Stratus ztC Endurance™ Platform

Overview
Intelligent, predictive fault tolerant computing
delivering 99.99999% availability

Presenter Name
Date
1 | © 2023 Stratus Technologies. All Rights Reserved.
Table of Contents

FRONT

Stratus ztC Endurance Introduction &

Overview (Slide 3)

Additional modules

• Technical details (Slide 25)

• Fault tolerant architecture (Slide 31)

• Performance benchmarks (Slide 39)

BACK

2 | © 2023 Stratus Technologies. All Rights Reserved.

What We Do

Deliver zero-touch computing for continuous availability of mission-critical applications

Simple Protected Autonomous

Easy to install and manage Fault tolerant Seamless failover/redundancy

Highly interoperable Self-monitoring and healing Automated replacement
Manageable by IT or OT Secure Remote management

3 | © 2023 Stratus Technologies. All Rights Reserved.

A Giant Leap Forward in Fault Tolerant Computing

Stratus ztC Endurance • Increased performance and scalability

• Intelligent, predictive availability layer
• Increased modularity and serviceability
• More remote management capabilities
• Extensible for future use cases

40 years of innovation • Built in fault tolerance

in fault tolerance • Continuous availability and no data loss
• Pro-active health monitoring
• Single OS image
• Bare-metal and virtualization
• Easily serviceable by IT or OT

4 | © 2023 Stratus Technologies. All Rights Reserved.

What is the Stratus ztC Endurance Platform?

Stratus ztC Endurance is an evolutionary, new family of

intelligent, predictive fault tolerant computing platforms that
delivers predictable, protected performance for next generation,
sustainable operations.

Using ztC Endurance, organizations are able to support

complex business operations at the edge or the data center
and run advanced software with the assurance of fault tolerant
computing that is simple to service and manage.
• x86-based computing platform with
built-in computing redundancy
Key benefits • 2U rack mount size
• 99.99999% system availability • Eight hot-swappable customer
replaceable unit (CRU) modules
• Serviceable by OT and IT
• Run “bare metal” OS or with hypervisor
• Predictive and proactive health monitoring

Seven nines (99.99999%) availability of the Stratus ztC Endurance computing platform is based on requirements that customers use parts authorized for Stratus ztC Endurance, maintain an
active Stratus Support contract, and perform system updates recommended by Stratus.

5 | © 2023 Stratus Technologies. All Rights Reserved.

Predictive – Resolves Issues Before They Occur

Stratus Automated Uptime Layer with Smart Exchange

• Continuous monitoring, diagnosing, self-healing,
and managing to maintain availability
• Identify>Isolate>Service approach
• Automated: requires no human intervention
• Transparent to operating systems and applications
(operates like single system)
• Active/standby approach to compute module
availability
• Automatically activates standby module and manages
transfer of data and instructions from active to standby

6 | © 2023 Stratus Technologies. All Rights Reserved.

 Serviceable – Easy for OT or IT to Maintain

• Redundant, modular architecture featuring

4 pairs of customer replaceable unit (CRU) modules
• Dual compute modules
• Dual I/O modules
• Dual storage module
• Dual Power supply units
• Replaceable by OT or IT without need for specialized
expertise or tools to remove or replace CRUs
• Supports uptime and efficient operation

7 | © 2023 Stratus Technologies. All Rights Reserved.

Performance – Drive Higher Levels of Performance

• Latest technology to deliver new levels of

performance
• Intel 4th generation Intel Xeon Scalable
“Sapphire Rapids” chips
• Up to 48 cores/96 threads per platform
• Leverages advanced RAS capabilities
• Advanced performance capabilities
• AVX, Turbo Boost
• High speed native NVMe storage – faster
read/write capability
• Resilient, high-performance Zefr DDR5
memory – for greater compute performance
• Gen 4 PCIe cards – faster I/O performance

8 | © 2023 Stratus Technologies. All Rights Reserved.

How It Works – Stratus Platforms Engineered for Unmatched Reliability
1. Fully redundant hardware
2 x I/O modules
2 x storage modules
(12 x NVMe drives)

2 x compute 2 x power supply

modules units (PSU)

How Benefit
• Four (4) pairs of redundant customer replaceable unit (CRUs) modules • Compute modules are hot swappable by OT
built with industry-standard components - compute, storage, I/O, and or IT users; does not require specialized tools
power module
• Platform predicts a failure and automatically
• Redundant modules leverage multi-path IO for availability moves workload from active compute module
• Stratus hardened drivers key for I/O and storage redundancy to standby compute module in seconds

9 | © 2023 Stratus Technologies. All Rights Reserved. See technical details

 How It Works – Stratus Platforms Engineered for Unmatched Reliability
2. Hardware-based fault tolerance

How
• PCIe Fabric switch connects “Active” compute module and “Standby”
compute module
• Moves CPU state, OS, and workload from “Active” compute module to
“Standby” compute module before a failure occurs

Benefit
• Hardware-based fault tolerance eliminates failover time
“Active” • No loss of in-flight data or pause in transaction processing
compute CRU • Does not require software modification or failover scripts
“Standby”
module Smart compute CRU
Exchange module
Fabric
switch

10 | © 2023 Stratus Technologies. All Rights Reserved.

