Cisco & OpenSource (context SDN & NFV)

Guillaume Gottardi – Solutions Architect – gottardi@cisco.com

How Cisco Participates In Open Source

• Contributions 100+ projects, for over 25 years

• Linux Kernel – 25th largest contributor in the world

• Protocol development: SCTP, XMPP, PTS/VNTAG, LISP, …

• OpenDaylight

• OpenStack

• OPNFV – Open Platform for NFV

• Apache SpamAssassin, OVS, …

• Eclipse Tigerstripe, Krikkit, …

• Sourcefire snort, razorback, clamav, …

• Mozilla Ikran, Open.h264

• ConfD

• And many, many more…

• Community participation

• Linux Foundation, Snort, Eclipse, Apache, FreeBSD, Mozilla, etc.

Networking Open Source Project Participation ~ 2 years

Cisco and OpenStack

• Cisco Validated Designs for production deployments

• Work closely and jointly with customers to design and build their OpenStack environment

• OpenStack based Global Intercloud hosted across Cisco and partners data centers

• Cisco Webex Service running on OpenStack

• Automation (Puppet) and architectures (HA) for production deployment and operational support

• Neutron/Nova Plug-ins for Cisco product lines – Nexus, CSR, ACI, UCS

• Code contributions across several services – Network. Compute, Dashboard, Storage

• Foundation Board member

Community Participation

Engineering/ Automation

Partners/ Customers

Cloud Services

OPNFV

• Open Platform for Network Function Virtualization (OPNFV)

• Realization of ETSI NFV Architecture

• Integration of open source components – “the glue”

• Interoperable across industry partners and usage models

• Active in upstream open source communities

• System integration as open community effort

OPNFV

• Open Platform for Network Function Virtualization (OPNFV)

• Realization of ETSI NFV architecture

• Integration of open source components –“the glue”

• Interoperable across industry partners and usage models

• Active in upstream open source communities

• System integration as open community effort

18

Evolution of Programmable Networking

• Many industries are transitioning to a more dynamic model to deliver network services

• The great unsolved problem is how to deliver network services in this more dynamic environment

• Inordinate attention has been focused on the non-local network control plane (controllers)

• Necessary, but insufficient

• There is a giant gap in the capabilities that foster delivery of dynamic Data Plane Services

Programmable Data Plane

Issues/Limitations with Existing Data Plane Solutions

• Known issues with Performance, Scalability & Stability

• Overly Complex Architectures

- Hard to evolve

- Slow rate of innovation

- Steep learning curve

• Hard to deploy/upgrade/operate

- slow cycles, too many kernel dependencies

• Lack of :

- automated end-to-end system testing frameworks

- leads to unpredictable system behavior

- support for diverse/custom hardware

- portability across compute platforms

- optimal use of compute microarchitectures

- network level instrumentation

- Few debugability features

- Few if any Statistics/Counters exposed

Fast Data Project – FD.io

• Collaborative open source project in Linux foundation

• High performance I/O services framework for dynamic computing

• User space I/O services framework

• Hardware, kernel, and deployment (bare metal, VM, container) agnostic

• 6WIND, Brocade, Cavium, Cisco, Comcast, Ericsson, Huaiwei, Inocybe, Intem, Mesosphere, Project Calico (Metaswitch), PLUMgrid, Red Hat

Introducing VPP – Vector Packet Processor Introducing Vector Packet Processor - VPP

§ VPP is a rapid packet processing development platform for

highly performing network applications.

§ It runs on commodity CPUs and leverages DPDK

§ It creates a vector of packet indices and processes them

using a directed graph of nodes – resulting in a highly

performant solution.

§ Runs as a Linux user-space application

§ Ships as part of both embedded & server products, in volume

§ Active development since 2002

DRAFT - Linux Foundation Confidential 4

NetworkIO

PacketProcessing

DataPlaneManagementAgent

BareMetal/VM/Container

VPP in the Overall Stack

fd.io Foundation 10

Hardware

Application Layer / App Server

VM/VIM Management Systems

Network Controller

Operating Systems

Data Plane Services

Orchestration

Network IO VPP Packet Processing

vSwitch FD.io evolution

VPP processes the vector of packets through a Packet Processing graph.

