either the original or a new one for DR) in a Read-Only state for recovery purposes. Statement D is incorrect because you cannot write to a detached store. Statement B is incorrect because one of the primary value propositions of Cloud Bank is portability—allowing you to connect a detached store to a completely different StoreOnce appliance in a different region for recovery. Finally, while HPE Services are available for complex DR planning, they are not a technical requirement for the software-defined reconnection of a detached store.