Pass the Pure Storage FlashArray Storage FlashArray-Storage-Professional Questions and answers with ExamsMirror

Similar to the previous question regarding directory services, Pure Storage Purity OS handles Role-Based Access Control (RBAC) overlaps by granting the most permissive role available to the user.

When configuring LDAP or Active Directory authentication on a FlashArray, administrators map directory groups to specific FlashArray roles (Array Admin, Storage Admin, Ops Admin, Read Only). If a user happens to be a member of multiple LDAP groups that are mapped to different roles on the array, Purity evaluates all mapped roles and automatically assigns the user the highest level of privilege during their session.

Since " Array Admin " has full administrative rights over the entire array (including hardware management, directory services configuration, and firmware upgrades) and sits higher in the hierarchy than " Storage Admin " (which is restricted to provisioning and managing storage objects like volumes and hosts), the system will seamlessly grant the user Array Admin permissions.

Here is why the other options are incorrect:

User will not be able to login (B): Purity is designed to handle this exact scenario smoothly. It resolves the conflict by defaulting to the higher privilege, rather than throwing an error or denying access.

User will have Storage Admin permissions (C): The system does not default to the lowest privilege or restrict access when a higher-level group membership is present and valid.

Requirement for Lossless Ethernet: NVMe over RoCE (RDMA over Converged Ethernet) requires a lossless fabric to function correctly. Unlike standard iSCSI which uses TCP for error recovery, RoCE assumes the network will not drop packets. If the network is " lossy, " performance degrades significantly.

The Role of PFC: Priority Flow Control (PFC) (IEEE 802.1Qbb) is the specific mechanism used in Data Center Bridging (DCB) to provide flow control on a per-priority basis. It allows the switch to send a " pause " frame to the sender when buffers are full, preventing packet drops.

Symptom Analysis: In the scenario provided, the array itself is not overloaded ( " within the performance envelope " ). However, the addition of a new workload increased traffic to the point where buffer congestion occurred. Because PFC was likely misconfigured (either on the FlashArray ports, the network switches, or the host NICs), the network dropped packets instead of pausing traffic. This leads to " go-back-N " retransmissions and massive latency spikes that affect all workloads sharing that fabric.

Pure Storage Best Practices: Pure Storage documentation for NVMe-RoCE emphasizes that PFC must be enabled and consistent across the entire path. If there is a mismatch in PFC configuration, the resulting packet loss will cause the symptoms described: extreme latency and potential service outages.

Analyzing the Graph: The provided capacity graph shows a sharp, vertical drop in Unique and Shared space (the teal/green sections) at approximately 10:45. Simultaneously, there is a massive spike in System space (the grey shaded area).

The " System " Space Spike: In Purity, when a volume is deleted and eradicated, the data isn ' t physically wiped from the flash modules instantly. Instead, the logical references are removed, and the blocks are moved into a " Pending Reclamation " state. In older versions of Purity reporting (or specific views), this pending work is often categorized under System space until the background Garbage Collection (GC) process can physically reclaim it.

The Decay Curve: Notice the grey " System " area gradually tapers down over the following hours (from 11:00 to 14:00). This is the visual representation of the array ' s Garbage Collection engine working in the background to physically free up those blocks.

Why Option C is incorrect: If SafeMode were enabled, the space would not move into " System " space in this manner; it would remain locked in the Snapshots or Destroyed bucket for the duration of the retention period, preventing the graph from dropping at all.

Why Option A is incorrect: If space reclamation stopped working, the grey " System " area would stay flat and would not decay over time. The fact that it is trending downward proves that reclamation is actively processing the deleted volume.

SNMP Capabilities on FlashArray: Pure Storage FlashArrays support SNMP (Simple Network Management Protocol) versions v2c and v3. The primary use case for SNMP on the FlashArray is for integration with third-party monitoring tools (like SolarWinds, Nagios, or Zabbix) to provide a real-time health and performance heartbeat.

Performance Metrics (MIBs): The Pure Storage Management Information Base (MIB) specifically exposes high-level array performance metrics. These include the " Big Three " of storage performance: Latency (response time in microseconds), Bandwidth (throughput in bytes per second), and IOPS (I/O operations per second).

Scope of Monitoring: While the FlashArray GUI and CLI provide deep granularity (per host, per volume, or per pod), standard SNMP queries typically focus on array-wide performance and health status. Detailed " Host and volume performance " (Option B) is generally more effectively monitored via the Pure Storage REST API or the Pure1 VM Analytics, as SNMP is less efficient for pulling large tables of per-object data.

Capacity vs. Performance: While SNMP can report on capacity (Option C) through specific OIDs, the question specifically asks about performance statistics . In the context of Pure Storage monitoring documentation, the core performance metrics provided via SNMP traps and polling are latency, bandwidth, and IOPS.

Understanding Space Reporting: To understand this behavior, you have to look at how Purity calculates capacity. Pure Storage uses a data reduction engine where data is deduplicated and compressed.

Volume Space vs. Shared Space:

Volume Space: This represents the unique data belonging to a specific volume that is not shared with any other volume via deduplication or snapshots.

Shared Space: This represents the data that is common across multiple volumes or snapshots. If you have two volumes that are clones of each other, most of that data is " Shared. "

The " Shift " Mechanism: When a volume is deleted (and potentially eradicated), the data it once shared with other volumes no longer needs to be " shared. "

Imagine Volume A and Volume B share 100GB of data. That 100GB is accounted for in Shared Space .

If you delete Volume B, that 100GB of data is now only referenced by Volume A.

Consequently, that 100GB is moved from the Shared Space bucket into Volume A ' s Volume Space bucket.