Openstack to VPP integration VPP vs OVS performance

https://fd.io/technology

VPP Feature summary VPP Feature Summary

DRAFT - Linux Foundation Confidential 8

14+ MPPS, single core

Multimillion entry FIBs

Source RPF

Thousands of VRFs

Controlled cross-VRF

lookups

Multipath – ECMP and Unequal

Cost

Multiple million Classifiers –

Arbitrary N-tuple

VLAN Support – Single/Double

tag

Counters for everything

Mandatory Input Checks:

TTL expiration

header checksum

L2 length < IP length

ARP resolution/snooping

ARP proxy

IPv4/IPv6 IPv4

GRE, MPLS-GRE, NSH-GRE,

VXLAN

IPSEC

DHCP client/proxy

CG NAT

IPv6

Neighbor discovery

Router Advertisement

DHCPv6 Proxy

L2TPv3

Segment Routing

MAP/LW46 – IPv4aas

iOAM

MPLS

MPLS-o-Ethernet –

Deep label stacks

supported

L2

VLAN Support

Single/ Double tag

L2 forwarding with EFP/

BridgeDomain concepts

VTR – push/pop/Translate

(1:1,1:2, 2:1,2:2)

Mac Learning – default limit of

50k addresses

Bridging – Split-horizon group

support/EFP Filtering

Proxy Arp

Arp termination

IRB – BVI Support with

RouterMac assignment

Flooding

Input ACLs

Interface cross-connect

========

TC5 120ge.vpp.24t24pc.ip4.cop

TC5.0 120ge.2pnic.6nic.rss2.vpp.24t24pc.ip4.cop

d. testcase-vpp-ip4-cop-scale

120ge.2pnic.6nic.rss2.vpp.24t24pc.ip4.2m.cop.2.copip4dst.2k.match.100

64B, 138.000Mpps, 92,736Gbps

IMIX, 40.124832Mpps, 120.000Gbps

1518, 9.752925Mpps, 120.000Gbps

---------------

Thread 1 vpp_wk_0 (lcore 2)

Time 45.1, average vectors/node 23.44, last 128 main loops 1.44 per node 23.00

vector rates in 4.6791e6, out 4.6791e6, drop 0.0000e0, punt 0.0000e0

Name State Calls Vectors Suspends Clocks Vectors/Call

TenGigabitEtherneta/0/1-output active 9003498 211054648 0 1.63e1 23.44

TenGigabitEtherneta/0/1-tx active 9003498 211054648 0 7.94e1 23.44

cop-input active 9003498 211054648 0 2.23e1 23.44

dpdk-input polling 45658750 211054648 0 1.52e2 4.62

ip4-cop-whitelist active 9003498 211054648 0 4.34e1 23.44

ip4-input active 9003498 211054648 0 4.98e1 23.44

ip4-lookup active 9003498 211054648 0 6.25e1 23.44

ip4-rewrite-transit active 9003498 211054648 0 3.43e1 23.44

---------------

Thread 24 vpp_wk_23 (lcore 29)

Time 45.1, average vectors/node 27.04, last 128 main loops 1.75 per node 28.00

vector rates in 4.6791e6, out 4.6791e6, drop 0.0000e0, punt 0.0000e0

Name State Calls Vectors Suspends Clocks Vectors/Call

TenGigabitEthernet88/0/0-outpu active 7805705 211055503 0 1.54e1 27.04

TenGigabitEthernet88/0/0-tx active 7805705 211055503 0 7.75e1 27.04

cop-input active 7805705 211055503 0 2.12e1 27.04

dpdk-input polling 46628961 211055503 0 1.60e2 4.53

ip4-cop-whitelist active 7805705 211055503 0 4.35e1 27.04

ip4-input active 7805705 211055503 0 4.86e1 27.04

ip4-lookup active 7805705 211055503 0 6.02e1 27.04

ip4-rewrite-transit active 7805705 211055503 0 3.36e1 27.04

Embedded Telemetry

Cisco Confidential

VPP vRouter/vSwitch: Local Programmability

1

4

Linux Host

Kernel

DPDK

VPP App External

App

Low Level API

• Complete

• Feature Rich

• High Performance

• Example: 500k routes/s

• Shared memory/message queue

• Box local

• All CLI tasks can be done via API

Generated Low Level Bindings - existing today

• C clients

• Java clients

• Others can be done

Cisco Confidential

VPP vRouter/vSwitch: Remote Programmability

fd.io Foundation 1

5

Linux Host

Kernel

DPDK

VPP App Data Plane

Management

Agent

High Level API: An approach

• Data Plane Management Agent

• Speaks low level API to VPP

• Box (or VM or container) local

• Exposes higher level API via some

binding

Flexibility:

• VPP does not force a particular Data

Plane Management Agent

• VPP does not force only *one* High

Level API

• Anybody can bring a Data Plane

Management Agent

• High Level API/Data Plane

Management Agent

• Match VPP app needs

netconf/yang REST Other (BGP)

Cisco Confidential

Continuous Performance Lab (CPL)

• Fully automated testing infrastructure

• Covers both programmability and data planes

• Continuous verification of code/feature

• Functionality and performance

• Code breakage and performance degradations identified before patch review

• Review, commit and release resource protected

• Fully open sourced test framework to be included at launch

Develop

Submit Patch

Automated Testing

Deploy

• Virtualisation / automation / orchestration has made real-time service provisioning possible

• Open source big data / analytics technologies now being widely applied outside of big web companies

• OSS architectures simply not keeping pace with the rest of industry

• No coherent industry direction on how OSS needs to change in the presence of these new technologies

Panda Context

Cisco Confidential 18 © 2013-2014 Cisco and/or its affiliates. All rights reserved.

