Pass the Pure Storage FAIS FlashArray-Implementation-Specialist Questions and answers with ValidTests

The FlashArray//XR2 and //XR3 (representing the FlashArray//X generation) utilize a standardized 3U (3 Rack Units) chassis form factor.

This 3U chassis is the foundation for the majority of the FlashArray//X line, including the //X10, //X20, //X50, //X70, and //X90 models. It is designed to be compact while providing sufficient airflow and space for the dual controllers, two data packs (up to 20 DirectFlash Modules), and power/cooling infrastructure.

Dimensions: Approximately 5.12 inches (13 cm) in height.

Rack Layout: Implementation Engineers must reserve 3U of contiguous space for the base controller chassis. If expansion shelves are included, they typically require an additional 2U each (for SAS or NVMe shelves).

Option B (4U) is not a standard height for current Pure Storage arrays. Option C (5U) refers to the FlashArray//XL , a larger, performance-optimized platform designed for maximum scale and density, which uses a significantly taller chassis to accommodate more expansion slots and power components. Therefore, for the specified R2/R3 X-series, 3U is the correct height.

When executing a cross-generational Hardware Non-Disruptive Upgrade (HWNDU) for a customer who has been with Pure Storage for several years, an Implementation Engineer will often encounter legacy storage media. Early generations of the FlashArray relied on SAS-based Solid State Drives (SSDs) before the industry entirely shifted to the much faster NVMe protocol (DirectFlash Modules).

Pure Storage is famous for its Evergreen architecture, which allows customers to upgrade their controllers without having to rip and replace their existing drives. However, a standard modern //XR4 controller natively communicates strictly via NVMe and PCIe lanes. If a customer is upgrading an array that still contains legacy SAS drives (in this scenario, 10 SAS drives mixed with 10 NVMe drives), a standard NVMe-only controller will not be able to read the SAS media, leading to immediate data unavailability.

To solve this, Pure Storage manufactures specific controller configurations. The SE (SAS-Enabled) controller models are explicitly designed with the necessary internal SAS Host Bus Adapters (HBAs) and routing logic to support legacy SAS media alongside modern NVMe drives. Conversely, the NS (Non-SAS) controllers strictly support NVMe media. Because this specific chassis still houses 10 SAS modules, the Implementation Engineer must verify that the Bill of Materials (BOM) shipped for the upgrade is explicitly the "FLASHARRAY//X50 R4, SE" variant to ensure a safe, non-disruptive migration.

Data evacuation from a shelf is a sensitive background process where data is migrated off specific physical media to free it for removal. While the purearray evacuate command initiates the data movement, the finalization of the process—specifically the logical removal of the shelf object from the system configuration to allow for safe physical disconnection—often requires engineering-level commands or verification.

Therefore, before Pure Storage Support can log in to run these specific completion scripts or verification checks, the Implementation Engineer must Enable Remote Assist (RA) .

Remote Assist creates a secure, outbound SSH tunnel that allows Support to access the array remotely.

Without RA enabled, Support cannot connect to the array to execute the necessary puredrive or internal maintenance commands to mark the shelf as "vacated" and safe to unplug.

Powering off the shelf (Option B) before this step could result in data loss or an array outage if the system still believes data resides there.

=========

FlashArray performance and capacity efficiency are optimized when drives are grouped into "Wide Write Groups" (typically 20 modules). Purity OS generally enforces strict homogeneity within a write group to ensure consistent performance and reliability. Specifically, mixing standard DirectFlash Modules (DFM) with DirectFlash Modules with Distributed NVRAM (DFMD) in the same write group is restricted by default because of the architectural differences in how they handle persistent memory and I/O acknowledgment.

However, in upgrade scenarios where a customer is expanding an existing 16-module DFM group to a 20-module group using newer DFMD spares or expansion packs, Purity includes an engineering tunable to permit this mixed configuration. The correct procedure is to physically install the new DFMDs and run the admission command. Because of the mix, the array might initially hesitate or create a separate group. To force the creation of a single 20-drive Wide Write Group (WWG) , the Implementation Engineer must work with Pure Storage Support to apply the specific tunable puretune --set ALLOW_DFMD_IN_DFM_WWG 1 . This overrides the default restriction, allowing the system to merge the disparate module types into a single, high-efficiency capacity pool.

=========

The FlashArray//XR4 architecture fundamentally changed how external capacity is attached compared to older SAS-based or early NVMe platforms. Because DirectFlash Shelves (DFS) utilize massive 100GbE RDMA over Converged Ethernet (RoCE) for their backend loops to maintain NVMe-level performance, specific port and slot assignments are heavily regulated.

On an //XR4 model (such as the //X70 R4 or //X90 R4), the controllers feature dedicated, integrated 100GbE ports on the motherboard (specifically eth18 and eth19) which are strictly reserved for the first DirectFlash Shelf loop.

However, when an Implementation Engineer needs to install a second DirectFlash Shelf, the integrated onboard ports are already fully saturated. Pure Storage does not support daisy-chaining NVMe-oF RoCE shelves in the same manner as legacy SAS enclosures. Instead, a new dedicated backend fabric loop must be established. To accomplish this, the engineer must install an additional 100GbE RoCE PCIe expansion card. According to the official //XR4 PCIe Slot Population Matrix, the designated location for this secondary backend expansion card is PCIe Slot 4 . Installing the card into other frontend-designated slots (like Slot 0 or Slot 3) will trigger a topology mismatch during the array's pre-flight hardware checks and prevent the new shelf from initializing.

