Validates ability to design Dell PowerScale (formerly Isilon) scale-out NAS solutions including infrastructure planning, capacity sizing, data management, and workload monitoring.
OneFSSmartConnectSmartPoolsCloudPoolsSyncIQSmartSyncSnapshotIQSmartLockSmartQuotasSmartDedupeZRBACInsightIQNDMP BackupGroupnets / IP PoolsAccess ZonesJob EngineHealthCheckisi statistics
Master Dell PowerScale architecture and design — from infrastructure node selection and OneFS networking through advanced data services, sizing methodologies, and operational monitoring tools.
›OneFS overview — single filesystem spanning all nodes, no SAN/NAS split, linear scalability to 252 nodes; journal-based protection replicated across nodes
›Networking — front-end client-facing interfaces (10/25/40/100GbE); InfiniBand back-end interconnect for intra-cluster data and journal traffic
›Groupnets — independent networking domains with separate routing tables, DNS, and NTP configurations per tenant or department
›IP Pools — subset of IPs within a groupnet/subnet assigned to a SmartConnect zone; control which nodes serve which client networks
›Access Zones — logical partitions: each zone has an independent share/export namespace, protocol settings, and authentication providers
›SmartConnect — DNS-based load balancing across node IPs in an IP pool; supports round-robin, CPU usage, connection count, and throughput policies
›Cluster sizing — usable vs raw capacity, protection overhead, journal space, node count minimums per protection level, hot-spare equivalence
Module 02
Management, Identity Management, and Advanced Applications
›SnapshotIQ — schedule-based or manual snapshots on directories, snapshot reserve capacity, snapshot aliases, and directory-level snapshot restore
›SmartLock WORM — enterprise mode vs compliance mode, clock authority, delete restrictions, retention date setting; meets SEC 17a-4 and other compliance mandates
›SmartPools — file pool policies: match files by directory, extension, or date; tier files to different node pool types; spillover policies when primary pool is full
›CloudPools — stub files on-prem, actual data in cloud (AWS/Azure/GCP/Virtustream); transparent cloud tiering with automatic cloud retrieve on access
›Authentication providers — local, LDAP, Active Directory, NIS, Kerberos; configured per access zone for multi-tenant identity isolation
›Role-based access control (RBAC) — built-in roles (SystemAdmin, AuditAdmin, BackupAdmin, NetworkAdmin) plus custom roles for least-privilege administration
›Zone-based RBAC (ZRBAC) — platform API namespace isolation per access zone; delegate admin rights within an access zone to tenant administrators
›SyncIQ — asynchronous replication of directories across clusters; supports schedule/manual policies, bandwidth throttling, and failover/failback workflow
›SmartSync — cloud sync policies, bi-directional sync with S3-compatible targets for hybrid cloud data workflows
Module 03
Sizing Considerations
›Data availability and protection — N+M protection levels: default N+2:1, N+3:2, N+4:2 mirrored; tolerate M drive and 1 node failures; FEC journal-based protection
›NDMP backup sizing — three-party NDMP with DMA, two-party NDMP, dedicated SmartConnect zone for backup traffic to isolate impact on production
›SmartQuotas — directory/user/group quotas with advisory, soft, and hard limits; grace periods for soft limits; automated quota reports
›SmartDedupe — post-process deduplication with schedule-based runs; dedupe domains; estimated savings analysis before enabling; shadow store concept for shared data blocks
›Inline data reduction — LZ4 compression on F-series all-flash nodes; savings reporting in cluster statistics
›iostat and netstat — system-level disk I/O monitoring (iostat) and network connection tracking, per-NIC statistics (netstat) for troubleshooting throughput issues
›iperf — bandwidth testing between nodes or clients to verify back-end InfiniBand and front-end Ethernet throughput; baseline capacity validation
›SNMP — MIBs for cluster monitoring integration with Nagios, Zabbix, and other SNMP managers; trap configuration for proactive alerting
›SRS (Secure Remote Services) — proactive monitoring, remote diagnostics, ESRS gateway integration; automated case creation on hardware failures
Complete PowerScale design curriculum — from infrastructure node selection and OneFS fundamentals through advanced data services, capacity sizing, and workload monitoring tools.
