Optimizing Performance in OpenStack CloudsSteve Croce, Senior Product Strategist, Cloud Solutions, Dell

Alok Prakash, Product Manager, Cloud Platforms Group, Intel Corporation

Today’s session

• Dell’s OpenStack Cloud Solution design tenants– Co-engineering – Reference Architectures– Extensions

• Intel: Why and how performance matters in OpenStack – The hidden impacts if performance is not managed– Noisy neighbors – Service Compute Units, Workload placement

OpenStack for the enterprise

• Repeatable, validated end-to-end

• Fully supported software, infrastructure

• Massively scalable, elastic infrastructure

• Solutions to get you agile

• Efficient deployment, rapid time to value

• Performance tuned configurations

• Extended lifecycle support

• Expertise to complement in-house skills

Snowflakes and Houses of Cards

Unstable

Not Repeatable

Not Optimized

Hard to Support

Impossible to Modify

It WILL collapse! One-of-a-Kind!

• Address the gaps around OpenStack for Enterprise use cases via a jointly engineered solution

• Deliver enterprise-grade, certified, highly-scalable, secure, and fully supported cloud infrastructure solutions

• Provide a fulfillment, deployment, and support experience like commercial software

• Support specific enterprise use cases for cloud applications, software defined storage, development and test, cloud applications

Dell Red Hat Cloud Solutions Make

OpenStack the defacto

OPEN

Cloud Solution

in the

Enterprise

Solution components

Dell PowerEdgeDell EqualLogicDell Networking

Dell reference architecture

Red Hat Enterprise Linux OpenStack Platform

Dell Professional ServicesRed Hat Professional Services Dell ProSupport

Dell Red Hat Enterprise Cloud

Solution Extensions

Co-Engineered Reference Architecture

• Validated and Prescriptive – Server configurations

– Network Design and configurations

– Storage configurations

– System sizing recommendations

• Robust, Reliable, Repeatable – Blueprint for success

– Capture and package joint expertise

– Use cases, best practices

– Accelerate your time to value

No more houses of cards

Cloud Management

Performance Management

Software Defined Networking

Solution extensions*

Application Management

* Consult Dell Ref. Architecture for complete details

∞ Investing in OpenStack

∞ Vendors you can count on!

∞ Validated with Dell Ref. Architecture*

∞ Deployment guidance*

∞ Sizing guidance and recommendations*

∞ Professional services and training

∞ Deployment services and tools

∞ Collaborative support

Dell and Intel Service Assurance Administrator

* Under development

Intel: Why performance matters

Intel: Why performance matters

Software Defined Infrastructure – Service Assurance

Automate On Standard IA Hardware

STATIC, MANUAL INFRASTRUCTURE

Storage Network Compute

Resource Pool

Orchestration Software

Infrastructure AttributesInfrastructure AttributesGovernance,

PolicyCooling

Power Trust and Compliance

Application A

Application B

Application C

Application D

Intelligent service monitoringand operational event alerts

SLA Requests

Monitoring

UtilizationPerformance LocationThermalsCompliancePower

Orchestration Software

Trust

Performance

PublicApplication Workloads & Data

Private

Physical and virtualized machine management

Challenges in Running Enterprise and Telco Workloads in the Cloud

CPU IOPS Memory Memory b/w CPU cache Instruction set

Capability Capacity Consumption Availability Resilience

DATA REGULATIONS

LEGACYWORKLOADS

Telemetry is hiddenCOMPLIANCE

Is my workload running on trusted infrastructure?

How do I avoid noisy neighbors?

Will my workload get

expected compute cycles?

