Dell-EMC Powerstore Technical Overview

PowerStore Technical
Overview
April 14, 2021
Presentation sections
• Hardware Overview • NVMe-FC
• PowerStore 1000T/X – 9000T/X Hardware • Storage Network Scaling
Overview • Volumes & Volume Groups
• PowerStore 500T Hardware Overview • Protection Policies
• PowerStore T Architecture Overview • Snapshots
• PowerStore X Architecture Overview • Thin Clones
• PowerStore Discovery • Block Remote Replication
• PowerStore Clustering • PowerStore metro node
• Resource Balancing • Virtualization Integration
• PowerStore Manager GUI • Serviceability
• CLI • Security Overview
• Rest API • Import External Storage
• Data Path • Services
• File
2 of 292 © Copyright 2021 Dell Inc.

For Presenters: What has been updated in the
presentation since the last version
• April update
– Content added:
▪ PowerStore 500T
▪ PowerStoreOS 2.0:
• Data Path enhancements
• Cluster support for PowerStore X
• NVMe-FC
• Storage Network Scaling
• Additional Import External Storage support
– Content deleted:
▪ System Limits section (this information can be found in the Support Matrix)

Hardware Overview

Overview
Terminology (Hardware) - PowerStore
PowerStore Term Definition Previous Terms
Component within the base enclosure that contains processors
Node Storage Processor (SP), Controller
and memory. Each appliance consists of 2 nodes.
Used to reference the enclosure containing both Nodes (Node A
Base Enclosure Disk Processor Enclosure (DPE), Array
and Node B) and 25x NVMe drive slots
Enclosures that can be attached to a base enclosure to provide
Expansion Enclosure Disk Array Enclosure (DAE), JBOD
additional storage in the form of SAS drives
Term used for solution containing a base enclosure and any
attached expansion enclosures. The size of an appliance could
Appliance N/A
be just the base enclosure or the base enclosure plus expansion
enclosures.
Multiple appliances in a single grouping. Clusters can consist of
Cluster one appliance or more. Clusters are expandable by adding more N/A
appliances (up to 4).
Module embedded to each node providing mgmt, host front-end
Embedded Module N/A
connectivity, SAS expansion ports, and selectable 4-port card
Card for each node that provides 4 ports for mgmt. and front-end
4-Port Card N/A
ports. Selectable in 25GbE and 10GbE Base-T options.

Overview
PowerStoreTM
• Overall family brand that encompasses all

models is called “PowerStore” PowerStore™
• “PowerStore T models” refers to SAN/NAS
only deployment mode of PowerStore
PowerStore T models PowerStore X models
• “PowerStore X models” refers to SAN and • PowerStore 500T • PowerStore 1000X
built-in ESX hypervisor deployment mode • PowerStore 1000T • PowerStore 3000X
of PowerStore • PowerStore 3000T • PowerStore 5000X
• PowerStore 5000T • PowerStore 7000X
• PowerStore 7000T • PowerStore 9000X
• From a marketing perspective, it is
• PowerStore 9000T
acceptable to omit “T” and “X” when
referring to a particular system
configuration
– i.e. “PowerStore 5000 models have these system
limits…”

Overview
PowerStore Models New
PowerStore PowerStore PowerStore PowerStore PowerStore PowerStore

500T 1000T/X 3000 5000 7000 9000
2x 12C @ 2.2G 4x 8C @ 1.8G 4x 12C @ 2.1G 4x 16C @ 2.1G 4x 20C @ 2.4G 4x 28C @ 2.1G
CPU per Appliance
Cascade Skylake Skylake Skylake Skylake Skylake
Memory per
192GB 384GB 768GB 1152GB 1536GB 2560GB
Appliance
NVRAM drives per
0 2 4
Appliance
Max Storage Drives
25 96
per Appliance
NVMe SCM,
Supported Drives NVMe SCM, NVMe SSD, SAS SSD
NVMe SSD
4x 25/10/1 GbE
Embedded Ports Optical (Optional) 4x 25/10/1 GbE Optical or 4x 10/1 GbE BaseT
2x10 GbE Optical
4x 32/16/8 Gb FC,
Support I/O Modules
4x 25/10/1 GbE Optical,
(2 slots per node)
4x 10/1 GbE BaseT
Supported Expansion
None 2.5” 25-Drive SAS SSD
Shelves

PowerStore drive support updates
PowerStore Update Drive Support List
PowerStore 1000T/X –
PowerStore 1000T/X –
PowerStore 500T 9000T/X
9000T/X
Storage Type Usage / Purpose GB 2.5” Base Enclosure 2.5” Expansion
2.5” Base Enclosure
(25 slot) Enclosure
(25 slot)
(25 slot)
User Data/Metadata 1,920 ✓
SAS SSD User Data/Metadata 3,840 ✓
User Data/Metadata 7,680 ✓
User Data/Metadata 1,920 ✓ ✓

NVMe SSD
User Data/Metadata 375 ✓ ✓
NVMe SCM User Data/Metadata 750 ✓ ✓

New in
User Data/Metadata 1,500 ✓ PowerStoreOS

2.0
NVMe NVRAM Cache 8 ✓
• All drive offerings are encrypted (SEDs)

• All drive offerings are FIPS certified except for NVMe NVRAM drives
PowerStore
PowerStore Manager
Cluster Overview
Appliance 1 Appliance 2 Appliance 3 Appliance 4
Dual Node
Base Enclosure Architecture
Expansion Scale Up
Storage with
Enclosure Extra Drives
Clustered
Mode Appliances
PowerStore PowerStore PowerStore PowerStore
Clusters Scale-Out with Appliances*
*Clustering is supported for PowerStore X appliances starting in PowerStoreOS 2.0

PowerStore 1000T/X – 9000T/X
Hardware Overview

Chassis Embedded Module
Base Enclosure (Rear View)

Node B
• 2 Nodes in inverted arrangement
Node A
• Each Node has Embedded Module
with: 4-Port Card 2X SAS
“Mezz 0”” Expansion I/O Module 1
– 1x 4-Port Card Slot – Mezz 0
I/O Module 00
– Embedded 2x1GbE Port (Mgmt/Service) 2X 1Gbe Ports Power Supply Unit
(PSU)
– Embedded SAS Expansion Ports (x2) 2X 1Gbe Ports
Power Supply
• Every appliance must contain a 4-Port Mezz 0 Unit (PSU)
– The 4-Port Card is used for connections such as cluster interconnect and management of the
appliance
– Customers will have 2 types of 4-Port Mezz cards to select from: Optical or Copper
• Each node has 2x I/O Module slots for configuring optional additional frontend ports
• Each node has a power supply unit (PSU) which can power both nodes if needed
Embedded Module
Overview
USB port
Mezz 0
Micro DB9
SAS ports 1GbE ports
The Embedded Module has the following components/connectors:
• One 4-port card (Mezz 0)
• One SAS controller, with two external MiniSAS HD ports
• One Broadcom 1GbE device with two RJ45 ports, in the field the service port will be used for
console access
• One Micro DB9 serial port, direct connect SP console (Not used by customers)
• USB port for system recovery in case dual M.2 failure and for password reset
4-Port Card
Overview
• 4-Port Cards are inserted in Mezz 0

slot
• 4-Port Cards are required and will be
available with the following
configurations:
– 4-Port 25GbE Optical
▪ 10Gb or 25Gb SFPs
• Can mix SFPs on same card
4-Port Mezz Card
▪ TwinAx
4-Port Card Option 1 4-Port Card Option 2
– 4-Port 10GbE/1GbE/100MbE BaseT
Optional
SFPs/TwinAx
• 4-Port Card is sometimes referred to
as “OCP” (Open Compute Project) 4-Port 25GbE Optical Card 4-Port 10GbE BaseT Card

I/O Modules
Connectivity Options
Slot 1 Slot 0
• Each node contains 2 I/O Module slots for added connectivity
• I/O Module configurations must match on both Nodes
• The following I/O Modules are optional for each appliance:
v Slot 0 Slot 1
– 4-Port 25GbE SFP based I/O Module

▪ Supports the following speeds based on installed Optional I/O Modules
SFP/TwinAx Cable
• 1GbE / 10GbE / 25GbE
▪ Supports 10GbE SFP, 25GbE SFP, Active and Passive I/O Module Option 1 I/O Module Option 2
TwinAx
SFPs/TwinAx
• No active TwinAx for 25GbE
– 4-Port 10GBaseT I/O Module
▪ Supports both 10GBaseT, 1GBaseT, 100MBaseT on 4-Port 25GbE SFP based I/O Module 4-Port 10GbE BaseT I/O Module
the same I/O Module
– 4-Port 32Gb Fibre Channel I/O Module
I/O Module Option 3
▪ Supports 16Gb FC SFP and 32Gb FC SFP
SFPs
• Both SFP types will auto negotiate to lower
speeds
(32Gb / 16Gb / 8Gb or 16Gb / 8Gb / 4Gb)
4-Port 32Gb Fibre Channel I/O Module

PowerStore Drive Support
NVMe NVRAM
• Every system has either 2 or 4 NVMe NVRAM Write Cache drives

depending on model
–Each NVRAM drive is 8GB in size and mirrored
–All writes to the system must be written to the NVRAM drives
prior to being acknowledged to the source
• There will be a maximum of 4 NVRAM drives depending on model
of the system and this is locked
–Customers cannot scale write cache
• NVMe NVRAM drives are self-encrypting, but not FIPS certified

NVMe SCM
• Customers have the option to use NVMe SCM drives in their

PowerStore system
• If only NVMe SCM drives are chosen, a minimum of 6 drives are
required per appliance
• NVMe SCM drives can now be mixed with NVMe SSD and SAS
SSD drives New in
PowerStoreOS 2.0
– When mixed, NVMe SCM drives will serve as a dedicated metadata tier, improving
performance with the low latency capability of NVMe SCM drives
– When mixed, NVMe SSD drives will server as the storage tier and a minimum of 6
NVMe SSD drives are required

NVMe SCM – System Configuration
6x Up to 21x 2 or 4
Minimum NVMe SCM NVRAM
Drives Drives drives

NVMe SSD
• Customers have the option to use NVMe SSD drives in their

PowerStore system
• If NVMe SSD is chosen, a minimum of 6 drives are required
per appliance
• The base enclosure with NVMe SSD drives can attach
expansion enclosure(s) with SAS SSD drives
– Note, in terms of ordering, the base enclosure must be completely filled
before expansion enclosures can be attached

NVMe SSD – System configuration (no expansion enclosures)
6x Up to 21x 2 or 4
Minimum NVMe SSD NVRAM
Drives Drives drives

NVMe SSD – System configuration (With expansion enclosure)
25x SAS
SSD
Drives
Up to 21x 2 or 4
NVMe SSD NVRAM
Drives drives

Summary
• All 25 drive slots support NVMe drives, SAS SSD drives are not supported in any
slot on base enclosure
• SAS SSD drives are only supported in attached expansion enclosures
Drive Type Location Purpose

NVMe NVRAM Slots 21-24 in base enclosure Cache
NVMe SCM Slots 0-20 in base enclosure User Data/Meta Data*

*When mixed with NVMe SSD,
NVMe SCM drives will only store
NVMe SSD Slots 0-20 in base enclosure User Data/Meta Data Meta Data
SAS SSD Slots 0-24 in expansion enclosure User Data/Meta Data

Power Supply Unit
Atlas * Poseidon *
• Supports 1800W High Line Only • Supports 2100W High Line
• Uses C13/C14 or C13/C20 cables only • Uses C19/C20 cables only

– C13/C14 is a more typical connector in the field
– C13/C20 used in our PowerStore racks that have primarily
C19 outlets
Poseidon Power Cord
Atlas Power Cords
* Low Line voltages (110 – 120 volts) require the use of a step-up transformer

PowerStore Country Specific Cables
C13 Country Specific Cables C19 Country Specific Cables
• Only used with Atlas PSU • Only used with Poseidon PSU
• Can be used with Base Enclosure for any country that has wall • Can be used with Base Enclosure for any country that has wall
outlets with 200-240 Volts outlets with 200-240 Volts
• Can be used with Tabasco Expansion Enclosure for any
country
• Requires >100 Volts
• Applies to following Base Enclosure models:

• Applies to following Base Enclosure models:
• PowerStore 5000X
• PowerStore 1000T / PowerStore 1000X • PowerStore 7000T / PowerStore 7000X
• PowerStore 3000T / PowerStore 3000X • PowerStore 9000T / PowerStore 9000X
• PowerStore 5000T
PowerStore 500T
Hardware Overview

PowerStore 500T
Base Enclosure (Front) Up to 25x NVMe
• 25x U.2 Drive Slots capable of supporting:

– NVMe SSD
– NVMe SCM
• Internal DRAM caching therefore no need for NVRAM drives

• 2U in height (rack units), 31.2” depth, 73.3 lbs fully loaded
– All PowerStore systems have the same depth

PowerStore 500T
Base Enclosure (Front) Up to 25x NVMe
• Each node will contain 1x Intel

S4510 240GB M.2 SATA drive
– Single SSD
• Each node will contain 96GB of

DDR Memory
• Each node will contain a single 12
core CPU (Intel 4214)
• PowerStore 500T platform
supports Unified or Block
Optimized only
– PowerStore T model

PowerStore 500T
Base Enclosure (Rear)
• Consist of 2x Nodes Embedded Module
– Node A – Bottom Half

– Node B – Top Half
Node B
• Each Node includes an “Embedded

Module” which contains: Node A
– 1x 4-Port Card (Optional) - “Mezz 0”

– 1x 2-Port Card (10GbE Optical ports) – “Mezz 1” 4-Port Card
I/O Module
Slot 1
“Mezz 0”
– Embedded 1GbE Port (x2)
– Embedded SAS Expansion Ports (x2) I/O Module
Slot 0
▪ Not used 2X 1GbE Ports
Power Supply
– Service ports for engineering use Unit (PSU)
2-Port Card
(Fixed 10GbE Optical)
“Mezz 1”