How It Works – Stratus Platforms Engineered for Unmatched Reliability
3. Stratus Automated Uptime Layer (AUL) with Smart Exchange

How
Alerts • Stratus Automated Uptime Layer software with Smart Exchange
Stratus End user
works across the ztC Endurance system
• Manages identification, isolation, and service of errors
Stratus AUL- • Monitors 500 points of platform health and component
Smart performance
Exchange • Initiates Smart Exchange if predicting unrecoverable error in
Software compute module

Benefits
• Provides platform self-diagnosis, self-healing, and pro-active
health monitoring including alerting
• Platform health alerts are sent via the Stratus ActiveService™
Network (ASN) or through standard protocols such as SNMP traps,
OPC UA, REST APIs, other

11 | © 2023 Stratus Technologies. All Rights Reserved.

See details about Smart Exchange
ztC Endurance Summary – Technical Advantages for IT and OT Environments

• Maximum Availability
• Fully Redundant Hardware
• Automated Uptime Layer with
Smart Exchange™
• No Failover Time/No Data Loss
• Operational Simplicity
• Single System - Single License
• Serviceable by OT and IT
• 24/7/365 Support Service
• Industry standard for IT
• Out of the box deployment
• System Life 7-10+ Years
• Lowest TCO for fault tolerance

12 | © 2023 Stratus Technologies. All Rights Reserved.

Stratus ztC Endurance Platform Family

Specifications 3100 5100 7100

Sizing Up to 12 VMs Up to 24 VMs 40 VMs and up

1 x Intel® Xeon® Silver 4410Y 2 x Intel® Xeon® Silver 4410Y 2 x Intel® Xeon® Gold 5418Y
Processors 2.0 GHz 2.0 GHz 2.0 GHz
3.9 GHz max (Turbo) 3.9 GHz max (Turbo) 3.8 GHz max (Turbo)

Cores/Threads 12 Cores/24 Threads 24 Cores/48 Threads 44 Cores/88 Threads

Min/max memory 64 GB/256 GB DDR5 128 GB/512 GB DDR5 128 GB/1024GB DDR5

Max internal storage 38.4 TB NVMe 38.4 TB NVMe 38.4 TB NVMe

Connectivity 1 x 1GbE, 2 x 10 GbE 1 x 1GbE, 2 x 10 GbE 1 x 1GbE, 2 x 10 GbE

Expansion slots Up to 5x PCIe Gen 4 Cards Up to 5x PCIe Gen 4 Cards Up to 5x PCIe Gen 4 Cards

Ethernet ports 1 x 1GbE, 2 x 10 GbE 1 x 1GbE, 2 x 10 GbE 1 x 1GbE, 2 x 10 GbE

OS Support VMware® vSphere, Microsoft Windows Server® with or without Hyper-V, Red Hat Enterprise Linux*

13 | © 2023 Stratus Technologies. All Rights Reserved. *Check with your local distributor for specific availability
 Fault Tolerant Architecture

14 | © 2023 Stratus Technologies. All Rights Reserved.

Stratus Approach to Continuous Availability

Identify
• Proactively monitor system health events
• Predictively identify system availability risk

Isolate
• Self-healing transfer of operational state to healthy module
• Include bare metal and virtualized state

Service
• Notification of failing module sent to Stratus
• Customer replaces failing module with good module

15 | © 2023 Stratus Technologies. All Rights Reserved.

How Smart Exchange Works

Identification of Storage-0

• Processor Faults
Active Customer • Intel processors support bus width reduction
Identify
Workload for bus failures IO-0
• QPI, DMI, PCIe
• Bus width and speed are minimized to
isolate failure
• Memory Faults Storage-1
• Coverage for transient and persistent fault
utilizing RAS
Standby Standby • Power / Electrical FaultsIO-1
OS • Thermal / Airflow Faults
PCI
Express Fabric

Compute Modules Passive Backplane Storage and I/O

Modules

16 | © 2023 Stratus Technologies. All Rights Reserved.

How Smart Exchange Works

Smart Exchange
Storage-0

Standby
Active Customer Transfer of operational state without application
Service
Workload downtime IO-0

1. Move Workload to Standby

2. Storage-1
Reconfigure PCIe Fabric

Active
Standby Standby 3. Promote Standby to Active
OS IO-1
PCI
PCI
Express Fabric
Express Fabric

Compute Modules Passive Backplane Storage and I/O

Modules

17 | © 2023 Stratus Technologies. All Rights Reserved.