Intel® Service Assurance Administrator enhances OpenStack Cloud Services

Nova Scheduler

Plug-In

Compute Node

Machine Flavor

Creator

Analysis & Remediation

Engine

Service AssuranceController

Monitoring Engine

Capacity Insight

REST API

Web Admin

Console

Compute NodeCompute Node Agent

The Challenge Workloads do not get predictable performance Limited trust of multi-tenant nodes Grey machines (unhealthy state) are not avoided Workload scheduling is less than optimal

www.intel.com/assurance

The Solution• Intelligent Workload Placement based on ability of nodes

to meet service level objectives• Compute node capacity and utilization metrics normalized

for use across generation of CPUs• VM performance monitoring and assurance with cache

contention and ‘Noisy Neighbor’ detection• Reporting of trust of VMs and nodes – BIOS, Hypervisor

whitelisting and attestation

Why Intel® Service Assurance Administrator?

Match workloads to platforms, based on capability and

capacity

Find and address software-defined infrastructure issues

Service assurance for trust-attestation

• Performance metric infrastructure capacity and utilization

• Cache Contention• System health• Resource metering

• Boot time trust attestation of nodes

• Whitelisting

Intel® Architecture Platform Monitoring and Control

“What size virtual machine must I use for my app?”

“My app is slow sometimes –How do I diagnose? “

Do I have ‘noisy neighbor’ VMs?”

“Are my VMs running on trusted platforms?”

AutomationEnhance OpenStack* to provision and monitor machine flavors with specified

service levels

Automated software-defined infrastructure

Intel® Service Assurance Administrator

EfficiencyIntegrate with IT operations tools to

determine probable root cause, report, and help remediate issues

Efficient service assurance and administration

Agility

Run workloads with confidence on software-defined infrastructure

Agile business service deployment

Enhanced Machine Flavors

Key Challenges in Cloud Performance AssuranceHow big a

VM do I need for

my workload?

What resources should be

reserved, how much burst capacity?

What metric do I use to specify and

measure performance?

What is the performance

capacity of the node?

Intel® Service Assurance Administrator defines Service Compute Unit as a performance metric for use by cloud administrators

What is the performance capacity

of the machine flavor?

Am I getting specified performance without

interference from ‘noisy neighbors’?

Linux

Compute node

VM VMVM VMKVM

AppAppApp

Services

VM

Monitoring Reporting

Platform monitoring to detect and avoid ‘noisy neighbors’

Cache ContentionMemory b/w

Resource Metering

Noisy Neighbor’s impact on VM performance

Share

d L3

C

ach

e

Core

Core

Core

Core

Core

Core

Core

Core

Shared Infrastructure

Share

d L3

C

ach

e

Core

Core

Core

Core

Core

Core

Core

Core

HypervisorOS Application & Services

VMVM VMVM

What is a Service Compute Unit (SCU)?

18

Provisioning Orchestration

Match workloads to platforms, based on capability and capacity

Service Compute UnitsPerformance metric for cloud

infrastructure capacity and utilization

vCPUApproach

Intel® Service Assurance

Administration

Resource Allocation

Static, processor dependent

SCU – dynamic, processor

independent

Machine Portability

Not portable across

processor family or

generations

SCU is portable across generations

of Intel CPU and across processor

familyPerformance Service Level Monitoring

NoneService Assurance of

SCU units

SCU = function of (frequency, throughput, instruction set efficiency, cache size) 1 GCEU = 2.5 SCUs (Approximate)1 EC2 = 2.5 to 3 SCUs (Approximate)SCUs can be set to define a floor (guaranteed capacity) and ceiling (burst capacity)

How much performance capacity are available on my host?

19

For each compute node, performance capacity status for OS and VMs are

monitored continuously, displaying how much resource is available, e.g. SCU and

number of cores available

How much cache contention are there in the compute nodes?