PowerStore 500T
Base Enclosure (Rear)
• Each node has 2x IO Module slots Embedded Module
for optionally adding more front end

ports – blanks shipped when no IO Node B
Modules are ordered
Node A
• Mgmt is done with the 1GbE ports

on the embedded module therefore 4-Port Card
I/O Module
Slot 1
“Mezz 0”
a 4-port card is not required for I/O Module
every PowerStore 500T appliance 2X 1GbE Ports
Slot 0
Power Supply
– Shipped with a blank bezel in the event a Unit (PSU)
2-Port Card
4-port is not ordered (Fixed 10GbE Optical)
“Mezz 1”
– NAS and Clustering not supported
without 4-Port Card

PowerStore 500T
4-Port Card (Mezz0)
• 4-Port Card (“Mezz0”) is used for cluster interconnect and front-end

connectivity
4-Port Card
• A 4-Port Card is optional. The following features are not supported “Mezz 0”
when a 4-Port Card is not ordered
– Unified Deployment mode (File services)
– Clustering
Top of Rack Switch
• 4-Port 25GbE SFP based Mezz Card (“Light Blade”)
– Supports 1GbE SFP to RJ45, 10GbE SFP, 25GbE SFP28, Active and
Passive TwinAx TwinAx Cables for Mgmt
▪ No active TwinAx for 25GbE and FE
– Mezz card will include 4x TwinAx cables to be used to connect appliance

to Top of Rack switch
Indus Backend
– The last two ports (ports 2 and 3) are reserved for future releases and not SFPs/TwinAx
available to customers
4-Port SFP based Mezz Card
(“Light Blade”)

PowerStore 500T
2-Port Card (Mezz1)
• 2-Port Card (“Mezz1”) is used for front-end connectivity and

replication
2-Port Card
• 2-Port Card is a fixed 10 GbE Optical card “Mezz 1”

PowerStore 500T
Connectivity Matrix
Slot Device Port 0 Port 1 Port 2 Port 3
iSCSI, iSCSI, iSCSI, iSCSI,

IO Module ClearskyX - 4x 25GbE
Replication Replication Replication Replication
iSCSI, iSCSI, iSCSI, iSCSI,
IO Module SnowslideX - 4x 10/1 GBaseT
Replication Replication Replication Replication
IO Module ColdspellX - 4x 32Gb FC FC, NVMe-FC FC, NVMe-FC FC, NVMe-FC FC, NVMe-FC
iSCSI, NAS, iSCSI, NAS,

4-Port Card (Mezz0) Light Blade Clustering, Clustering, Reserved Reserved
Replication Replication
iSCSI, iSCSI,
Fixed Ports (Mezz1) Fixed Ports - 10GbE SFPs N/A N/A
Replication Replication

PowerStore 500T
Drive Support
• All 25 drive slots support NVMe drives

• Expansion enclosures are not supported with PowerStore 500T
Drive Type Location Purpose

NVMe SCM Slots 0-24 in base enclosure User Data/Meta Data*
NVMe SSD Slots 0-24 in base enclosure User Data/Meta Data

*When mixed with NVMe SSD,
NVMe SCM drives will only store
Meta Data

Platform
Power Supply Unit
• PowerStore 500 supports two 1450W PSU models

• 1450W PSU
– Supports high-line power
• 1450W (100-120V) Native Low Line PSU

– Support low-line power
• Uses C13/14 or C13/C20 cables only

– C13/C14 is a more typical connector in the field
– C13/C20 used in our PowerStore racks that have primarily C19 outlets
• DC power is not supported

PowerStore T Architecture
Overview
(applicable to 500T – 9000T models)

Overview
PowerStore T Model
• PowerStoreOS installed directly on purpose built hardware

– 2U2N
– All NVMe Base Enclosure
• Unified Storage array
– SAN (FC/iSCSI)
– NAS (NFS/SMB/FTP/SFTP)
– vVol (FC/iSCSI)
• Active-Active architecture
– Each node has access to the same storage
– Active-optimized/Active-unoptimized front end connectivity

Overview
PowerStore T Storage Configuration
• PowerStore T has two user configured deployment modes

• Different deployment modes are called Storage Configuration
– Unified:
▪ Default storage configuration (factory state)
▪ Supports SAN, NAS, and vVol
▪ Resources shared between block and file components
– Block Optimized
▪ Alternate storage configuration (requires reboot)
▪ Supports SAN and vVol
▪ Resources dedicated to block components
• Storage Configuration is selected at time of Initial Configuration

– Changing Storage Configuration requires a factory reset
– Currently, multi-appliance clusters support a maximum of one Unified appliance.

Overview
PowerStoreOS
• Implemented through multiple docker

containers
• Docker
– An environment and set of specifications for
container based deployments
• Containerizing PowerStoreOS comes

with many advantages
– Easier serviceability
– Greater potential for integration

Networking
System Bond
• PowerStore T models contain a system bond by default

• Ports 0 & 1 of 4-Port Card are automatically bonded together in LACP mode
• This bond is essential to the networking configuration and cannot be removed
• System bond can provide high availability to cluster data and metadata traffic
• Link aggregation configuration on the switch side is optional
– If switches are not configured to support LACP, Ports 0 & 1 will operate in Active/Passive mode

Networking
Storage Network Scaling
• PowerStore T models support 32Gb FC I/O Modules for SAN and NVMe-FC connectivity
– Traditional block and vVol storage presented to external hosts over FC New in
PowerStoreOS
2.0
• iSCSI Storage network can be scaled up to support more iSCSI targets

• The following components are available for iSCSI port scaling:
– Ports 2 & 3 of 4-Port Card
– 4-Port I/O Modules
• No additional link aggregation can be configured

– System bond cannot be expanded
• Workflow:
1. Add additional IPs to the existing storage network (unless unused IPs exist)
2. Map unmapped ports to the additional IPs

Deployment
Configure Switches - Ethernet Switches (Dell)
• Dual ethernet switches are required to provided high availability

• Dell EMC PowerSwitches running OS10 Enterprise Edition (OS10EE)
– S4148F-ON
▪ 48 Fixed 10 GbE SFP+ ports
▪ 8 Fixed 50 GbE QSFP28 ports
– S4148T-ON
▪ 48 Fixed 10 GBASE-T ports
– S5248F-ON
▪ 48 Fixed 25 GbE SFP28 ports
▪ 2 Fixed 200 GbE QSFP28-DD ports

Deployment
Configure Switches - Ethernet Switches (Third-party)
• Customers can use third party switches

– Reference E-Lab for complete list of qualified switches
• Ethernet switches are used for a variety of network traffic

– Traffic is determined by model and deployment type
▪ iSCSI (PowerStore T and PowerStore X)
▪ Replication (PowerStore T and PowerStore X)
▪ Native Import (PowerStore T and PowerStore X)
▪ NAS (PowerStore T Unified)
▪ vMotion (PowerStore X)
▪ Management (PowerStore X)

Deployment
Configure Switches - Ethernet Switches - Topology
• Virtual Link Trunking interconnect (VLTi) (or third-party equivalent)

– Recommended configuration
– Creates a single logical switch out of two physical switches
– Automatic configuration of VLANs on VLTi interface
– Synchronized MAC tables support faster failover
– Supports link aggregation across two physical switches
▪ Best practice is to configure LACP on switch side to support the PowerStore T system bond
▪ Provides full bandwidth potential across system bond
• Alternative Layer 2 connectivity options

– Dynamic Link Aggregation Group (LAG), Static LAG, or reliable L2 uplinks
– Provides connectivity between two physical switches
– System bond will enter degraded mode
▪ Bond ports presented in active / passive states
▪ Bandwidth potential limited to 50% on system bond
▪ High availability remains
Deployment
Configure Switches – Out of Band Management
• PowerStore T model deployments require Out of Band (OOB) Management switch

connectivity
– 1x OOB Management switch is required
– 2x OOB Management switch is supported for high availability
– The document and this presentation outline 1x OOB Management
• OOB Management switch can be configured with or without a management VLAN

– Switch ports must support untagged native VLAN traffic for system discovery

Deployment Example Config
Cabling
• Management Network
– 1x Out of Band (OOB) Management switch
▪ 2x OOB Management is supported for HA
– Onboard 1GbE Ports
• Data Network
– 2x Top of Rack (ToR) Ethernet Switches
– Bonded 4-Port Card Ports 0 & 1
– Layer 2 interconnect
• Diagram depicts minimal required cabling for full

feature support
• In PowerStoreOS 2.0, only the management Management Network
network cabling is required to complete the initial Data Network
configuration wizard
PowerStore X Architecture
Overview
(applicable to 1000X – 9000X models)

Overview
PowerStore X Model
• ESXi runs directly on purpose built hardware

– 2U2N
– All NVMe Base Enclosure PowerStoreOS End User
– Dual-socket Intel Xeon architecture VM
• PowerStoreOS runs in a virtual machine

End User
– Referred to as the Controller VM VM
• Capabilities:
End User
– SAN (FC/iSCSI) VM
– vVol (FC/iSCSI)
– Embedded Applications (Virtual Machines)
– Clustering of multiple X model appliances is supported
New in
PowerStoreOS
2.0
Hypervisor
• Active-Active architecture
– Each node has access to the same storage
– Active-optimized/Active-unoptimized front end connectivity

Overview
ESXi
• VMware ESXi 6.7P04 installed directly onto each node

• PowerStore X model appliance contains ESXi Cluster of two ESXi hosts
– 1 ESXi host per node
• Requires existing vCenter and license for deployment

– VMware vSphere Enterprise Plus or vSphere Remote Office Branch Office license
– Customer can purchase license or use existing
• VMware components are automatically configured

– ESXi Cluster
– vSphere HA
– Distributed Virtual Switch
– Etc.
• Customer Virtual Machines will leverage PowerStore storage and data services
Overview
Controller VM
• PowerStoreOS runs inside of a Controller VM

– VMware Virtual Machine
– “Virtualized” instance of the PowerStore series OS
• One Controller VM fixed per node

– Will never fail-over to other node
• Reserves 50% of node resources

– CPU and Memory
• Key platform devices passed through directly to Controller VM

– Removes latency involved with ESXi layer
• Stored on M.2 device local to host and node

Overview
Diagram
passthrough passthrough
Primary Node Secondary Node

Overview
PowerStore X Model Capability Diagram
Server PowerStore X Model
PowerStore
ESXi Cluster
Storage FC/iSCSI FC/iSCSI iSCSI Internal

PowerStore Datastore

Networking
Teaming
• PowerStore X implements various port teaming configurations

– This provides high availability for network traffic
• PowerStore X requires 4-Port Card Ports 0 & 1 cabled at a minimum

– ESXi networking is configured to support cabling all 4 ports on the 4-Port Card
• The use of these ports by specific networks will be detailed

Networking
Port Mapping Table
REQUIRED OPTIONAL
4-Port Card Port 0 Port 1 Port 2 Port 3
Port Group PowerStore Network Uplink 2 Uplink 1 Uplink 4 Uplink 3
PG_MGMT Management, Initial Discovery Network Active Active Standby Standby
PG_MGMT_ESXi Management Active Active Standby Standby
PG_Storage_INIT1 Storage Unused Active Unused Unused
PG_Storage_INIT2 Storage Active Unused Unused Unused
PG_Storage_TGT1 Storage Standby Active Standby Standby
PG_Storage_TGT2 Storage Active Standby Standby Standby
PG_Storage_TGT3 Storage Standby Standby Standby Active
PG_Storage_TGT4 Storage Standby Standby Active Standby
PG_vMotion1 vMotion Standby Standby Standby Active

Deployment
Switches
• Dual ethernet switches are required to provided high availability
• Dell EMC PowerSwitches running OS10 Enterprise Edition (OS10EE)

– S4148F-ON
▪ 48 Fixed 10 GbE SFP+ ports
▪ 2 Fixed 40 GbE QSFP+ ports
– S4148T-ON
▪ 48 Fixed 10 GBASE-T ports
▪ 2 Fixed 40 GbE QSFP+ ports
– S5248F-ON
▪ 48 Fixed 25 GbE SFP28 ports
▪ 2 Fixed 200 GbE QSFP28-DD ports
• Customers can use third party switches

– Reference E-Lab for complete list of qualified switches
Deployment
Switches - Configuration
• Virtual Link Trunking interconnect (VLTi) (or third-party equivalent)

– Recommended configuration
– Creates a single logical switch out of two physical switches
– Automatic configuration of VLANs on VLTi interface
– Synchronized MAC tables support faster failover
• Alternative Layer 2 connectivity options

– Dynamic Link Aggregation Group (LAG), Static LAG, or reliable L2 uplinks
– Provides connectivity between two physical switches
• Reference Configuring Switches and External Networks

Deployment
Configure Switches - Ethernet Switches - Management
• PowerStore X model deployments do not use an Out of Band (OOB) Management

switch
• Management is in-band on 4-Port Card Port 0 & 1
– This ensures high availability for management traffic to vCenter
• Switch ports must support untagged native VLAN traffic for system discovery

Deployment Example Config
Cabling
• Data Network
– 2x Top of Rack (ToR) Ethernet Switches
– Bonded 4-Port Card Ports 0 & 1
– Layer 2 interconnect
• Diagram depicts minimal required cabling

– Cluster will alert on not having dual redundant switches
Data Network

PowerStore Discovery
• Discovery Utility
• Static IP Discovery

Once you get a system, what are the next steps?
2
1 Connect workstation to
3
the same L2 network and
Once the rack and stack of
discover the system Deploy a new cluster
the appliance is completed
• Cabling completed (To be covered in • Run through the Initial
a later presentation) • Use the PowerStore Discovery Configuration Wizard
Utility to discover appliance(s) (ICW)
• Power on appliance(s)
• Select appliance(s) and click Create
Cluster

Discovery Utility

PowerStore Discovery Utility
• PowerStore Discovery Utility is a Dell EMC open-source discovery tool

– Available on Dell EMC Support Site
– No installation required
– Utilizes Zero-configuration networking (zeroconf) to discover appliances
• Use cases:
– Initial discovery
▪ Discovery of unconfigured appliances
– Adding appliances
▪ Adding appliance to an existing configured cluster
– Finding already configured cluster(s)
▪ Once a cluster is configured, it can be discovered again from the tool

PowerStore Discovery Utility
Pre-requisites (Physical)
L2 Network
• Cabling required for
discovery
– PowerStore T uses the 1Gb
Onboard Management port
PowerStore T
– PowerStore X uses first two
ports of the 4-Port Card
• Workstation (laptop or PowerStore X

host/VM) has connection
via Ethernet cable to the
same L2 native network as Workstation
the appliance(s)
– Additional cabling required for
deployment and host access

Initial Configuration
Initial Configuration

Manual discovery procedure – Service Port
• You can also discover the system by connecting
to the service port: PowerStore appliance
– Connect your workstation's Ethernet adapter to the service
port on node A of the base enclosure.
– Set the IP address of your workstation to 128.221.1.249 with
netmask 255.255.255.0, and no gateway address defined.
– Ensure that you can ping the IP address of node A's service
LAN port (128.221.1.250).
– In a web browser, go to https://128.221.1.250
– Log on to PowerStore Manager and begin the initial
configuration process using the default credentials
(admin/Password123#). Workstation
– Disconnect the workstation from the service port after the
initial configuration process is complete.