For a Pure Storage FlashArray to fully utilize its predictive support capabilities, it must maintain a persistent, secure outbound connection to the Pure1 cloud infrastructure. This allows the array's Callhome Connection Manager (CCM) to transmit telemetry data, hardware alerts, and heartbeat signals. However, in many enterprise environments, storage management networks do not have direct, unfiltered access to the public internet. Instead, all outbound HTTP/HTTPS traffic must be routed through a corporate proxy server.

To facilitate this, the Implementation Engineer must configure the FlashArray to tunnel its Call Home traffic through the customer's proxy. The officially supported Purity CLI command to achieve this is purearray setattr --proxy https:// < proxy_address > : < port > .

By executing this command, the system modifies its global network attributes. Purity immediately updates the internal routing daemon, instructing it to encapsulate all outgoing Pure1 management traffic and direct it to the specified proxy IP or hostname on the designated port (e.g., port 8080 or 3128). Options B, C, and D are fabricated commands that do not exist within the Purity CLI framework. Once the proxy is set, the engineer should always validate the configuration by running purearray test phonehome to ensure the proxy is properly allowing the traffic to pass through to the Pure Storage cloud endpoints.

To view the currently configured Network Time Protocol (NTP) servers on a FlashArray, the correct command is purearray list --ntpserver .

Command Hierarchy: While NTP is a networking function, in the Purity CLI object model, time synchronization is considered a global array attribute rather than a specific interface setting. Therefore, it falls under the purearray command namespace.

Usage: Running purearray list --ntpserver returns a list of the IP addresses or FQDNs of the time servers the array is currently using to sync its system clock.

Invalid Options:

purenetwork list (Option B) displays IP interface configurations (IPs, subnets, gateways) but does not list service-level pointers like NTP.

puredns list (Option A) is strictly for DNS nameservers and suffix configurations.

Importance: Verifying NTP is critical for log timestamp accuracy, ActiveCluster mediation, and Kerberos authentication.

FlashArray//XL architectures offer the ability to leverage DirectMemory Modules (DMMs) to drastically accelerate read-intensive workloads. DMMs are specialized drives built on Storage Class Memory (SCM) technologies, such as Intel Optane. They act as a high-speed read cache tier, sitting seamlessly between the array's ultra-fast NVRAM (which handles incoming writes) and the standard DirectFlash Modules (which provide the bulk capacity).

Because the entire purpose of a DirectMemory Module is to provide the absolute lowest read latency possible, physical proximity to the controller's CPUs and direct access to internal PCIe lanes is mandatory. Therefore, DMMs must be installed directly into the primary FlashArray chassis in specifically designated slots.

Installing DMMs into an external DirectFlash Shelf (DFS) is strictly prohibited and physically unsupported. Routing read-cache traffic across an external NVMe-oF (RoCE) backend fabric—even at 100GbE speeds—would introduce microsecond latency overheads that completely defeat the performance benefits of Storage Class Memory. Furthermore, DMMs share the exact same physical U.2/NVMe drive slot form factor as standard capacity DFMs (making Option C false), and they are supported across multiple array models beyond just the //XL 170 (making Option B false). An Implementation Engineer must strictly validate the chassis population rules before inserting DMMs to ensure the system recognizes the cache tier correctly upon initialization.

Executing an intra-series Hardware Non-Disruptive Upgrade (HWNDU) from a FlashArray//XR3 to an //XR4 is a highly orchestrated procedure. It involves failing over active I/O responsibilities, physically removing legacy controller nodes, and seamlessly introducing newer generation hardware without causing any downtime for the connected hosts.

Before an Implementation Engineer physically unplugs or removes a secondary controller—or initiates a failover to take the primary controller offline—it is an absolute prerequisite to verify that the frontend host multi-pathing is healthy. If a host is accidentally single-pathed to the controller that is about to be removed, an all-paths-down (APD) event will occur, causing immediate application disruption.

To safely validate this from the array's perspective, the engineer must execute the iobalance --sampletime 30 command within the Purity CLI. This specialized diagnostic command forces the array to actively sample and monitor the incoming SCSI/NVMe read and write commands across all frontend target ports for a continuous 30-second window. The output produces a detailed matrix showing exactly how traffic is distributed across both controllers and all active host initiators. If the data confirms that I/O is cleanly balanced and redundantly flowing across the surviving controller's paths, the engineer can confidently proceed with the physical removal phase of the HWNDU.

This scenario highlights a critical architectural transition point within the Pure Storage FlashArray product line. The arrays utilize dedicated NVRAM modules to safely ingest and acknowledge incoming writes before destaging them to the capacity flash drives.

On the lower to mid-range models (such as the //X10, //X20, and //X50), the system requires only two NVRAM modules to maintain a highly available, mirrored write cache. These are physically installed in the central chassis slots NVB0 and NVB1. However, the higher-end, performance-heavy models (such as the //X70 and //X90) boast significantly more processing power and ingest bandwidth. To support these massive workloads without creating a cache bottleneck, the architecture mandates four NVRAM modules, filling slots NVB0, NVB1, NVB2, and NVB3.

Therefore, when performing an inter-generational, cross-model Hardware Non-Disruptive Upgrade (HWNDU) from an //X50 to an //X70 , the Implementation Engineer is traversing that architectural boundary. While performing the controller swaps, the engineer is strictly required to install two additional, newly shipped NVRAM modules into the previously empty bays NVB2 and NVB3. If they fail to add these modules, the new //X70 controllers will detect a cache hardware deficiency and refuse to initialize, halting the upgrade process. Options B, C, and D do not cross this 2-to-4 NVRAM threshold.