20

Using Cache Monitoring Technology, console displays LLC Cache Contention

status for the OS and VM on each compute node

Extending flavors with VM bursting and assurance capabilities

21

VM will only be created on nodes that have been trust-attested to be safe during

boot

VM will reserve 1 SCU of compute performance, and burst up to 2 SCU

based on available compute resources

Extending flavors with trust and core-pinning capabilities

VM will reserve one dedicated core from the CPU compute resource

VM will only be created on nodes that have been trust-attested to be safe during

boot

Finding out the performance consumption level for each compute node at the OS level

Allocating SCU for OS level consumption of

compute resource

Finding out VM core-pinning status and metrics Core Pinning, VM received dedicated core from the

CPU

VM SLO status include number of cores assigned

and VM uptime

Learning more about each VM SCU utilization metrics

Real Time VM SCU utilization and core

pinning metrics

Conducting probable cause analysis of “Noisy Neighbors” problems

Cloud Administrator can take remedial action (e.g. evacuate a VM to a different node) once Intel® SAA has identified Noisy Neighbors (VMs aggressively using shared compute resources in the platform, like CPU cache) and VMs that are affected.

Intel® Service Assurance Administrator Key Feature SummaryFeature Customer pain point Solution

Compute throughput metering

How do I know how many VMs and workloads I can put on a compute node?How do I specify how much compute capability is needed by my VM?

SCU, service compute unit, a portable compute performance metric • Characterize node’s compute capacity• Specify required VM capacity• Measure utilization

Host OS and App Services Assurance

How do I know that applications or OS services such as Ceph or backup are not creating issues for my VMs?

Enable creation of Assurance SLA to grant compute quota to Host Apps and OS services

Contention Aware Scheduling

How do I detect ‘noisy neighbor’ VMs and affected VMs? How do I automate the VM scheduling process?

Intelligent Workload Placement based on ranking of node’s ability to meet service level objectives, using telemetry

Grey Machine Avoidance How do I avoid systems that might not be healthy, e.g. running hot with fan failure?

Gather system health info via IPMI protocols and use as a metric in workload placement

Trusted Compute Pool –Policy enforcement

How do I find out that my workloads are running on trusted nodes?

Use Intel® Trusted Execution Technology for boot time attestation of each nodes FW/SW

Contact the Dell Red Hat OpenStack Solutions team• openstack@dell.com

Engage a Dell Solution Center (contact your Dell Rep) • dell.com/solutioncenters

Dell Cloud resources sites• Dell.com/openstack• Dell.com/redhat• youtube.com/Dell

Intel SAA Resources Site• www.intel.com/assurance• Contact Intel SAA team by clicking “Contact Us”

Resources & Contacts

Questions?

Thank you!

Intel Confidential — Do Not Forward31

Legal Disclaimers:

INFORMATION IN THIS DOCUMENT IS PROVIDED IN CONNECTION WITH INTEL PRODUCTS. NO LICENSE, EXPRESS OR IMPLIED, BY ESTOPPEL OR OTHERWISE, TO ANY INTELLECTUAL PROPERTY RIGHTS IS GRANTED BY THIS DOCUMENT. EXCEPT AS PROVIDED IN INTEL'S TERMS AND CONDITIONS OF SALE FOR SUCH PRODUCTS, INTEL ASSUMES NO LIABILITY WHATSOEVER AND INTEL DISCLAIMS ANY EXPRESS OR IMPLIED WARRANTY, RELATING TO SALE AND/OR USE OF INTEL PRODUCTS INCLUDING LIABILITY OR WARRANTIES RELATING TO FITNESS FOR A PARTICULAR PURPOSE, MERCHANTABILITY, OR INFRINGEMENT OF ANY PATENT, COPYRIGHT OR OTHER INTELLECTUAL PROPERTY RIGHT.

A "Mission Critical Application" is any application in which failure of the Intel Product could result, directly or indirectly, in personal injury or death. SHOULD YOU PURCHASE OR USE INTEL'S PRODUCTS FOR ANY SUCH MISSION CRITICAL APPLICATION, YOU SHALL INDEMNIFY AND HOLD INTEL AND ITS SUBSIDIARIES, SUBCONTRACTORS AND AFFILIATES, AND THE DIRECTORS, OFFICERS, AND EMPLOYEES OF EACH, HARMLESS AGAINST ALL CLAIMS COSTS, DAMAGES, AND EXPENSES AND REASONABLE ATTORNEYS' FEES ARISING OUT OF, DIRECTLY OR INDIRECTLY, ANY CLAIM OF PRODUCT LIABILITY, PERSONAL INJURY, OR DEATH ARISING IN ANY WAY OUT OF SUCH MISSION CRITICAL APPLICATION, WHETHER OR NOT INTEL OR ITS SUBCONTRACTOR WAS NEGLIGENT IN THE DESIGN, MANUFACTURE, OR WARNING OF THE INTEL PRODUCT OR ANY OF ITS PARTS.