PowerStore Static IP Discovery

Overview
What’s New
• There are now three methods to discover PowerStore to run through Initial
Configuration Wizard (ICW) of a PowerStore system
PowerStore T
1 0
x4
B
B B
3 2 1 0 3 2 1 0
A 3 2 1 0
1GbE
12
10GbE
A
SS
2200W B 3 2 1 0
0
1
1GbE
1GbE
2200W
1
0 1 2 3 B
0
10GbE
SS
1GbE
12
0 1 2 3 A 0 1 0 1
2 3 2 3
A B A
x4
0 1

Overview
What’s New
– Service port (recommended)
▪ Direct connect to PowerStore and launch ICW
PowerStore T
1 0
x4
B
B B
3 2 1 0 3 2 1 0
A 3 2 1 0
1GbE
12
10GbE
A
SS
2200W B 3 2 1 0
0
1
1GbE
1GbE
2200W
1
0 1 2 3 B
0
10GbE
SS
1GbE
12
0 1 2 3 A 0 1 0 1
2 3 2 3
A B A
x4
0 1
IP: 128.221.1.10 IP: 128.221.1.250/251

Overview
What’s New
– Discovery tool
▪ Leverage discovery tool to launch ICW
PowerStore T
1 0
x4
B
B B
3 2 1 0 3 2 1 0
A 3 2 1 0
1GbE
12
10GbE
A
SS
2200W B 3 2 1 0
0
1
1GbE
Network
1GbE
2200W
1
0 1 2 3 B
0
10GbE
SS
1GbE
12
0 1 2 3 A 0 1 0 1
2 3 2 3
A B A
x4
0 1
IP: 169.254.0.5 IP: 169.254.x.x

Overview
What’s New
– Discovery tool
▪ Leverage discovery tool to launch ICW
– PowerStore static IP discovery

▪ Supported on both PowerStore T and PowerStore X model appliances
▪ Open web browser to launch ICW with a well defined static IP address PowerStore T
1 0
x4
B
B B
3 2 1 0 3 2 1 0
A 3 2 1 0
1GbE
12
10GbE
A
SS
2200W B 3 2 1 0
0
1
1GbE
Network
1GbE
2200W
1
0 1 2 3 B
0
10GbE
SS
1GbE
12
0 1 2 3 A 0 1 0 1
2 3 2 3
A B A
x4
0 1
IP: 169.254.0.5 IP: 169.254.0.10

169.254.0.20
169.254.0.30
169.254.0.40
72 of 292 © Copyright 2021 Dell Inc. 169.254.0.50
How Does It Work
Two Phase Login
• Launch web browser to static IP addresses http://169.254.0.10

– Backup IP addresses 169.254.0.20, 169.254.0.30, 169.254.0.40, 169.254.0.50 are also available and
point to the same location
– Login with default credentials
– A list of unconfigured appliances will be presented
– Select appliance to be configured and click “Go To Initial Configuration”
• Launch initial configuration

– A new ICW session is created redirecting the user to the Zeroconf IP of the selected appliance
– Login with credentials
– Complete ICW and cluster configuration

PowerStore Clustering

Clustering New in
Details PowerStoreOS
2.0 – X Model
Clustering
• PowerStore T & X models support clustering

– You can mix different appliance models (all T or all X, no T & X combination in the same cluster)
• Maximum size of 4 appliances

Clustering
What’s New in PowerStoreOS 2.0
Also new in
PowerStoreOS
2.0
• Automatic primary appliance selection

– Create cluster utilizes the most effective appliance as primary appliance
• Create cluster roll forward

– Multi appliance clusters will attempt to continue with creating the cluster as a best effort in the event of
a single appliance failure
• Create cluster REST API

– REST interface for creating a cluster and adding an appliance

Clustering
Primary Appliance
• Each Cluster contains a Primary appliance that runs all control path services
• Primary appliance specific services include:
– Global Management IP
– Primary Management DB
– Cluster high availability (pacemaker)
• All other appliances are standby appliances from primary control path perspective
– They run a subset of control path services to manage themselves and communicate with primary
– These appliances still serve I/O, only standby in regard to Primary appliance specific services
• Primary appliance specific services will failover to standby appliance if needed

– Effectively creating new Primary appliance

Clustering
Primary Appliance
• Primary appliance can be configured in Block optimized or Unified mode

– Unified mode will support file services
– All additional appliances are configured automatically as Block optimized
• The following Cluster configurations are supported:

– Unified Cluster
▪ Primary appliance is deployed as Unified storage configuration running file services
▪ Appliances 2-4 (depending on cluster size) are deployed as Block optimized
– Block optimized Cluster
▪ All appliances are deployed as Block optimized
• Cannot convert Block optimized Cluster to Unified

– Cannot add a Unified appliance to a Block optimized Cluster
• File resources and services will always run on the primary appliance of a Unified Cluster
– File does not failover to a new Primary appliance
– File is highly available on the Primary appliance
Clustering
Resources
• Individual resources reside on a

single appliance only
– Volumes
– vVols
– NAS Servers
– File Systems
• Data for a single resource is not

striped across multiple appliances
• Resource balancing and migration will
be handled in a future presentation

Clustering
Creation
• How to create and modify a Cluster

– Initial Configuration (1-4 appliances)
– Add Appliance
– Remove Appliance
• Cluster creation can only be done via the GUI

– REST and CLI are not supported

Resource Balancing

Overview
What is Resource Balancer?
1-appliance Cluster
• The ability of PowerStore to
use analytics to balance
storage resources (volumes)
Appliance 1
? New
Volumes
– Which node to assign a new

volume to on an appliance

Overview
4-appliance Cluster
• The ability of PowerStore to
use analytics to balance Appliance 1
storage resources (volumes)
– Which node to assign a new
volume to on an appliance
– Which appliance to assign a new Appliance 2
volume to in a multi-appliance
cluster ? New
Volumes
Appliance 3
Appliance 4

Overview
4-appliance Cluster
• Appliance assignments
determined by: Appliance 1
– Current storage space utilization
› Storage trends and forecasts
– System limits
› Max volumes per appliance / Appliance 2
volume group
– Appliance status and health
› Offline, failures, read-only (100%
? New
Volumes
full) Appliance 3
– Performance metrics are not
considered
– Resource Balancer does not
proactively or automatically move Appliance 4
existing volumes from one
appliance to another
Overview
What is Resource Balancing?
4-appliance Cluster
• The ability to migrate storage
resources between
Appliance 1 90%+
appliances in a cluster
– Manual migration
– Assisted migration
Appliance 2 ?
• Leverage capacity
monitoring, forecasts, and
alerts with suggested Appliance 3
?
?
remediation options
Appliance 4
?

Volume Placement
Placement Options
• Placement of new storage volumes on a multi-

appliance cluster
– Automatic or manual
• Automatic placement of new volumes (default)

– Placement determined by the system
– Governed by metrics listed earlier (Slide 65)
• Manual placement of new volumes

– The administrator chooses the desired appliance

Manual Migration
Remediation for Capacity Alert
• 2-appliance cluster
• Appliance 1 is over 90%

full
• Appliance 2 has lots of

storage capacity
• Event is recorded as a
major alert
• Alert details suggests

remediation options
• Remediation in this
example: manually move
a volume

Assisted Migration
Remediation for a Time-to-Full Alert
• 2-appliance cluster
• Appliance 1 is forecast to run out

of space in 8 days
• Appliance 2 has lots of storage

capacity
• Event is recorded as a major alert
• Alert details suggests assisted

migration, cleanup, or add drives
• Remediation in this example: move

volumes with assisted migration

PowerStore Manager GUI

PowerStore Manager GUI Overview
What is PowerStore Manager?
• HTML5-based GUI for managing PowerStore

clusters
• No Java required
• Supported with popular browsers: Chrome,
Firefox, Internet Explorer, Edge, Safari
• Simple, modern, intuitive management and
monitoring interface
• No client install required
• Connect by HTTPS
– HTTP or HTTP redirect to HTTPS is not
supported
Compute and Storage
• Configure a
new host server
under Compute
• Provision and
map a new
volume to the
host server
under Storage

Compute
• On the menu bar, click

Compute
• From the drop down,
click Hosts & Host
Groups
• Click Add Host
• Provide a host name,
OS, description and
click Next

Compute
• Select iSCSI or
Fibre Channel and
click Next

Compute
• Select the host

initiators and
click Next

Compute
• Review the
summary and
click Add Host

Compute
• New host is
added

Storage
• Click Volumes
List or Storage to
configure a volume
for the host
• In this example:
click Storage and
to the right of
Volumes in the
drop down

Storage
Create Volumes
wizard
• Enter volume
properties and
click Next
• The default
Performance
policy is
medium, but can
be changed

Storage
• Select a host or
host group
– Set LUN ID or
allow automatic
generation
– Use LUN 0 for
boot-from-SAN
• Click Next
• Review the
summary and click
Create

Storage
• Click Refresh
• Volume now listed
• Complete steps on
the host to
initialize and
format the new
volume
• Configure MPIO
settings on the
host

CLI

Introduction
PowerStore CLI
• Standalone Client
– Translates CLI into REST API calls in the background
• CLI still used frequently for administration

– Easy to use, OS commands can be used to filter
• Allows batch jobs without programming skills

• Available for Windows and Linux
– RPM for SLES 12, RHEL 7.x, Oracle Linux 6.5, 7
– Windows 7,8,10 / Server 2012, 2012R2, 2016
– Requires administrative privileges for installation
• Objects almost identical to REST-API

Using PowerStore CLI
Start using PSTCLI
• Option “-h/-help” shows additional information

[user@host ~]$ pstcli -h
Enter interactive session.
pstcli [-d <address>] [-port <number>] [-u <user_name>] [-p <password>]
[-ssl { interactive | reject | accept | store }] [-header] -session
Perform an action on an object on the destination system.

pstcli [-d <address>] [-port <number>] [-u <user_name>] [-p <password>]
[-ssl { interactive | reject | accept | store }] [-header] <object> <action> [<qualifiers>]
Get detailed help on client options:

pstcli help { -session | CMD | -version | -save_cred | -removeCred | -removeAllCreds | -default |
-certList | -certDel | -certClear | -clearCache }
Additional help topics will be provided once you connect to the remote server. Please provide destination
address to obtain remote server command help.

REST API

Functionality
• The REST API allows you to interact with PowerStore Management functionality,
including:
– System settings and monitoring
– Host and remote system connections
– Network settings
– Storage management
– Data protection
– Support configuration
• Presents a single, consistent interface to manage

• SSL encryption allows secure connection from client to system

SwaggerUI
Swagger UI – https://<PowerStoreClusterIP>/swaggerui

Data Path

PowerStore Data Path
Overview
• A PowerStore’s read and write path depends on the system model

– Described in detail in later slides
• PowerStore’s data path includes hardware and software algorithms working

together to accept and store user as efficiently as possible
• PowerStore’s Dynamic Resiliency Engine (PowerStore DRE) automatically
consumes the drives within an appliance and creates appropriate redundancy
using all the drives
– Space is consumed from the drives to store data after it passes through deduplication and
compression

PowerStore 1000T/X – 9000T/X Data Path
Data Layout – Storage Devices
• The software operating environment is

installed on (2) M.2 devices within each
node
– No “VAULT” System Drives
• A portion of the DRAM Memory is used

as a read cache
• Dual ported NVRAM Write Cache is
accessible by both nodes
– No mirroring write cache data to peer controller
• Hardware offloaded compression engine

– Quick-Assist Technology

Data Layout – Storage Devices – NVRAM Devices
• Cache Mirror Mirror

– 2 or 4 Devices
– Write Cache & TX Data
NVRAM Device
NVRAM Device
NVRAM Device
NVRAM Device
Data Drive
Data Drive
– Data@Rest Encryption
• Battery Protected
...
– Node BoB
19 20 21 22 23 24
A B

I/O Flow – Write Path
1. The appliance receives a Write on

Node B
2. The write request is stored in the
DRAM memory and NVRAM Write
Cache
3. A token is passed between the nodes
to update the peer that a new write
has been received for a particular
location
4. The system acknowledges the host

I/O Flow – Read Path
1. The system receives a Read

operation
2. If the data is in the DRAM cache, fulfill
the Read request
3. If the data is not in the DRAM cache,
copy the data from the drives

I/O Flow – Late Bind
After the Host has been acknowledged, the

system still processes the Write request in the
background
1. Write request is deduplicated and
compressed
2. Continue to fill a 2MB stripe of compressed
and deduplicated Write requests
3. Once the 2MB stripe is full, destage to drives

