FAIS Practice Questions

Newer versions of Purity often include more advanced space accounting and reporting mechanisms. The 'unreclaimable space' alert indicates that the host operating system or hypervisor has deleted data but has not sent the corresponding UNMAP/TRIM commands to the array to free up the blocks. The alert is informational, not an error. The correct resolution is for the host administrator to run space reclamation utilities (like esxcli storage vmfs unmap in VMware or fstrim in Linux) to allow the array to reclaim the dead space.

Correct power redundancy involves balancing the PSUs across the independent power sources. On a FlashArray, PSUs are paired for internal redundancy (0 and 2 are a pair, 1 and 3 are a pair, serving different controllers/components). To achieve maximum fault tolerance, one PSU from each pair should be connected to each PDU. Connecting PSU0 and PSU2 to PDU A, and PSU1 and PSU3 to PDU B, ensures that the failure of an entire PDU (e.g., PDU A) will still leave one operational PSU for each internal redundant pair (PSU1 and PSU3), allowing the array to continue running without interruption.

Pure Storage FlashArray systems are designed exclusively for installation in 4-post racks to ensure proper support, stability, and airflow. They are not supported in 2-post racks. The correct and professional action is to halt the installation and inform the customer of this mandatory requirement. Attempting to use a conversion kit is not a vendor-supported solution and could lead to equipment damage and safety issues.

The purehw list command is specifically designed to display the status of all hardware components within the FlashArray. This includes controllers, power supplies, fans, SAS modules, DirectFlash modules, and network cards. Running this command after an upgrade is the most effective way to get a consolidated view and confirm that all physical components are recognized and operating in a healthy state.

The NV-RAM (Non-Volatile RAM) modules serve as the write cache for the FlashArray. When a host sends a write I/O, it is written to NV-RAM on both controllers, and an acknowledgement is sent back to the host immediately. This process provides very low latency for write operations. The data in NV-RAM is then de-staged to the flash modules in the background. The non-volatile nature of the modules ensures that in-flight writes are protected against power failure.

The correct procedure is to first bring the software to a compatible level before introducing new hardware. The Purity//FA software on the existing R2 chassis must be upgraded to the minimum version that supports the R4 hardware before the physical chassis swap occurs. Attempting to bring up the R4 chassis with an unsupported Purity version on the DFMs will result in failure. Upgrading after the swap is not possible if the controllers cannot boot with the old software version.

A 'shelf mismatch' error during initial boot often indicates an improper SAS cabling connection between the controllers and the expansion shelves. Reversing the SAS cables (e.g., connecting a controller's 'out' port to another controller's 'in' port, or mis-cabling the shelf loop) is a common installation error that would prevent the controllers from correctly discovering and identifying the attached shelves, thus halting the setup process.

The correct action is to proactively address the discrepancy between the customer's provisioned resources and Pure Storage's best practices. The specialist should communicate the need for a second, separate IP address and subnet for the second replication interface to ensure high availability and prevent a single point of failure in the replication network. Proceeding with a non-redundant configuration would be a disservice to the customer.

This statement is false. While data evacuation is one method, it is not always required. If the FlashArray has sufficient free space and available drive slots, capacity consolidation can often be performed non-disruptively by adding the new, larger data packs, allowing Purity to rebalance data onto them, and then removing the older, smaller data packs one by one. A data evacuation is typically only necessary in scenarios with very high capacity utilization or when a full chassis swap is required.