Intel may make changes to specifications and product descriptions at any time, without notice. Designers must not rely on the absence or characteristics of any features or instructions marked "reserved" or "undefined". Intel reserves these for future definition and shall have no responsibility whatsoever for conflicts or incompatibilities arising from future changes to them. The information here is subject to change without notice. Do not finalize a design with this information.

The products described in this document may contain design defects or errors known as errata which may cause the product to deviate from published specifications. Current characterized errata are available on request.

Intel product plans in this presentation do not constitute Intel plan of record product roadmaps. Please contact your Intel representative to obtain Intel's current plan of record product roadmaps.

Intel processor numbers are not a measure of performance. Processor numbers differentiate features within each processor family, not across different processor families. Go to: http://www.intel.com/products/processor_number.

Contact your local Intel sales office or your distributor to obtain the latest specifications and before placing your product order.

Copies of documents which have an order number and are referenced in this document, or other Intel literature, may be obtained by calling 1-800-548-4725, or go to: http://www.intel.com/design/literature.htm

Code names featured are used internally within Intel to identify products that are in development and not yet publicly announced for release. Customers, licensees and other third parties are not authorized by Intel to use code names in advertising, promotion or marketing of any product or services and any such use of Intel's internal code names is at the sole risk of the user

Intel, and the Intel logo are trademarks of Intel Corporation in the United States and other countries.

*Other names and brands may be claimed as the property of others.

Copyright ©2013, 2014, 2015 Intel Corporation. Intel Corporation. All rights reserved. Intel and the Intel logo are trademarks of Intel Corporation in the U.S. and/or other countries. *Other names and brands may be claimed as the property of others.

The OpenStack Word Mark and OpenStack Logo are either registered trademarks/service marks or trademarks/service marks of the OpenStack Foundation, in the United States and other countries and are used with the OpenStack Foundation's permission. We are not affiliated with, endorsed or sponsored by the OpenStack Foundation, or the OpenStack community."

Intel Confidential — Do Not Forward

Intel's compilers may or may not optimize to the same degree for non-Intel microprocessors for optimizations that are not unique to Intel microprocessors. These optimizations include SSE2, SSE3, and SSE3 instruction sets and other optimizations. Intel does not guarantee the availability, functionality, or effectiveness of any optimization on microprocessors not manufactured by Intel.

Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors. Certain optimizations not specific to Intel microarchitecture are reserved for Intel microprocessors. Please refer to the applicable product User and Reference Guides for more information regarding the specific instruction sets covered by this notice.

Notice revision #20110804

32

Optimization Notice

Dell and Red Hat Enterprise Cloud SolutionsPrescriptive, practical, right-sized choices fit for you

PoC System• Dev/test• Concept testing• Prototyping• Rapid deploy

Pilot System• Mid-scale production• Scale testing• Sizing options, 1-3 racks • 10G networking • High Availability• Optimized Inktank Ceph

Scale out System

• Data center scale• Production workloads• PowerEdge R, C Series• 10G networking • High Availability• Optimized Inktank Ceph• Tailored designs• Customized integrations

with OpenStack ecosystem technologies

Shift your focus from infrastructure to service

IT silos

Serv

er

Sto

rag

e

Ne

two

rkin

g

Soft

war

e

Man

agem

ent

Management

Open standards-based hardware

Infrastructure control

Enterprise applications & workloads

Mu

lti-

ven

do

r, c

ross

-pla

tfo

rm

un

ifie

d m

anag

emen

t

Orchestration stack