PowerStore 500T Data Path
Data Layout – Storage Devices
• The software operating environment is

installed on a M.2 device within each
node
– No “VAULT” System Drives
• A portion of the DRAM Memory is used

as a read/write cache
• Hardware offloaded compression engine
– Quick-Assist Technology

I/O flow – Write path
1. The appliance receives a Write

on Node B
2. The write request is stored in the
node’s DRAM memory
3. The data is mirrored to the peer
node’s DRAM memory
4. The system acknowledges the
host

I/O Flow – Read Path
1. The system receives a Read

operation
2. If the data is in the DRAM cache, fulfill
the Read request
3. If the data is not in the DRAM cache,
copy the data from the drives

I/O Flow – Late Bind
After the Host has been acknowledged, the

system still processes the Write request in the
background
1. Write request is deduplicated and
compressed
2. Continue to fill a 2MB stripe of compressed
and deduplicated Write requests
3. Once the 2MB stripe is full, destage to
drives

PowerStore - Global Deduplication
Overview
• Deduplication savings within a PowerStore appliance works across nodes within

an appliance
– Data entering a node can be compared and deduplicated against data received on the peer node
– Potentially increases the amount of savings that can be achieved within an appliance
• The fingerprint cache is not mirrored across the nodes

– Each node contains a fingerprint cache containing unique fingerprints created on that node
▪ When data enters a node and a new/unique fingerprint is generated and that node owns that fingerprint
– To achieve global deduplication, fingerprint compares will occur across nodes using internal links
– By not mirroring the fingerprint cache, more fingerprints can be held in the appliance’s memory
Deduplication utilizes a 4 KB granularity

PowerStore - Fingerprint Cache Size
Overview
• The size of the fingerprint cache is based on the capacity of the appliance when
the software boots on the system
– The fingerprint cache does not expand as drives are added to the appliance
– This size is not exposed to the user
• System memory and drive capacity is used to create a fingerprint cache

– The fingerprint cache spans the memory within a node and the drives
– Reduces the usable capacity of the system
• Goal: Increase the amount of deduplication that can be achieved by tracking a

larger number of fingerprints
– A larger fingerprint cache allows for a greater chance of deduplication occurring

PowerStore – Persisted Fingerprint Cache
Overview
• The contents of the fingerprint cache within system memory is updated to a backup
location on the data drives periodically
– This operation is not configurable and is hidden from the user
• The periodic update process is based on the number of entries that are different
than the backup location
– Only the new entries are updated to the backup location when an update occurs
– The threshold varies by model
• As the fingerprint cache update process is based on a threshold, some new

fingerprints can be lost if the node reboots
– Deduplication savings is not lost if a fingerprint is lost
– The ability to deduplicate to that piece of data is lost

PowerStore Dynamic Resiliency Engine
Overview
• Within PowerStore’s Dynamic Resiliency Engine (DRE), all drives within the
system are automatically consumed within an appliance and the appropriate
amount of redundancy is applied
– Proprietary algorithms are used to store and protect data within the system
– Resiliency sets are use as fault domains to improve the reliability while minimizing spare space
overhead
▪ Resiliency sets are also known as fault resiliency sets in customer facing documentation
– Having multiple failure domains increases the reliability of the system
• Drive fault tolerance is the amount of concurrent drive failures, per resiliency set,
that a system can sustain without causing a Data Unavailable/Data Loss (DU/DL)
situation
– The protection scheme within the resiliency set defines how many failures can occur

Overview (Continued)
• In the PowerStoreOS 2.0 release, the drive tolerance level can be set to single
drive failure or double drive failure during the initial configuration of an appliance
• Initial configuration could be initial cluster creation or when the appliance is being
added to an existing cluster
• Configuring the drive tolerance level sets the data protection tolerance level for all
resiliency sets created within the appliance
• The drive tolerance level is set for the lifetime of the appliance and cannot be
changed without a non-data-in-place factory reset
– Pre-2.0 systems utilize single drive failure protection
• Different appliances within a cluster can have different tolerance levels

Tolerance Level: Single Drive Failure vs. Double Drive Failure
• Tolerance Level: Single Drive Failure

– Default value
– Single drive parity protection is used within each resiliency set for user data
– Up to one simultaneous drive failure per resiliency set without encountering DU/DL
– Metadata and other mirrored data uses 2-way mirroring
• Tolerance Level: Double Drive Failure

– Dual drive parity protection is used within each resiliency set for user data
– Up to two simultaneous drive failure per resiliency set without encountering DU/DL
– Metadata and other mirrored data uses 3-way mirroring
– Reduced capacity in similar configurations when compared to single

Resiliency Sets & Supported configurations
• Tolerance Level: Single Drive Failure

– Resiliency set maximum drive count = 25 drives (Same as previous releases)
– Each resiliency set reserves one drive worth of spare space
– 4+1 and 8+1 widths supported; width set based on drive count at initialization
▪ 4+1 requires a minimum of 6 drives
• Tolerance Level: Double Drive Failure

– Resiliency set maximum drive count = 50 drives
– Each resiliency set reserves one drive worth of spare space
– 4+2, 8+2, and 16+2 widths supported; width set based on drive count at initialization
• Widths do not change as drives are added to the system

Benefits
• Shorten Rebuild Times

• Reduce Drive Overhead
• Mix Drive Capacities
• Single-Drive Granularity Scale-Up
• Distributed Spare

PowerStore Data Path
PowerStore DRE – Drive Types
• Mixing Drive Types and sizes
D0 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 D11

Architecture
Resiliency Sets & Supported configurations – Tolerance Level: Single Drive
Node 1 Node 2
8+1
User User User User User User User User User User User User
Data0 Data0 Data0 Data0 Data0 Data0 Data0 Data0 Data0 Data1 Data1 Data1
User User User User User User User User User
Data4 Data4 Data4 Data4 Data4 Data4 Data4 Data4 Data4
Spare Spare Spare Spare Spare Spare Spare Spare Spare Spare Spare Spare
Drive Count = 12 127 of 292 © Copyright 2021 Dell Inc.

Architecture
Resiliency Sets & Supported configurations – Tolerance Level: Double Drive
Node 1 Node 2
8+2
User User
Data4 Data4
Spare Spare Spare Spare Spare Spare Spare Spare Spare Spare Spare Spare
Drive Count = 12 128 of 292 © Copyright 2021 Dell Inc.

PowerStore Data Path – Performance Policies
I/O Priority mechanism
• PowerStore performance policies allow a user to define the I/O priority for volumes
within an appliance
– Can be customized at time of creation or can be changed at any time
– This is supported only for block volumes at this time
– This is not a limits based policy approach rather it is a shared based I/O prioritization mechanism
– When in affect, I/O for volumes set to High will receive priority over resources set to Medium or Low
• The performance policy setting on a volume can be set to Low, Medium or High
– Medium is the default setting at resource creation
• The performance policies only take affect when there is contention of resources

File

PowerStore File Overview
• PowerStore File enables clients to access data over file protocols:

– Server Message Block (SMB)
– Network File System (NFS)
– File Transfer Protocol (FTP)
– SSH File Transfer Protocol (SFTP)
• Currently File Replication is not available

• File is only available on PowerStore T model appliances
– Currently PowerStore X model appliances do not support File

Block Optimized or Unified
• PowerStore T model appliances can be configured as Block Optimized or Unified

(block and file)
– Selection determines resource allocation on the appliance
– PowerStore X model appliances do not have this option as they do not support NAS
• Must be decided during initial configuration

– Select Unified if there is a chance that you will need file functionality
– No option to convert between Block Optimized and Unified afterwards
• PowerStore Sizer will have performance details for each configuration

– Block-optimized systems have slightly higher block IOPS potential
• If Unified is selected, NAS installation is performed automatically after cluster

creation completes

Unified Appliance
• File functionality is only available on the primary appliance in the cluster

– Remaining appliances are configured as Block Optimized
• Only the capacity on the primary appliance is available for File

– Capacity available on other appliances within the same cluster can be used for volumes and vVols
• Both nodes on the primary appliance are used for File

– Active/active architecture enables load balancing and high availability

NAS Servers
• NAS servers enable access to the data on file systems

– Contains protocol and environmental configuration
– Required before creating file systems
• NAS servers are used to enforce multi-tenancy

– NAS Servers are logically segregated from each other
– Clients of one NAS Server do not have access to data on other NAS Servers
– IP multi-tenancy is not available
• Each NAS server has its own independent configuration

– E.g., DNS, LDAP, NIS, interfaces, protocols, etc.

Anti-Virus
• Shared bank of virus-checking

servers
• Can deploy multiple vendors’ engines
Client Virus-Checking concurrently
Server
• Virus-checking server only reads part
of files
• File access is blocked until it is
checked
– Scan after update
– Scan on first read
– Automatic access-time update
• Notification on virus detect

Supported Protocols
• NFS
– NFSv3
– NFSv4 - 4.1
– Secure NFS
• SMB – Standalone or Domain Joined

– SMB1
– SMB2
– SMB3 – 3.1.1
• Multiprotocol - Access using both SMB and NFS simultaneously

– Automatically enabled when both the SMB and NFS protocols are enabled on the NAS Server
• FTP/SFTP

NAS Server High Availability
• In the event of a PowerStore node failure, NAS Servers automatically failover from
one NAS node to the other
– Failover generally completes within 30 seconds to avoid host timeouts
– NAS Servers are automatically moved to the peer node during NDU
– Failback is a manual process
• New NAS Servers are automatically assigned round-robin across nodes

– All file systems associated with the NAS Server are served by the NAS Server’s current node
• Current Node
– The node that the NAS server is currently running on
– Changing this moves the NAS server to run on a different node
• Preferred Node
– The node that the NAS server should be ideally running on
– Acts as a marker that is based on the round-robin algorithm
NDMP Backups
• NDMP – Network Data Management Protocol

– A backup and recovery protocol used to transport data between NAS and backup systems
• Components
– Primary Storage – Source system to be backed up, such as PowerStore
– Data Management Application (DMA) – Backup application that orchestrates the backup sessions,
such as Dell EMC NetWorker
– Secondary Storage – The backup target, such as Data Domain
• PowerStore supports 3-way NDMP

– 3-way NDMP transfers both the metadata and backup data over the LAN
– 2-way NDMP is not supported
• Both full and incremental backups are supported

File Systems
• A file system can be created once a NAS Server is available

– Once created, file systems cannot be moved from one NAS Server to another
• The file system creation wizard prompts for:

– NAS Server
– File System Details
▪ Name
▪ Description (optional)
▪ Size
– NFS Export Details (if enabled on the NAS Server)
▪ Name
▪ NFS access configuration
– SMB Share Details (if enabled on the NAS Server)
▪ Name
▪ Advanced SMB options
– Protection Policy
SMB Options
• Advanced SMB Settings (File System) • Advanced SMB Settings (Share)

– Sync Writes Enabled – Continuous Availability (SMB3)
– Oplocks Enabled (Default) – Protocol Encryption (SMB3)
– Notify on Write Enabled – Access-Based Enumeration
– Notify on Access Enabled – Brach Cache Enabled
– Offline Availability
▪ None (Default)
▪ Manual
▪ Documents
▪ Programs
– UMASK (022 Default)

NFS Options
• Minimum Security – Minimum security allowed when connecting to a NFS export

– Sys
– Kerberos
– Kerberos with Integrity
– Kerberos with Encryption
• Access Levels – Can be set for Default Access or individual hosts

– No Access
– Read/Write
– Read-Only
– Read/Write, Root Only
– Read-Only, Root Only

File System Shrink and Extend
• File systems can be shrunk and extended at any time

– Cannot shrink to a value lower than the Used size
• Shrink and extend operations take effect immediately

– Refresh the client to see the size update
• Minimum Size: 3GB
• Maximum Size: 256TB

File System Quotas
• Quotas - Provides the ability to regulate capacity consumption

• User Quotas – Limits the capacity consumed by an individual user on the file
system
– All users are identified by their UNIX User ID
• Tree Quotas – Limits the capacity consumed on a specific directory on the file
system
– All files in the directory and subdirectories contribute towards the limit
• Default Quotas – Applied to all users on the file system automatically

– Ability to configure exceptions to the default
• Can create user quotas in a tree quota

– Limits capacity by users in a specific directory

NVMe-FC

NVMe-FC front end connectivity
Overview
• NVMe-FC stands for Non-Volatile Memory express over Fabrics

– In the first release of PowerStore, NVMe support was only in the drive level (Back-End)
• In the PowerStoreOS 2.0 release, NVMe-FC is supported on both PowerStore T and

PowerStore X models
– New Protocol for hosts to access storage systems across a network fabric using NVMe protocol (Front-End)
– Using Fibre Channel (NVMe)
PowerEdge PowerStore
HBAs
Switch

Benefits
• Highly scalable and parallel

• Better efficiency
• End-to-End NVMe Model
• NVMe drive level
• NVMe extended over the network fabric

Connectivity Considerations
• NVMe-FC is over Fibre Channel (FC)

– Front end connectivity 32Gb FC I/O modules
– NVMe-FC will be supported on any 32G FC switches running Brocade FOS 8.1.0 or later and Cisco
NX-OS 8.1(1) or later
• Fibre Channel front end port simultaneously supports SCSI and NVMe access
– User can choose whether to use both protocols on same port or use each protocol on separate ports
– Always support both protocols with no option to disable one of them
• Host Bus Adapter (HBA) supports 32Gb Gen6/7

– Marvell QLogic QLE series
– Broadcom Emulex LPe series
• NVMe Expansion shelf not supported in the PowerStoreOS 2.0 release

Connectivity Setup
• Users can setup NVMe host through:

