Dell PowerStore Operate 2023 flashcards

32 cards. Read the term and the definition here. The deck at the top of this page is optional.

Active-active nodes

Both nodes in a PowerStore appliance process I/O simultaneously; if one node fails, its workloads fail over to the surviving node, which continues serving from shared dual-ported NVMe drives.

Base enclosure drive slots

The PowerStore base enclosure has 25 NVMe slots, but the first four are reserved for NVRAM and boot modules, leaving 21 usable data drive slots before expansion enclosures are added.

PowerStore T versus X

T models are native unified block-and-file appliances running PowerStoreOS directly; X models run an ESXi-based controller with embedded vSphere for block-only workloads.

Cluster scale-out limit

A PowerStore cluster contains from one to four appliances; adding appliances increases capacity and performance, and all appliances must run compatible PowerStoreOS versions.

NVRAM role

Each node contains NVRAM that caches writes and protects them against power loss, allowing acknowledged writes to be preserved until destaged to the NVMe drives.

Initial Configuration Wizard

The ICW discovers unconfigured appliances, sets management networking, creates the cluster, and establishes the admin account; it is reached through the service port or management network.

Management versus storage network

Management traffic (1 GbE) carries PowerStore Manager and API access; front-end storage traffic uses separate 25 GbE or FC ports, and the two networks should not be shared.

FC zoning rule

Use single-initiator, multiple-target zones on the SAN switch, and zone both nodes' FC ports to each host HBA so every host retains a path after a node failure.

NTP requirement

Configure NTP during installation; unsynchronized clocks break replication, snapshot scheduling, and certificate validation across the cluster and remote systems.

PowerStore Manager roles

Built-in roles include Administrator with full control, Storage Administrator for storage objects, VM Administrator for X-model virtualization, and Read-Only for monitoring.

SupportAssist connectivity

SupportAssist sends alerts and telemetry to Dell and enables remote support and automated case creation; it requires outbound connectivity and must be enabled and verified.

Event log versus audit log

The event log records system events such as alerts and state changes; the audit log records user and API actions, and both can be exported for review.

LDAP directory integration

Directory services are configured under Settings with a bind account, base DN, and user or group mapping, allowing AD users to log in with assigned PowerStore roles.

Volume group purpose

A volume group collects volumes managed and protected together; snapshots, replication, and protection policies apply to the group as a single unit.

Thin provisioning

Volumes are thin by default: capacity is allocated only as data is written, so the sum of advertised volume sizes can exceed physical usable capacity.

Host object initiators

A host object lists initiators such as FC WWPNs or iSCSI IQNs and NVMe NQNs; a volume is accessible only after the host is mapped to it.

Host group use case

A host group bundles hosts that need identical access to the same volumes, such as cluster nodes; map volumes to the group so every node sees the same LUNs.

Protection policy

A protection policy combines a snapshot rule and optionally a replication rule; assigning it to a volume or volume group enables scheduled local and remote protection.

Snapshot retention behavior

A snapshot rule sets a schedule and retention count; when the count is exceeded, the oldest snapshot is deleted automatically to free pool capacity.

Asynchronous replication RPO

Asynchronous replication sends changes on a schedule, and the RPO defines the maximum data loss window if the source site fails before the next transfer.

Metro replication

Synchronous replication between two sites with zero RPO and active-active host access; it requires low-latency links and supported distance and configuration limits.

Replication failover

Failover promotes the destination copy to read/write for production use; failback later resynchronizes changes and returns production to the original site.

iSCSI target portal

Hosts connect to a PowerStore iSCSI target portal IP on bonded front-end Ethernet ports; CHAP authentication can be enabled to validate initiators.

NVMe over TCP front end

NVMe-TCP uses the same 25 GbE front-end ports as iSCSI; the host connects using an NQN and target portal, and the subsystem NQN identifies the target.

Multipath requirement

Each host should have at least two paths, one through each node, using MPIO or PowerPath, so a node or link failure does not interrupt I/O.

Jumbo frame consistency

MTU 9000 must match on host interfaces, switch ports, and PowerStore front-end ports; a mismatch causes fragmentation, degraded performance, or failed connections.

Capacity overcommit ratio

Subscription divided by usable capacity gives the overcommit ratio; thin volumes can push this above one, so monitor it to avoid running out of physical space.

Pool deduplication and compression

These are storage-pool settings; enabling them on an existing pool starts a background scan and does not immediately reduce data already written.

Latency percentile reading

Average latency hides outliers; a rising 99th-percentile latency with flat IOPS usually points to a host or network bottleneck rather than array saturation.

Non-disruptive upgrade

PowerStoreOS upgrades proceed node by node on a healthy cluster, with I/O continuing on the peer node; failed pre-upgrade health checks abort the upgrade.

Drive replacement rebuild

Identify the failed slot from the alert, replace the drive with a supported part, and the system rebuilds automatically from the mirrored data without host interruption.

Support bundle collection

Collect logs and support materials from PowerStore Manager or CLI before opening a case; Dell support requires the bundle to analyze the issue.