Net Result: The total physical space used on the array remains the same initially, but the accounting shifts. You see a decrease in Shared Space and an identical increase in the Volume Space of the remaining volumes that held those deduplication references.

Understanding Alert 51: On a Pure Storage FlashArray, Alert 51 signifies that a Protection Group ' s replication is lagging behind its scheduled completion time. This does not necessarily mean the connection is " down, " but rather that the volume of data being sent is exceeding the available throughput or is being queued behind other tasks.

The Triage Process:

Open Alerts: You must check for related alerts (like Alert 20 for " Replication Connection Down " ) to determine if the delay is caused by a total link failure or just congestion.

Replication Jobs in Progress: Because FlashArray uses a specialized engine to manage replication, having multiple large snapshots from different Protection Groups replicating simultaneously can saturate the " replication pipe. " Checking active jobs helps determine if there is a scheduling " traffic jam. "

Replication Bandwidth: Comparing the current outgoing replication throughput against the historical average or the physical limit of the replication ports helps identify if the delay is due to a sudden increase in Data Change Rate (churn) or a reduction in network performance.

Why Option B is incorrect: If a Protection Group were disabled, replication wouldn ' t be " delayed " —it would be stopped, which triggers a different alert state. Cabling issues usually result in " Connection Down " alerts rather than just " Delayed " alerts.

Why Option C is incorrect: Disconnecting replication is a destructive troubleshooting step that will only increase the lag and RPO. You should always analyze the existing data flow before breaking the connection.

In the Pure Storage Purity operating environment, a volume can only be a member of one Volume Group at a time .

When an administrator navigates to a Volume Group in the GUI and clicks to add members, the system filters the inventory and only displays volumes that are currently " unassigned " (not belonging to any Volume Group). If a volume is already residing inside another Volume Group, Purity intentionally hides it from this available list to prevent conflicting overlapping memberships. To resolve this, the administrator must first navigate to the volume ' s current Volume Group, remove the volume from that group, and then it will become available to add to the new one.

Here is why the other options are incorrect:

It is already part of a Protection Group (B): Protection Groups (pgroups) manage snapshot and replication schedules. A volume can absolutely be a standalone member of a Protection Group while simultaneously being added to a Volume Group. Being in a pgroup does not hide it from the vgroup selection list.

It is protected by SafeMode (C): SafeMode is a ransomware protection feature that prevents the manual eradication of destroyed volumes and snapshots before their retention timer expires. It does not dictate or restrict logical organizational containers like Volume Groups.

Command Purpose: The purehost monitor command is the primary tool in the Pure Storage CLI for observing real-time performance and connectivity health from the perspective of the hosts connected to the FlashArray.

The --balance Flag: When the --balance flag is added, the output shifts from general performance (IOPS, bandwidth, latency) to showing how I/O is distributed across the available paths (controllers and ports). This is critical for identifying " unbalanced " loads, which usually point to misconfigured MPIO (Multi-Path I/O) on the host side (e.g., a host only using one controller ' s ports).

Interval vs. Repeat:

The --interval flag specifies the time in seconds between each sample. In option C, --interval 15 tells the array to refresh the data every 15 seconds.

The --repeat flag (seen in option A) is used to limit the total number of samples taken before the command exits. However, in standard troubleshooting, the administrator typically wants a consistent stream of data until manually stopped (Ctrl+C).

--resample (seen in option B) is not a valid flag for the purehost monitor command in Purity.

Best Practice: When troubleshooting multipathing, Pure Storage recommends monitoring the balance to ensure that the " Relative I/O " percentage is roughly equal across all active paths. Large discrepancies often indicate that the host ' s MPIO policy is set to " Failover Only " instead of the recommended " Round Robin " or " Least Queue Depth. "

API Evolution: Pure Storage introduced REST API 2.x to provide a more scalable, standardized, and performant way to automate FlashArray management. It uses a different authentication method (OAuth2 with API Clients) compared to the API Token-based method in 1.x.

Feature Freeze on 1.x: As of Purity 6.x and beyond, Pure Storage has designated REST API 1.x as " Legacy. " While 1.x is still supported for backward compatibility to ensure older scripts don ' t break, all new Purity features (such as specialized ActiveDR commands, advanced File Services, or new hardware capabilities) are only developed and exposed via REST API 2.x .

Side-by-Side Support: Contrary to option C, both versions are supported side-by-side on the same array. An administrator can run a script using 1.x for volume creation and another script using 2.x for performance monitoring simultaneously without contacting support.

Feature Parity: REST API 2.x has long since reached and exceeded the capabilities of 1.x. It offers improved filtering, pagination, and a more consistent object model (e.g., /volumes instead of multiple nested endpoints).

Best Practice: Pure Storage strongly recommends that all new automation projects use REST API 2.x to ensure access to the full suite of Purity features and to future-proof infrastructure-as-code (IaC) workflows.

Pure Storage strongly recommends an MTU size of 9000 (Jumbo Frames) for replication networks—such as those used for Asynchronous Replication, ActiveCluster, and ActiveDR—as well as for iSCSI and NVMe/TCP data networks.

A 9000-byte MTU significantly reduces protocol overhead and CPU processing load on the storage controllers by allowing a much larger payload of data to be transmitted inside a single network packet. During heavy replication, this drastically increases throughput and maximizes bandwidth efficiency.

Here is why the other options are incorrect:

1500 (B): While 1500 bytes is the standard default MTU for Ethernet and is exactly what Pure Storage recommends for the management ports (vir0), it is not the recommended optimization for high-throughput replication traffic. (Note: If your network cannot support 9000 end-to-end, 1500 must be used to prevent packet fragmentation, but 9000 remains the best-practice recommendation).

4200 (A): This is an arbitrary number and is not a standard network MTU size used in Pure Storage environments.