– PowerStore Manager
– PowerStore CLI
– REST API
• Connectivity flow:
– Set up Fibre Channel Front End ports (zoned)
– Create Host or Host Groups and select NVMe as protocol
▪ Add initiator(s)
▪ nqn. is the NVMe identifier similar to the iqn. for iSCSI
– Create Volume/Thin Clone or Volume Groups
▪ Not support with vVol
– Map the NVMe Host to the Volume(s)

Add Host – NVMe Initiator option

Add Host – NVMe Host Initiator

Add Host – NVMe

Host Groups – NVMe Protocol

Volumes
• A Volume is not tagged as NVMe or SCSI upon creation

• A Volume can be only attached to NVMe host or SCSI host
– Cannot be attached to different protocol at the same time
– Can switch between the two protocols but requires unmap and then map
• A Volume with NVMe host can still do Volume migration

• Upon Volume creation NVMe Unique IDs are allocated (in addition to SCSI wwn):
– NSID – Volume ID on host perspective
– NGUID – NVMe Global Unique Identifier (equivalent to SCSI wwn)
– Both IDs assigned internally by the system
▪ Not user configurable

Volume Creation – NVMe Protocol

Volume Creation

Supported Operating Systems - Host
• VMware ESX 7.0

• RHEL 7.6-7.8, 8.0-8.3
• SLES 12 SP4-SP5, 15


Networking
• In PowerStoreOS 1.0, a single storage network could support a single

VLAN or subnet.
• PowerStoreOS 2.0 now supports multiple subnets over the Storage
Network
– Support for multitenancy and traffic isolation New in
PowerStoreOS
– PowerStore supports up to 8 storage networks of IPv4 and /or IPv6 addresses 2.0
– PowerStore supports up to 8 storage networks per interface of IPv4 and /or IPv6
addresses
– Improved GUI workflows to create networks and assign interfaces

Networking
Storage Network Scaling GUI Workflow
• Navigate to the settings network IP page

• Click on the storage tab
• Create a new storage network
• Map the network to desired interfaces to appliance(s) in the cluster

Add Storage Network
Demo

Add Storage Network
Demo

Add Storage Network
Demo

Add Storage Network
Demo

Add Storage Network
Demo

Add Storage Network
Demo

Add Storage Network
Demo

Add Storage Network
Demo

Volumes & Volume Groups

Block Resources
Storage
• Volumes
• Volume Groups

Volume Groups
Overview
• A Volume Group is a logical container for a group of volumes or volume thin clones
– Provides a single point of management for one or more resources
Volume Group 1 Volume Group 2

App1Volume-1 App1Volume-2 App2Volume-1 App2Volume-2
Application 1 Application 2
Volume Groups
Considerations
• A Volume can only belong to one Volume Group at a time

• All Volumes must reside on the same appliance
– To utilize a Volume Group you must plan Volume creation accordingly
• If a Protection Policy is assigned to a Volume Group, you cannot assign a

Protection Policy to an individual resource within the group
• Host/Host Group Mapping cannot be done at the Volume Group level
– Host access must be configured by other means
• Single volume restore operations are only allowed when “write order consistency”
is disabled on the Volume Group

Protection Policies

Protection Policies
Overview
• A Protection Policy is a set of user defined rules used to establish local or remote
data protection across storage resources
– Users do not configure snapshot schedules or replication sessions on a resource, but
rather assign a Protection Policy to it
• A Protection Policy consists of rules which define what level of protection to apply
• When a Protection Policy is assigned to a resource:
– The Snapshot Rule is automatically applied
– Replication is automatically configured

Protection Policies
Overview - Example
• Snapshot Rules:
– Hourly snapshots
– Daily snapshots Protection
– Weekly snapshots Policy
• Replication Rules:
– Asynchronous Replication RPO
▪ 1 hour RPO

Protection Policies
Use Cases
• Create one or more Protection Policies and use them across multiple resources
– I.E. Create a Gold, Silver, and Bronze service levels and assign them as needed
Example:
• Create a Protection Policy Containing:
– 3 Snapshot Rules:
▪ Hourly snapshots
▪ Daily snapshot at midnight
▪ Weekly snapshot taken on Sundays
– 1 Replication Rule:
▪ Remote Replication with 1 hour Recovery Point Objective (RPO)

Snapshots

Snapshots
Overview
• Snapshots are crash-consistent point-in-time copies of data that are stored in

supported storage resources
– Snapshots are the local data protection solution within PowerStore
– Note: A crash-consistent snapshot is not application consistent. Application consistency
requires I/O to be quiesced at the host, host buffers flushed, and a snapshot taken.
• Snapshots are not full copies of the original data

– Pointer based, redirect on write technology
– You should not rely on snapshots for mirrors, disaster recovery, etc.
• Snapshots consume overall system storage capacity as changes to the source are
made
– Ensure that the system has enough capacity to accommodate snapshots

Snapshots
• Snapshots are supported on volumes, volume groups, virtual machines, file

systems, and thin clones
• Snapshots can be created:
– Manually
– Automatically via a user created Snapshot Rule within the system
▪ To have snapshots automatically created via a predefined schedule on a storage resource, a
Protection Policy containing a Snapshot Rule must be added to the storage resource
– Automatically by a feature within the system
▪ Such as replication
▪ Hidden from the user

Snapshots
Redirect on Write Technology
D’
B’
A’
B
C
D
A
Application 1
A B C D B’ A’ D’
Monday Tuesday
Volume/Volume Group Snapshots
Overview
• Volume and Volume Group snapshots cannot be attached/accessed by hosts

– To access the data within a snapshot, create a Thin Clone
• Volume snapshot names are case insensitive and are limited to 128 characters
– Example of a valid snapshot name: SnApShOt`~!@#$%^&*()-_=+[]\;',./{}|:"<>?
– Names for existing snapshots can be edited
– Snapshots names need to be unique within the Volume family
• Taking a snapshot of a Volume Group creates a snapshot on all members of the

Volume Group at particular moment in time
– Snapshot consistency will depend on the “write order consistency” setting

Volume/Volume Group
Storage Resource Family
Storage Resource
Snap 1 Snap 1 Thin Clone
S1TC Snap 1
Snap 2 Snap 2 Thin Clone
S1TC Snap 2
S2TC Snap 1
Snap 3
S2TC Snap 2

File Snapshots
Overview
• There are 2 different file snapshot types, both used for data protection:
– Protocol (Read-Only)
▪ Read-Only snapshot that can be exported as an SMB share and/or NFS export
• Access is provided by the parent NAS Server
▪ This is the default type created by a Snapshot Rule
– .Snapshot (Read-Only)
▪ Read-Only snapshot that can be accessed through Previous Versions or .snapshot
• Note: You may customize the name when taking a manual snapshot, but editing
the name or the access type of a file system snapshot is not possible

Snapshots
Snapshot Aging
• When manually creating a snapshot, the user can customize when the snapshot is
retained until
– To ensure a snapshot doesn’t expire, choose “No Automatic Deletion”
– Can change existing snapshots to “No Automatic Deletion” if needed
• Snapshots created by a schedule are automatically assigned an expiration

Date/Time
– Can change existing snapshots to “No Automatic Deletion” if needed
• The snapshot aging service runs hourly in the background and cleans up expired
snapshots as needed
– Snapshots are deleted in batches to improve performance and efficiency

Volume/Volume Group Snapshots
Data Recovery
• Need access when some amount of data has been deleted or corrupted?
1.Find a snapshot which contains the data needed
2.Create a Thin Clone of the snapshot
3.Provide host access to the Thin Clone to access the data
• All data within the source has been deleted or corrupted?

2.Utilize the Restore function to return the source volume to the desired point in time

File Snapshots
Data Recovery
• Need access when some amount of data has been deleted or corrupted?
2.Depending on the snapshot type:
a. Protocol (Read-Only) – Create an SMB share and/or NFS export
b. .Snapshot – Access the snapshot data via Previous Versions or .snapshot
• All source data has been deleted or corrupted?

2.Utilize the Restore function to return the source file system to the desired point in time

Snapshot Rules
Overview
• Snapshot Rules are used to automatically take snapshots on a supported storage

resource
– Defines the Frequency to take a snapshot, and the Retention period for snapshots of a
given storage resource
• To have the system automatically take snapshots on a resource:

1.Create a Snapshot Rule
2.Add the Snapshot Rule to a new/existing Protection Policy
3.Add the Protection Policy to the storage resource

Snapshot Rules
• When creating a snapshot rule, the schedule is entered in your local time
– Stored within the system in UTC time format
• Snapshots created by a Protection Policy will have the following naming scheme:
– Block snapshots:
▪ Snapshot Rule Name.Resource Name.Timestamp with nano-time
• Example: Hourly Snapshots.Volume 1.2019-07-18T17:00:00Z 702319493
– File snapshots:
▪ Snapshot Rule Name_Resource Name_Timestamp with nano-time
• Spaces are automatically replaced with an underscore
• Example: Hourly_Snapshots_FS1_2019-10-30T07:00:01Z_614817807

Snapshot Rules
Modifying and Deleting Snapshot Rules
• A Snapshot Rule can be edited at any point in time, even while in use
– Storage resources will automatically inherit the new settings once the rule is updated
within the Protection Policy
– If the Retention is changed, all existing snapshots created by the rule will be updated
• A Snapshot Rule cannot be deleted while assigned to a Protection Policy

• When deleting a Snapshot Rule, the user is provided the option to delete all
snapshots created by that rule
– Checkbox (Disabled by default)

Restore Operation
Overview
• The restore operation is used to replace the contents of a parent storage resource
with data from an associated snapshot
– Returns the parent resource to a previous point in time of itself
– Only pointer updates occur (near instantaneous)
– Snapshot/Thin Clone hierarchy is preserved
• When completing a restore operation, it is highly suggested to:

– Take a backup snapshot of the resource before proceeding
▪ Within the GUI, a checkbox is provided to take a backup snapshot before proceeding with the
restore operation (Enabled by default)
– To prevent data corruption during the restore, it is recommended that you first quiesce
any applications running on the production host and flush the host cache
• For virtual volumes, utilize vSphere to revert to a previous snapshot

Restore Operation
Example
Goal:
Restore to Monday Snapshot
Application 1
A B C D B’ A’ D’
Monday Tuesday Backup

Volume/Volume Group Refresh Operation
Overview
• The refresh operation is used to replace the contents of a resource with data from
a resource within the same family
• When completing a refresh operation, it is highly suggested to:

refresh operation (Enabled by default)
– To prevent data corruption during the refresh, it is recommended that you first quiesce
• Snapshots of Volumes, Volume Groups, and Thin Clones cannot be refreshed

Volume Thin Clone Refresh
Example
Goal:
Refresh: Thin Clone
From: Parent Resource Host 1
Protection
Policy
A B C D B’ A’ D’ C’ E
RPO
Host 2
Monday Backup Thin Clone
Volumes/Thin Clones - Supported Operations
Object to Refresh Refresh From Is Allowed?

Volume Thin Clone Yes
Volume Thin Clone Snapshot Yes
Volume Volume Snapshot No*
Thin Clone Parent Volume Yes
Thin Clone Volume Snapshot Yes
Thin Clone Thin Clone Snapshot Yes**
*Utilize the Restore operation to complete this.

**Refreshing from a snapshot of a peer, in-family, Thin Clone is supported. If the snapshot is of the
Thin Clone being refreshed, utilize the Restore operation.

Volume Groups/Thin Clones - Supported Operations
Object to Refresh Refresh From Is Allowed?

Volume Group Thin Clone Yes
Volume Group Thin Clone Snapshot Yes
Volume Group Volume Snapshot No*
VG Thin Clone Parent Volume Yes
VG Thin Clone Volume Snapshot Yes
VG Thin Clone Thin Clone Snapshot Yes**
*Utilize the Restore operation to complete this.

**Refreshing from a snapshot of a peer, in-family, Thin Clone is supported. If the snapshot is of the
Thin Clone being refreshed, utilize the Restore operation.

File System Refresh Operation
Overview
• A file system or file system thin clone refresh operation replaces the contents of a
snapshot with the current view of the parent object
• When completing a refresh operation, it is highly suggested to:

refresh operation (Enabled by default)
– To prevent data corruption during the refresh, it is recommended that you first quiesce

File System Refresh Operation
Overview – File Systems
• When refreshing a file system snapshot

– The creation time of the snapshot is preserved, and the Last Refresh Time value is
updated
– Host accessing the restored snapshot will not know that it was refreshed (The original
creation date/time preserved)
• File system Thin Clones behave the same way

File System Snapshot Refresh
Example
Goal:
Refresh: Snapshot 1
From: Parent Resource Client 1
A B C D B’ A’ D’ C’ E
Snapshot 1 - Protocol (Read-Only)

Thin Clones

Thin Clones
Overview
• A Thin Clone is a read-write copy of a resource that shares blocks with the parent
resource
• Thin Clones are not full copies of the original data
– Pointer based, redirect on write technology
– Creating a Thin Clone is near instantaneous
• Protection Policies can be assigned to Thin Clones

– Allows the Thin Clone to be protected during usage
– The Protection Policy can be different than the one assigned to the source

Thin Clones
• A Thin Clone of a Volume, Volume Group (VG) or File System can be created
using the latest image of the Volume, File System, Snapshot, or another Thin
Clone
– The Thin Clone is automatically assigned to the same NAS Server
• When created, the Thin Clone is displayed in the same page as the parent
resource
• Parent resources can be deleted without deleting the associated Thin Clone
– No blocks used by the Thin Clone are removed
– The Thin Clone becomes its own resource
– Only unique blocks within the parent resource and snapshots will be reclaimed

Thin Clones
Use Cases
• Test and development environments

– Ability to create multiple copies of the same data for multiple teams
– When used with a Protection Policy, ability to replicate* and create snapshots
▪ *If replication is supported
• Parallel processing
– Ability to create and provide multiple copies of the same data to multiple servers/clients
for parallel processing
• System deployment
– Quickly utilize a common image to easily deploy new environments