›Node types — H-series (high-performance NVMe/HDD hybrid), F-series (all-flash NVMe for latency-sensitive workloads), A-series (archive/deep archive high-capacity HDDs), B-series (backup targets); Gen6 hardware: chassis-based vs 1U node form factors
›OneFS filesystem — single global namespace across all nodes, no separate SAN or NAS split; linear scalability from 3 to 252 nodes; parallel data and metadata distribution
›Front-end networking — client-facing interfaces at 10/25/40/100GbE; bonding and LACP for resilience; InfiniBand back-end for intra-cluster journal replication and data movement
›Groupnets — independent networking domains providing separate routing tables, DNS resolvers, and NTP servers per tenant; foundation for multi-tenancy
›IP Pools — a subset of IP addresses within a groupnet subnet assigned to a SmartConnect zone; controls which IP range a set of nodes presents to clients
›Access Zones — logical cluster partitions with independent share/export namespace, protocol settings, and per-zone authentication providers; enables multi-tenant isolation
›SmartConnect — DNS-based load balancing across node IPs in a pool; delegation zone delegated from corporate DNS; policies: round-robin, CPU usage, connection count, or throughput
›Cluster sizing — calculating usable vs raw capacity (protection overhead, journal overhead, hot-spare equivalence); node count minimums per N+M protection policy; performance headroom planning
›SnapshotIQ — schedule-based or on-demand directory-level snapshots; snapshot reserve capacity allocation; snapshot aliases for application-consistent access; granular file/directory restore
›SmartLock WORM — enterprise mode (cluster admin can delete after retention) vs compliance mode (no deletion until retention expires); clock authority configuration; retention date enforcement
›SmartPools — file pool policies matching files by directory path, file extension, or access/modification date; tier matched files to target node pool types; spillover policy when primary pool is full
›CloudPools — transparent cloud tiering: stub files remain on-prem, actual data stored in AWS S3/Azure Blob/GCP GCS/Virtustream; automatic cloud retrieve on file access; cloud recall on anchor
›Authentication providers — local users, LDAP, Active Directory (Kerberos + NTLMv2), NIS, and MIT Kerberos; configured per access zone; multi-provider stacking and search order
›RBAC — built-in roles: SystemAdmin (full access), AuditAdmin, BackupAdmin, NetworkAdmin; custom roles with granular privilege sets for least-privilege administration
›ZRBAC — platform API namespace isolation per access zone; delegate zone-specific admin rights to tenant admins without granting cluster-wide privileges
›SyncIQ replication — asynchronous directory replication between clusters; scheduled or manual policies; bandwidth throttling windows; failover (break association) and failback (resync) workflow
›SmartSync — cloud sync policies; bi-directional sync with S3-compatible targets; data mobility for hybrid/multi-cloud workflows
›Node minimums — N+2:1 requires minimum 3 nodes; N+3:2 requires minimum 4 nodes; higher protection levels and mixed node types have additional minimums documented in sizing guides
›NDMP backup — three-party NDMP (cluster as data mover, DMA server, backup device on separate host); two-party NDMP (local); dedicated SmartConnect zone for backup IPs to separate backup traffic from production
›SmartQuotas — directory, user, and group quota types; advisory (notify only), soft (allow with grace period), and hard (block writes) limits; grace period configuration; scheduled quota reports to storage admins
›SmartDedupe — post-process deduplication running on a schedule; dedupe domains define which directories are eligible; savings analysis report estimates reduction before committing; shadow store shared data blocks
›Inline data reduction — LZ4 compression enabled on F-series all-flash nodes at write time; per-directory enable/disable; savings ratio reported in isi statistics and InsightIQ dashboards
›Performance sizing — baseline IOPs per node type (F-series vs H-series vs A-series); latency SLA requirements; client concurrency and connection counts; protocol overhead comparison (SMB3 multichannel vs NFSv4 vs S3)
›InsightIQ — performance analytics platform; collects real-time and historical statistics for nodes, cluster, protocols, and individual clients; dashboards for latency, throughput, and IOPs; file system analytics: hot files report, top clients, and protocol breakdown
›HealthCheck — comprehensive cluster health assessment report: node status, drive health, network interface checks, quota utilisation, protection integrity, and firmware version compliance
›iostat — system-level disk I/O monitoring on each node; track read/write throughput and I/O wait per storage device; useful for identifying spindle vs NVMe bottlenecks
›netstat and iperf — netstat for active network connections and per-NIC statistics; iperf for bandwidth testing between nodes or clients; validate InfiniBand back-end and front-end Ethernet throughput against design targets
›SNMP monitoring integration — cluster SNMP MIBs for integration with Nagios, Zabbix, and enterprise SNMP managers; configure SNMP v2c/v3 communities; trap settings for proactive alerting on drive failures, node events
›SRS (Secure Remote Services) — proactive monitoring via ESRS gateway; remote diagnostics and log collection by Dell support; automated case creation on hardware failures; reduces MTTR
›Job Engine — system background jobs: AutoBalance (redistribute data after node add), FlexProtect (re-protect after drive/node failure), MediaScan (data integrity scan), SmartPoolsTree (apply file pool policies), SnapshotDelete; priority tiers (low/medium/high); impact policies limit CPU/disk impact; schedule windows; monitor running/paused/completed status via isi job status
Earn the D-PSC-DS-23 Dell Proven Professional certification and demonstrate your expertise in designing PowerScale scale-out NAS infrastructure, capacity sizing, advanced data services, and workload monitoring.