Block Remote Replication

Overview
Use Cases - Remote Data Protection
• Redundancy
– Increase data redundancy and fault tolerance levels
– Failover to a secondary site
• Compliance
– Require additional footprint.
– for instance insurance companies, bank, and government sites
• Migration
– Migrate data between storage systems
– Tech Refresh

Replication Concepts
Supported Replication with PowerStore T / PowerStore X
• Supports PowerStore T model Arrays and PowerStore X model Arrays – also mixed
• RPO based Asynchronous Remote Replication

The acceptable amount of data, measured in units of time, which may be lost due to a failure
– With RPO the also frequency of synchronization between Source and Destination is defined
– Factor of time, not amount of changed data
– PowerStore supports fixed RPO values between 5min and 24hr
• Replication is supported for following Storage Resources

– Volumes
– Volume Groups
– Thin Clones
• IP based only
• Policy based
• Leverage PowerStore Snapshot technology

Replication Topologies
Overview
Supported Replication Topologies with PowerStore platform

• One-to-One (1:1)
– Uni Directional
▪ One or multiple Volumes are replicated from one cluster to one other cluster
– Bi-Directional
▪ Various Volumes are replicated in different directions between two clusters
• One-to-many (1:n)
– Different Volumes are replicating to multiple destination clusters
– No Support for fan-out or cascaded/chain replication
• Many-to-One (n:1)
– Individual Volumes from different cluster are replicating to a single cluster

Remote Replication
Diagram
Storage Network
WAN
Node Node Remote System Node Node

A B A B
Appliance 1 • Clustername Appliance 1
• Connection for
Volume Volume
Group - Management
Replication Sessions Group
- Data
Manage Replication
V1 Volume Pair 1 V1
•• SSL Certificates
Volume Pair
V2 •• Latency
Volume&Pair
Policy Rule2 V2
• RPO
Volume Pair 3
V3
• Systems V3
Cluster 1 Cluster 2
WAN
Management Network

Replication Data Connection
Network Port
• Replication Traffic is routed via Storage Network and their corresponding Network Interfaces
• One individual Port or system bond mapped for Storage Network can be tagged for replication
traffic. Default is bond0 / first two ports on Mezz Card and share host-, and replication workload
• Source and Destination System may use different configuration, but symmetric port tagging on
Nodes within same Appliance
• As replication Port can be tagged

on IP Ports located on
– Mezz Card Ethernet Ports
– Ethernet I/O Modules

Network Port
• Replication Traffic is routed via Storage Network and their corresponding Network Interfaces
• One individual Port or system bond mapped for Storage Network can be tagged for replication
traffic. Default is bond0 / first two ports on Mezz Card and share host-, and replication workload
• Source and Destination System may use different configuration, but symmetric port tagging on
Nodes within same Appliance
• As replication Port can be tagged

on IP Ports located on
– Mezz Card Ethernet Ports
– Ethernet I/O Modules

Physical Port
Mapping of Storage Network and replication Tag must be the same on Node A and Node B
IoModule1 IoModule0 4-Port Mezz
B
A
4-Port Mezz IoModule0 IoModule1
Default bond tagged for replication

Remote System
Configuration Summary
• Can be initiated on Source or Destination System

• Pairs two systems in order to establish replication session
– “Remote System” is always a 1:1 relationship
– Up to 16 “Remote System”
• Sets Bi-Directional Management Connection over SSL via Cluster-IP’s

- No further password maintenance after connection is successful established
- Certificate based authentication Source and Destination System
- SSL-Certificate encryption
Automatic Exchange on UI, but separate step on PSTCLI and REST-API
- Sets up iSCSI configuration in both direction between Clusters

- Data Connection based on Latency setting for different iSCSI Portals
– Standard iSCSI Portal on Port 3260 for Low Network Latency (< 10ms)
– Use iSCSI Portal on Port 3261 for High Network Latency (>= 10 ms)

Replication Session
Configuration Summary
• Requires Remote System

• Configuration on source
– Protection Policy
– Replication Rule
▪ Destination System
▪ Recovery Point Objective (5min - 24h)
▪ Alert threshold
• Manage and Maintain Replication

– Creates identical Volume on destination with same attributes but Read-Only
– Starts initial sync
– Sets Scheduler Events and drives replication based on configured RPO
– Executes Replication Session operations like Failover, Pause, and Resume

Replication Session
Monitoring
• There are various ways to monitor the Replication Session

– PowerStore Management UI
▪ Protection -> Replication
▪ Storage -> Volumes (Group) -> Protection -> Replication
– PowerStore CLI
▪ replication_session show
– REST API
▪ GET /api/rest/replication_session?select=*
# pstcli -d 10.1.2.34 -u admin -ssl accept replication_session show –format nvp

1: id = rep-b8a94bc9-a17e-4abb-9118-ff4742b49245
last_sync_timestamp = 08/08/2019 09:05:21 AM
local_resource_id = 2c79110d-a89e-4041-82d4-ccd408c84c1b
remote_resource_id = 4d237db1-073a-45fd-99a5-a952f28477da
progress_percentage =
2: id = rep-ff83db99-025d-4f78-b33e-8a867df98ac7
last_sync_timestamp = 08/08/2019 09:05:21 AM
local_resource_id = ae2a5053-cdb6-4fe7-9a48-8276b0610f37
remote_resource_id = 854d9d80-5c9d-4827-b15f-6ff2f764a16b
progress_percentage =

Replication Session
Performance Metrics for Replication
Level GUI REST API CLI
Volume ✓ ✓ ✓
Volume Group ✓ ✓ ✓
Remote System ✓ ✓
Appliance ✓ ✓
Cluster ✓ ✓

Replication Session
Status
Replication Session is set, but idle
Synchronization in progress – either manually triggered or based on RPO
Replication is paused – either manually or after Failover
System itself triggered “Pause” – may caused by remote Cluster down
Replication is resuming
Switching Source and Destination
Replication failed over, but paused

Replication Session
Replication Control
Status Synchronize Failover Reprotect Pause Resume
Operating normally X X X
Paused X
on Source
Failing Over
Failed Over X(1) X(1)
Synchronizing X X
Operating normally X X
on Destination
Paused X X
Failing Over X
Failed Over X(2) X(1)
Synchronizing X X
(1) New Source/Destination after Failover
(2) Forced Failover
Replication Session
DR aspects
• When Session is DR Failing-Over

– Replication Session can be Deleted
– All other operations are restricted
• When in Failed-over state

– Volume could be resized on DR Site volume
– Volume Group membership could be changed on DR site Volume Group
– Possible to Reprotect
– Possible to Delete session
– Issue Snapshot

Replication Session
DR aspects - Reprotect
• Erases any outstanding

changes remaining on
Original Source that were
never synced due to DR
Failover
• Activates Replication session

after Planned or DR Failover
• Starts processing RPO

scheduler and starts to
synchronize

Replication Session
Conclusion - Protection Policy
• A Protection Policy can be assigned only on Source

– During creation of Volumes or Volume Groups (VG), or
– On existing Volumes or VG
• Removal of a Protection Policy is only possible on Source

– Independent from current State
• Size changes on Source are replicated to Destination

other attributes can/must be modified on Source and Destination individually
• A Volume remains on Destination when Protection Policy with Replication Rule
is removed on a Volume / VG

Replication Session
Conclusion - Protection Policy
• Volume Groups (VG)

– Only Volumes without Protection Policy can be added into a VG
– Protection Policy need to be assigned to whole VG
– New Volumes in a VG inherit Replication Policy
• Snapshots
– Snapshots on Source are replicated in chronological order during RPO sync

Limits for Replication
Max Remote Clusters 8

Max Protection Policies 32
Max Replication Rules 32 (max 1 per Policy)
Max Replication sessions per Appliance 125
Max Volume replications per Appliance
- At RPO of 15 minutes+ 500
- At RPO of <15 minutes 75
Min RPO 5 min
Min RPO Best Practice 15 min
Max RPO 24 hours
Max Snapshots per Volume (incl. RPO Snaps) 256
Note: As updates will periodically occur, please reference the support matrix posted on E-Lab for the most up to date limits information.

PowerStore metro node

Introducing PowerStore metro node
Hardware add-on feature for PowerStore (block storage)
• Enables automated business continuity with zero Recovery Point Objective (RPO)
and Recovery Time Objective (RTO)
– Zero RPO means there’s an expectation for zero data loss
– Zero RTO means that there is no loss of data access
• Provides true active-active synchronous replication over metro distances

• Provides the ability to automatically initiate an instant site failover through its
unique virtual machine (VM) witness technology (expected FY 2022 Q2)
• Also supports local configurations for
– Continuous application availability
– Data mobility to non-disruptively relocate workloads
– Provides storage technology refresh without application downtime

Key features and functionality
• Low cost and designed for six 9’s availability
• Configurations
– Metro node metro - two clusters supported within 5ms distance
▪ Up to 10 ms depending on application
– Metro node local allows mirroring for two arrays within a data center
▪ Or mirroring across appliances within PowerStore multi-appliance cluster
• 32 Gb/s Fibre Channel support (16Gb/s supported now, 32Gb/s expected FY 2022 Q2)
• 64TB virtual volume support
• Embedded management server
• Support for Ansible 1.1 automation
• Events alerts and support
– iDrac based HW monitoring
– Dial home support
• FRU support
Hardware components
Metro node Hardware Configuration
• Based on proven Dell EMC hardware technology

– Intel Xeon Silver 4208 CPUs (8 cores per CPU), Dual CPU
– 64GB RDIMM Memory, 2 x 16GB DIMMs per CPU
– 480GB SATA Solid State Drive, read intensive, < 1 DWPD
▪ Using UEFI Boot
– 8x 2.5” Drive Chassis (no PERC/RAID support)
– Basic PE Bezel w/no LCD with metro node brand insert
• IO Adapter configuration targeting

– Three PCIe Low Profile IO Cards
– One NDC (Network Daughter Card)
• BMC (Baseboard Management Controller)

– iDRAC 9 Enterprise / racadm cli interface

Hardware components
HBA and NIC Cards
• Fibre Channel HBA Adapter (PCIe slots 1 & 2 – CPU1)

– Emulex LPe35002 32Gb Dual Port FC, LP HBA
• NIC Cards
– Intel x710 10GbE Quad Port Base-T (PCIe slot 3 – CPU2)
– Intel x710 10GbE Quad Port SFP+ (rNDC slot – CPU1; for LOM connections)

Port layout and labeling
PortMap – Port Names
• Fibre Channel Ports (32Gb)

– FC1, FC2 (PCIe Slot 1)
– FC3, FC4 (PCIe Slot 2)
• 10GbE RJ45 Ports (PCIe Slot 3)

– Management Ports - MGMT1, MGMT2
– Service Ports – SVC
– Customer Ports – CUST
• 10GbE SFP+ Ports (rNDC bottom right)

– Local COM Ports – LCOM1, LCOM2
– WAN Ports – WAN1, WAN2
• Power Connections – Black and Gray (far right)
• iDRAC Connection – Rear Port – DO NOT USE (We cross connect to peer iDRAC thru LCOM on RMII interface)

Cable connectivity
CMA cable kit
• Detailed Contents:
– Two 1U CMA’s (Cable Management Arms)
▪ CMA allows server to slide forward for service without disconnection
of any cables.
– Pre-dress cables:
▪ MGMT1 & MGMT2 Cat6 Shielded Ethernet Cables (Lime and Violet,
end labels)
▪ LCOM1 & LCOM2 SFP+ DAC (direct attach passive copper) cables
(black, end labels).
▪ Two Black Power cords (no labels)
▪ Two Gray Power cords (no labels)
– One Red Cat6 Shielded Eth Cabled (Red, no labels, fastened Assembled CM Arms
to the loop)
– PortMap Tag (gets fastened to the cable loop)
– Expectations are for customers to run Fibre Channel and
Ethernet cables through here as well
• Kit is installed during system setup and commissioning

Virtualization Integration

vSphere Integration Overview
• PowerStore features deep vSphere integration which enables:

– Discovering VMs
– Managing VM snapshots
– Automatic registration of the VASA Provider (PowerStore X models only)
– Automatic mounting of Storage Containers (PowerStore X models only)
• Benefits:
– Perform monitoring and active management from PowerStore Manager
– Monitor events without requiring continuous polling
– Automation for configuration process
• vCenter must be registered in PowerStore Manager to enable this functionality

– Each cluster can only be registered to a single vCenter instance at a time

vCenter Registration
PowerStore X Models PowerStore T Models
vCenter is automatically registered as part of

vCenter is not registered by default
initial configuration
vCenter cannot be unregistered. vCenter

vCenter can be unregistered if vSphere
IP/credentials can be updated, but must point to
integration is no longer needed
the same vCenter
VASA provider is not automatically registered in

VASA provider is automatically registered in
vSphere. VASA provider should be manually
vSphere and vVol support is enabled, by default
registered to enable vVol support

PowerStore X Models in vSphere

Seamless Integration
• Native vSphere features can be leveraged between PowerStore and external ESXi hosts
– Easily integrate PowerStore in to your vSphere environment
• vMotion - Move VM compute on to PowerStore X model nodes

• Storage vMotion – Move VM storage on to a PowerStore Storage Container
– Can be on to a PowerStore T or PowerStore X model appliance

Storage Containers
• PowerStore arrays automatically provision a default Storage Container

– Name: PowerStore <Cluster_Name>
• Once the VASA Provider is registered, this Storage Container becomes accessible
– Exposes all of the storage available on the cluster as a vVol datastore
– Enables external hosts to use it for VM storage
• Hosts must be registered using iSCSI or Fibre Channel

– All registered hosts are automatically provided access to this Storage Container
• Additional Storage Containers can also be created

• A quota can be enabled on Storage Containers
– Enables under or over-provisioning capacity to vSphere
– High Water Mark determines when an alert will be generated for the administrator

VM Details
• PowerStore discovers VM details and displays them in PowerStore

Manager
• Subscribes to VM object change events
– No need for continuous polling
• Available for all vVol-based VMs provisioned on the PowerStore Storage
Containers
– Regardless of internal or external compute

VM Snapshot Details
• VM Snapshot information is also displayed and can be managed from PowerStore

Manager
– Available for vVol-based VMs that are provisioned on PowerStore’s Storage Containers
– Manual snapshots can be taken and deleted
– Protection policies can be assigned for automatic snapshot creation/deletion
• VMware has a limit of 31 snapshots per VM

– If the limit is reached, the oldest scheduled snapshot is deleted
– Manually created snapshots are never deleted automatically
• PowerStore sends a maximum of 20 simultaneous snapshot operations to vSphere

– Additional operations are queued in order to avoid overloading the system

VAAI Support
• Block
– Atomic Test & Set (ATS) with Compare and Write (CAW) – Manages locking of files on
shared volumes
▪ Enables multiple ESXi hosts to write to the same volume without corruption
– Block Zero – Hardware-assisted zeroing using Write Same
▪ Speeds up VM provisioning
– Full Copy – Hardware-assisted copying using XCOPY
▪ Saves network traffic by offloading the copy operation to the storage
– Thin Provisioning and Unmap – Free capacity reporting and enables space reclaim
▪ Maximizes efficiency by reclaiming unused space
• File VAAI primitives are currently not supported

VMware Plugins
• Virtual Storage Integrator (VSI) - Enables provisioning, managing, and monitoring

storage from vSphere
• vRealize Operations Manager (vROps) Plugin – Monitor performance, capacity,
and analytics of the PowerStore system along with your virtualized environment
• vRealize Orchestrator (vRO) Plugin – Streamline VMware and PowerStore tasks
through automation workflows
• Storage Replication Adapter (SRA) – Leverage Site Recovery Manager for
VMware disaster recovery

Serviceability

PowerStore Licensing
Automatic Licensing PowerStore Cluster
Appliance _1 Appliance _2
• In PowerStore, the license will be installed
automatically during the Initial Configuration License License
Wizard File File
– All inclusive license e.g. replication, snapshot, Point of
migration…etc. Contact
Primary
• License will be issued to each Appliance in Appliance
HTTPS
the cluster
– Management will be centralized for the customer to run
from the primary Appliance
– The primary Appliance will be the point of contact for
all the Appliances
• License will be obtained by sending request

to Dell EMC via HTTPS
– Port 443 needs to be open

SupportAssist
Overview
• Customers can use SupportAssist to provide remote support from Dell EMC to the
PowerStore cluster in case of issues
– Gathering Data Collects
– Troubleshooting
• SupportAssist is a rebranded from ESRS with a new simple Connect Home solution
– Same backend of the ESRS channel
– New and simple frontend
– Does NOT require any support account or user’s information
• Each Appliance will be connected directly to Dell EMC
– Can enabled from the ICW or from the Settings page
– Centralized management
• SupportAssist types:
– Direct Connect
– Gateway Connect
• It should also be noted that CloudIQ supports PowerStore
Support Materials
Overview
• Support Materials also known as Data Collects

• To help troubleshoot PowerStore clusters and find what is the issue
– System degraded state
– Performance problem
– DU/DL
• Support Materials include:

– System logs
– Configuration details
– Diagnostic information
• Collects from both Nodes (A and B)

• Used mainly by Dell EMC
Disabled Support Notification
Overview
• An administrator can pause the alert notifications that sent to Dell EMC Support
during a maintenance time
– E.g. unplugging cables, swapping drives ..etc.
• Disable Support Notifications can disable the call home to one or multiple
Appliance(s) in the cluster
– Enabled/disable at anytime
• The Maintenance Window Duration counted by hourly

– 1 hour min
– 48 hours max
• The Maintenance Window Duration can be changed at anytime during the suspend

SSH Management
Overview
• SSH – Secure shell a cryptographic network protocol

• A user can enable/disable SSH in the Appliances level at any point of time
– Login to the primary node
• SSH is enabled by default before the Appliance gets initialize

– Once the Appliance is configured SSH is disabled
• Benefits:
– Execute service commands in the Appliances
– Run scripts and user CLI commands

Service Mode
Overview – What’s Service Mode?
• PowerStore’s node may encounter hardware or software failures that trigger the
Node unable to boot
– E.g. Take SLIC out
• Node would be placed into the Service Mode state if failures occur
– E.g. Boot loop, or D@RE lockbox copies are corrupted
• While in Service Mode the Appliances does NOT serve any customer I/O on that
Node
– It fail over to the other Node

Monitoring
Overview
• The cluster provides history of events

to the user in case of issues accrue
Event
• In order to keep the cluster in a healthy
state, and running well, proper alerts
have to be visible to the user
– Diagnose and to be able to fix the problem
Alerts
• Send alerts and be able to notify
support
• Monitoring in PowerStore consists of:
Notify (SupportAssist)
– Events
– Alerts
– Jobs

SMTP
Overview
• SMTP (Simple Mail Transfer Protocol) is a TCP/IP protocols used in sending and
receiving e-mail
• The SMTP server is used to send alerts notifications from the PowerStore cluster
to the user’s email address
• It can be enabled and disabled at anytime
• In order to enable SMTP:
– Server address
– From Email Address
– Port
• Send Test Email is available to check if the SMTP Server is set up correctly
– Send a test email by entering an email address

Reboot and Power Down
Overview
• An administrator has the ability to Reboot and Power down a Single Node in
PowerStore Appliance
– Done through the GUI, REST API or service script
– Customers can perform maintenance on their Nodes
▪ Replaced a failed hardware, such as, SLIC or power supply… etc.
• When Reboot or Power Down the Primary Node, the docker containers failover to
the secondary Node
– Primary will be switch over to the other Node
• In order to bring a Node back after a Power down, accessing a service script is
required
– Running the service script svc_node power_up will power up the nodes

Non Disruptive Upgrade
• Non-Disruptive Upgrade
• Customer Driven
• Software, Hotfixes and Disk Firmware
• One Node at a Time for Availability

Security Overview

Overview
Security Topics
• For PowerStore, the following high-level security practices have been

implemented:
– Physical
▪ Data at Rest Encryption (D@RE)
– Software
▪ User Authentication
▪ User Role Authorization
▪ Certificate Management
▪ Auditing Log

Data at Rest Encryption (D@RE)
Overview
• PowerStore supports Data at Rest Encryption (D@RE) which encrypts all user
data that is written to a storage drive at 256-bit Advanced Encryption Standard
(AES)
– PowerStore utilizes Self Encrypting Drives (SEDs) to implement D@RE
• Support array-based self managed keys with the option of downloading user
backup of keys
• Benefits:
– Compliance of National Institute of Standards and Technology (NIST) Special Publication 800-111 and
others
– Protects user data in the event of physical security breaches

Active Directory / LDAP authentication
Feature description
• Independent from File / NAS LDAP configuration

• Centralized authentication against
– Microsoft Active Directory (AD) or
– OpenLDAP / Native LDAP
• PowerStore role mapping based on

– Single user
– AD/LDAP group
• One or multiple servers are supported

• SSL encryption
SSL requires Certification authority (CA) trust certificate for AD/LDAP server certificate

Active Directory / LDAP authenticaiton
AD and LDAP – not the same, but similar
• LDAP specifies the application protocol, used by many different Directory Services
– Microsoft Active Directory (WindowsTM oriented)
– OpenLDAP
• Object database „““““hello

– Microsoft Active Directory
– Berkeley DB is an example of a backend used by OpenLDAP
• Major differences (out of the box)

– The schema which defines allowed attributes for objects is different – for example
▪ User ID attribute : “uid“ (LDAP) vs. “samAccountName“ (AD)
▪ Object class for group memebership : “groupOfNames“ (LDAP) vs. “groups“ (AD)
– Advanced usage of AD relies on Kerberos, while OpenLDAP prefer SASL (Simple Authentication
and Security Layer)

LDAP Introduction
• Lightweight Directory Access Protocol

• Standardized application protocol based on X.500
• Hierarchical / Tree Database with roots, branches and leaves
User A
• Root could represent the organizational domain
• One or multiple branches can
– Represent an organizational structure
– Organize objects by regions
– Reflect an own hierarchical structure
• Leaves are objects like User, Group, and Computer

• Each object has unique Distinguished Name
Example: dn: cn=User A, ou=People, ou=EMEA, dc=dell, dc=com

Demo // Settings > Users > Directory Services
Active Directory OpenLDAP
lab5
cn=readonly,dc=lab5 •••••••••

LDAP User/Group to PowerStore Role mapping // Settings > Users > Users > LDAP
A group configuration allows

all group members and members from
a nested group to log in with the associated acount role

AD/LDAP Group to PowerStore role mapping / RBAC
• LDAP groups or individual LDAP users can be mapped to a Role on PowerStore

Import External Storage

Overview
• Known in the GUI as Import External Storage

• Import External Storage is a native capability in PowerStore to import storage
resources from Dell EMC Midrange systems
– Block only resources, including:
▪ LUNs/Volumes
• VMFS Datastores (for PS Series only)
• Windows RDM (for PS Series only)
▪ Consistency Groups
• PowerStore also has support for Agentless Import

– The disruptive variant for native import which involves application downtime
– For users who are not inclined to have a host plugin footprint on host(s)
– Supports older firmware/OE versions of source arrays

Overview
Client
• Where are we starting?

– Client attached to an old array,
which hosts the data
• Where do we want to be?

– Wants to move application data to
a new PowerStore system
• Solution
– PowerStore allows for native
migration options of this client’s
data from the old array to the new
PowerStore Source
PowerStore
system

Overview
Use cases
• Data migration
– Move your data to a future proven
infrastructure, PowerStore
• Application migration
– Move your existing applications to
PowerStore

Overview for Non-disruptive Import
Main components
Client
• Source system Host plugin

– Dell EMC’s Midrange arrays
– Upgrade may be applicable
• Client(s) with a host plugin

installed
• PowerStore appliance
– Orchestrator
▪ Native software that orchestrates
the import
Orchestrator
Source PowerStore
system
Architecture for Non-disruptive Import
Connectivity
Client
• Front-end: Between client, Host plugin

source system, and
PowerStore
– iSCSI
iSCSI or FC iSCSI or FC
– Fibre Channel (FC): Zoning
required
– Protocols must match between
source and destination iSCSI only
• Back-end: Between source
array and PowerStore Orchestrator
– Only iSCSI support
Source PowerStore
system
Overview for Non-disruptive Import
3
2 Cutover
Import • Cutover is allowed once
1 • Add source system
Setup the import session is in
• Select source storage a Ready to Cutover
• Zoning for the front-end resources to import
connectivity state (source and
• Click Begin Import PowerStore are in-
• Add iSCSI connectivity
between Source system • Import session is sync)
and PowerStore system created
• Auto-cutover is
• Install Host plugin • Path flip to PowerStore
available
• Background copy from
source system starts • Once cutover, there
is no rollback

Overview for Agentless Import
Main components
Client
• Source system
– Dell EMC’s Midrange arrays
– Upgrade may be applicable
• PowerStore cluster
– Orchestrator
▪ Native software that orchestrates the
import
Orchestrator
Source PowerStore
system
Architecture for Agentless Import
Connectivity
Client
• Front-end: Between client,

source system, and
PowerStore
– iSCSI
iSCSI or FC iSCSI or FC
– Fibre Channel (FC): Zoning
required
– Protocols must match between
source and destination iSCSI only
• Back-end: Between source
array and PowerStore Orchestrator
– Only iSCSI support
Source PowerStore
system
Action by user
Overview for Agentless Import New for Agentless
1 2 3
Setup Import Cutover
• Zoning for the front-end • Add source system • Cutover is allowed
connectivity • Select source storage once the import session
• Add iSCSI connectivity resources to import is in a Ready to Cutover
between Source system • Map host(s) state (source and
and PowerStore system • Click Begin Import PowerStore are in-
• Add host(s) to • Import session is sync)
PowerStore created • Auto-cutover is
• Path flip to PowerStore available
• Enable destination
• Once cut over,
volume
there is no rollback
• Start copy
• Background copy from
source system starts

Overview
Support for Agentless Import
• Supported Source systems and version:

– VNX2 Version: 05.33.008.5.x or later
– Dell EMC Unity Version: OE 4.4 or later
– PS Series (EqualLogic) Version: 9.1.9 or later
– SC Series (Compellent) Version: 07.03 or later
– XtremIO X1 Version: 4.0.25-x or later
– XtremIO X2 Version: 6.2 or later
**For the most up-to-date supported versions, refer to the PowerStore ESSM
Support Matrix**

Overview New PowerStoreOS 2.0
Support for Non-disruptive Import
Multipath VNX2 PS Series Dell EMC Unity SC Series

Operating System OS Version
Module iSCSI FC-FE FC-BE iSCSI iSCSI FC-FE FC-BE iSCSI FC-FE FC-BE
DC Array Plugin
VMware vSphere 6.7
EQL MPIO NA NA NA
DC Array Plugin
Linux SLES 15
EQL MPIO NA NA NA

Overview
Source system required parameters
• Name
– Only for XtremIO
• Management IP Address
• iSCSI IP Address
– Multiple comma separated iSCSI IP
addresses can be entered
• Username
• Password
• CHAP Setting (Optional)

Dell Technologies Services
for PowerStore

Dell Technologies Services
Ultimate choice & flexibility throughout the life of your PowerStore appliance
Dell EMC Dell EMC

Dell EMC Dell EMC Dell EMC
ProDeploy ProSupport
Migration Residency Optimize
Enterprise Enterprise
Services Services for Storage
Suite Suite
Move data Speed adoption Get ahead of

Plan, install & Optimize
efficiently with with hands-on problems
configure with performance with
minimal PowerStore before
confidence ongoing guidance
disruption expertise they happen
With data-driven technologies to connect and manage your service experience

Consulting Services | Education Services | Managed Services
27
Internal Use - Confidential ©
273 3of 292 © Copyright
Copyright 2020
2021 Dell
Dell Inc.
Inc.
What sets us apart
Consistent end-to-end services across the
street or across the globe
• Onsite deployment experts for complex
environments or when short on time or staff
• Technical expertise in clustering, virtualization and

networking
• Services sized right for your needs, from remote

guidance and evaluations to hands-on
operational residents
• ProDeploy Plus training credits you can use for

PowerStore courses or certification
• Anytime Upgrade program, providing flexible

upgrade options to future proof your PowerStore
investment
• ProSupport Plus adds proactive, predictive support

and an assigned account-level expert
• Optimize for Storage brings in a technical storage

expert to optimize performance & efficiencies year
round
Internal Use - Confidential
Outside experts help drive fast, accurate results
Forrester survey of IT managers
Use IT providers to
70% deploy new technologies
more quickly*
Say services from IT providers

72% help organizations manage
spikes in demand for IT
expertise
Let certified services experts handle
the details so you don’t have to. Think deployment services
80% allow IT staff to spend more
time on innovation
2019 Forrester Thought Leadership Paper commissioned by Dell EMC

Faster deployment with less IT effort
Dell Technologies certified engineers
with PowerStore expertise
ProDeploy Plus
✓ Experts in network SAN/NAS requirements, • 24x7 onsite hardware
planning and validation ProDeploy installation and system
software configuration
✓ Clustering management planning
• 24x7 onsite hardware • In-region project
✓ VMware software expertise, trained and installation management
certified on vSphere Enterprise and vCenter • 24x7 remote system • 30 days of post-
software configuration deployment assistance
• Implementation • Training credits
4x Faster planning
and deployment1
Planning
with ProDeploy for Enterprise Additional Deployment Time
1Based on a September 2020 Principled Technologies Test Report commissioned by Dell Technologies comparing in-house
deployment vs. Dell EMC ProDeploy for Enterprise deployment service. Full report: http://facts.pt/JPiIlWm
Dell – Internal Use – Confidential 27 © Copyright 2021 Dell Inc.
276 of 292
6
Deployment Services | For PowerStore
Get production-ready with less time and effort
Additional services
Migration Services
ProDeploy Plus ProDeploy &
ProDeploy Plus
Recommended
Add-on Services Residency Services
• 2-Host Addition
Data Sanitization
• File Addition*
ProDeploy • Local Protection
for Enterprise
(snaps or thin clones)
Minimum for multi- Data Destruction
appliance clustering • Remote Replication for Enterprise
• Hardware Components
Hardware & system software • Expansion Enclosures
installation services Additional Deployment Time
NOTE: Custom deployment is also available. *PowerStore T models

Dell – Internal Use – Confidential © Copyright 2021 Dell Inc.
277 of 292
Dell EMC | Data Migration Services
The easy way to transport data
Dell EMC 3rd party Private Public
Consistency Transparency Integrity Value

Repeatable, dependable Detailed visibility of the process validated through post migration Minimal disruption and
processes and milestones verification customer effort
Move data with ease to PowerStore

PLAN AND MANAGE MIGRATIONS
PREPARE EXECUTE FOLLOW-UP
27 © Copyright 2020
2021 Dell Inc.
278 of 292
8
Dell EMC Residency Services
Set your business up for future

success and growth
1 Unmatched 2 Elevated 3 Individualized
expertise technology service
Highly specialized Get new solutions working Dell Technologies’
experts with best in to their full potential, extensive global network
class technical individualized to your provides utmost flexibility
certification across environment to drive value via onsite, remote and
products and solutions back to the business short-term engagements
45% 50% 40% 47% 2 months saved

Improvement in Improvement in technology Reduction in cost of Increase in 23 fewer incidents per month
IT staff efficiency performance IT service delivery technology utilization when transitioning to new technology
* IDC InfoBrief, sponsored by Dell Technologies, The Value of Resident Engineers on your IT Transformation Journey, June 2020
Secure data on retired, returned or redeployed systems
Data Sanitization & Data Destruction for Enterprise
Enable Compliance Safely and securely protect data from

US government NIST & unauthorized access
Certification
• Data Sanitization renders data unrecoverable
through a process of overwriting the data*
Mitigate Risk
Secures and protects data • Data Destruction physically destroys the
from unauthorized access
device (available Nov 17, 2020)
Customers receive a Certificate of Compliance aligned

Sustainability to NIST 800-88 standards to satisfy requirements
Proper, responsible regarding secure handling of regulated data.
recycling
*Data Sanitization requires systems to be fully operational and healthy; Data

Destruction applicable to non-operational systems.
Support Services | For PowerStore
Get ahead of problems before they happen
Additional support services
ProSupport Plus Anytime Upgrade Keep Your Hard Drive

Select or Standard For Enterprise
Recommended
Pre-requisite:
ProSupport, ProSupport Plus or
ProSupport One for Data Center
Keep Your Component
for Enterprise
ProSupport Optimize for Storage
Pre-requisite:
Available with 4-hour mission critical or next ProSupport, ProSupport Plus or Onsite Diagnosis
Available with 4-hour mission critical or next ProSupport One for Data Center
Available withbusiness
4-hourdaymission critical
or next business
business day day
Note: Customers with 1000+ Dell EMC infrastructure assets may be eligible for ProSupport One. Learn more.
PowerStore includes a 1-year limited warranty on hardware only.
281 of 292
Optimize for Storage
PowerStore technical experts work remotely
to optimize performance & efficiencies
• Highly skilled Dell EMC technical expert -- a trusted

advisor to your storage management team
• Year-round guidance on the operational health of your
storage solution
• Deep-dive performance evaluations & recommendations
98% CUSTOMER SATISFACTION1

It’s like a personal coach for
Up to 21% fewer Severity 1 events2
your PowerStore system
Requires CloudIQ and ProSupport or ProSupport Plus
Because even top performers
1Jan
need expert guidance
2020 voice-of-the-customer survey. 2Based on internal analysis of active service tags as of June 2020 comparing
the percentage of Severity 1 cases for SC Series systems with Optimize for Storage vs. systems without it. Actual
results may vary.
282 of 292
Optimize for Storage ProSupport
ProSupport Optimize
with ProSupport Enterprise Suite Plus for Storage
Remote technical support 24x7 24x7
Hardware,
Covered products Software
Hardware, Software
Next business day Next business day

Onsite hardware support or Mission Critical or Mission Critical
Automated issue detection and proactive case creation  
Self-service case initiation and management  
Access to software updates  
Priority access to specialized support experts 
3rd party software support 
Assigned Service Account Manager (SAM) 
Personalized assessments and recommendations 
Semiannual systems maintenance 
In-region Technical Account Manager, a storage expert 
Daily system monitoring 
Deep-dive configuration and performance evaluations1 
Annual strategic planning meeting2 
Ongoing best practice advice, technical guidance and knowledge transfer 
Assistance with critical support issues 
1Monthly or quarterly options available. 2Strategic planning meeting is onsite for monthly option. Note: Availability and terms of service vary by region and by product.

Maintain complete control of highly sensitive data
Keep Your Hard Drive for Enterprise | Keep Your Component for Enterprise
Retain possession of failed drives or Greater Security

components when receiving replacements Secures classified,
proprietary or sensitive data
without incurring additional cost
• Keep Your Hard Drive includes standard, Solid-
State Drive (SSD), Serial ATA (SATA), Hard Disk Industry Compliance
Drives (HDDs), including PCIe and NVMe Helps comply with data
privacy regulations
• Keep Your Component includes all drives
plus all other customer and field replaceable
components such as Boss cards, power supplies, Complete Control
motherboards, memory and more Sensitive data never leaves
your control

Data-driven Technologies
to connect and manage your service
experience

Unlock smarter, adaptive services
for your PowerStore solution
Connect & Monitor

SupportAssist Secure Remote Services
Adaptive AI for predictive, effortless support
Manage & Optimize

MyService360 CloudIQ
Actionable intelligence for services and product health

Connectivity with SupportAssist Enterprise
Support technology for your Dell EMC devices
Enhance your support experience

with secure connectivity
• Proactive, automated issue detection, case
creation and notification
• Remote issue resolution
• Best-in-class security from onboarding to
resolution
Choose your path to connectivity

Directly connect or connect to a Gateway to manage your connected devices all in one place
➢ SupportAssist Enterprise 4.0.5 or SRS 3.38 or higher

MyService360 An intuitive, streamlined experience
Actionable intelligence to proactively • Standardized navigation and features
manage services health and risks for
PowerStore Distill priorities for action
• Interactive KPIs identify relevant services metrics
New! Heads-Up Display with Do more to optimize your environment

tailored KPIs by category • Monitor PowerStore health with CloudIQ in near-real time
Get the right information on-demand

• One-click access from tables to a curated summary of
the most important data by product
Current TechDirect users can access technical support

for PowerStore. Please review details and scope for
Incident Management Example both TechDirect and MyService360.
Data Protection | Storage | Converged Systems
Dell – Internal Use – Confidential 288 of 292 © Copyright

© Copyright
2021 Dell
2020Inc.Dell Inc.
Proactive Insights That Drive Results
Technologies that fuel a better storage experience—data-driven health scores and proactive support
Connected through SupportAssist & Secure Remote Services
CloudIQ ProSupport Enterprise Suite

Make storage management tasks easier Identify, prevent and resolve issues faster with
with 24x7 proactive health and predictive automated, proactive support and issue resolution.
storage insights for better, faster
decision-making.
up to1 up to1 up to1
97/100 31% 36% 89%

Average Health Score for Fewer Faster issue Faster response
CloudIQ-connected arrays issues resolution to critical issues
with Dell EMC ProSupport Plus for Enterprise
Keeps you in control Prevents and solves issues for you
1 Based on an August 2020 internal analysis of service requests from August 2019 to August 2020 for Dell EMC storage, data protection and hyperconverged products comparing service requests for
connected products with ProSupport Plus for Enterprise vs. products without it. Connectivity is via Secure Remote Services. Actual results may vary..
28
289 of 292 © Copyright 2020
2021 Dell Inc.
9
Dell Technologies
Education Services

Education Services
Cutting-edge training and certifications
What are they?

• State-of-the art training and
certification programs for
PowerStore
• Instructor-led classroom
• Virtual classroom
• On-demand courses
• Labs
• Blended learning
Key Benefits:
• Keep internal staff up to date and
certified on latest technologies
• Improve IT staff efficiency
• Maximize infrastructure utilization
For details, go to: https://education.dellemc.com/content/emc/en-us/home/training/storage/powerstore.html

Dell-EMC Powerstore Technical Overview

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dell-EMC Powerstore Technical Overview

Uploaded by

Copyright:

Available Formats

PowerStore Technical

2 of 292 © Copyright 2021 Dell Inc.

3 of 292 © Copyright 2021 Dell Inc.

4 of 292 © Copyright 2021 Dell Inc.

5 of 292 © Copyright 2021 Dell Inc.

• Overall family brand that encompasses all

6 of 292 © Copyright 2021 Dell Inc.

PowerStore PowerStore PowerStore PowerStore PowerStore PowerStore

7 of 292 © Copyright 2021 Dell Inc.

User Data/Metadata 1,920 ✓

SAS SSD User Data/Metadata 3,840 ✓

User Data/Metadata 7,680 ✓

User Data/Metadata 1,920 ✓ ✓

User Data/Metadata 3,840 ✓ ✓

User Data/Metadata 15,360 ✓ ✓

User Data/Metadata 375 ✓ ✓

NVMe SCM User Data/Metadata 750 ✓ ✓

User Data/Metadata 1,500 ✓ PowerStoreOS

NVMe NVRAM Cache 8 ✓

• All drive offerings are encrypted (SEDs)

Appliance 1 Appliance 2 Appliance 3 Appliance 4

PowerStore PowerStore PowerStore PowerStore

Clusters Scale-Out with Appliances*

*Clustering is supported for PowerStore X appliances starting in PowerStoreOS 2.0

10 of 292 © Copyright 2021 Dell Inc.

Base Enclosure (Rear View)

• 4-Port Cards are inserted in Mezz 0

15 of 292 © Copyright 2021 Dell Inc.

– 4-Port 25GbE SFP based I/O Module

16 of 292 © Copyright 2021 Dell Inc.

• Every system has either 2 or 4 NVMe NVRAM Write Cache drives

17 of 292 © Copyright 2021 Dell Inc.

• Customers have the option to use NVMe SCM drives in their

18 of 292 © Copyright 2021 Dell Inc.

19 of 292 © Copyright 2021 Dell Inc.

• Customers have the option to use NVMe SSD drives in their

20 of 292 © Copyright 2021 Dell Inc.

21 of 292 © Copyright 2021 Dell Inc.

22 of 292 © Copyright 2021 Dell Inc.

Drive Type Location Purpose

NVMe SCM Slots 0-20 in base enclosure User Data/Meta Data*

SAS SSD Slots 0-24 in expansion enclosure User Data/Meta Data

23 of 292 © Copyright 2021 Dell Inc.

• Supports 1800W High Line Only • Supports 2100W High Line

• Uses C13/C14 or C13/C20 cables only • Uses C19/C20 cables only

Poseidon Power Cord

Atlas Power Cords

24 of 292 © Copyright 2021 Dell Inc.

• Applies to following Base Enclosure models:

26 of 292 © Copyright 2021 Dell Inc.

• 25x U.2 Drive Slots capable of supporting:

• Internal DRAM caching therefore no need for NVRAM drives

27 of 292 © Copyright 2021 Dell Inc.

• Each node will contain 1x Intel

• Each node will contain 96GB of

28 of 292 © Copyright 2021 Dell Inc.

• Consist of 2x Nodes Embedded Module

– Node A – Bottom Half

• Each Node includes an “Embedded

– 1x 4-Port Card (Optional) - “Mezz 0”

29 of 292 © Copyright 2021 Dell Inc.

• Each node has 2x IO Module slots Embedded Module

for optionally adding more front end

• Mgmt is done with the 1GbE ports