OSS Analytics is a big data problem! i.e. OSS analytics applications can be addressed by performing a

query function against the entire OSS data set Fault management = (event data)

Performance management = (metric data)

Billing mediation = (event data, metric data)

Capacity management = (metric data)

Security analytics = (metric data, route data)

…

Relationship Between Orchestration and Analytics

Orchestration OSS Analytics

Data Center Core Access &

Aggregation

User

Sta

te

Da

ta

Related as loosely coupled

but tightly integrated systems

OSS analytics is responsible for

collecting data from the

infrastructure, monitoring and

analysis

The “F_APS” in FCAPS

Orchestration is responsible

for service provisioning and

pushes state to the

infrastructure

The “C” in FCAPS

We have platforms for

orchestration in NSO, ODL,

Openstack

We need a companion

platform for OSS Analytics

PlAtform for Network Data Analytics - Vision

• Simple, scalable, open big data / analytics platform

• Forms a generic big data analytics platform supporting different types of analytics applications for cloud based networks and services

- Operational Intelligence, e.g. OSS

- Business intelligence, e.g. BSS

• Leverage rapid innovation in Big Data analytics space

Orchestration

Controllers

Customer

Devices

Applications

QoE Monitoring

Infr

astr

uct

ure

an

d

serv

ice-

leve

l dat

a C

ust

om

er-l

evel

d

ata

Data

Distribution

Data Store

& Processing

Open Data Platform

Producers:

Data aggregation

Event Data

Log Data

Metric Data

Network Telemetry

Data Sources

App

App

App

App

App

App

App

Consumers:

Data analysis

Applications

Inventory Topology Geography Geography

Context:

Horizontally

Scalable Data

Platform

Applications

App

App

App

Live stream

PlAtform for Network Data Analytics - Principles

Orchestration

Controllers

Customer

Devices

Applications

QoE Monitoring

Infr

astr

uct

ure

an

d

serv

ice-

leve

l dat

a C

ust

om

er-l

evel

d

ata

Data

Distribution

Data Store

& Processing

Master Data

Store

Open Data Platform

Batch

Processing

Stream

processing

Real Time

Data Store

Deep H

isto

rical Q

uery

R

eal T

ime Q

uery

Producers:

Data aggregation

Event Data

Log Data

Metric Data

Network Telemetry

Data Sources

Inventory Topology Geography Geography

Context:

• Decouple data aggregation (publishers) from data analysis (consumers) – allow any OSS app the potential to access any data source

• Simple, scalable, open data distribution platform

- Scale-out architecture with support for horizontal scale in all core components

- Very highly available core platform

- Low and predictable latency

• Immutable dataset with minimal filtering/processing on ingress

• Analytics based approach to analysis functions

• Support for streaming apps, real-time queries and batch processing

Live stream

App

App

App

App

App

App

App

Consumers:

Data analysis

Applications

App

PlAtform for Network Data Analytics - Benefits

Data

Distribution

Data Store

& Processing

Master Data

Store

Open Data Platform

Batch

Processing

Stream

processing

Real Time

Data Store

Deep H

isto

rical Q

uery

R

eal T

ime Q

uery

Producers:

Data aggregation

Event Data

Log Data

Metric Data

Network Telemetry

Data Sources

Capacity Analytics

Billing (Mediation)

Business Intelligence

Fault Analysis

Perf Analysis

Log Analytics

Security and Threat Analysis

Inventory

Consumers:

Data analysis

Applications

Inventory Topology Geography Geography

Context:

Live stream

SNMP

Logs

SNMP

Monit,

Collectd,

Logstash,

Ceilometer

Netflow

• An open system architecture

• Collect data once – allow any analysis application to mine any data source, leveraging the full value of the OSS dataset

• Extensible – add new OSS analysis functions quickly and seamlessly with minimum of development cost

• Leverage rapid innovation in Big Data analytics space

• SDO and OSS “partnership” – a collaborative loop

• SDOs to align with properly governed OSS projects

• Proven, neutral third-party management, proper licensing, support infrastructure, public participation

• Linux Foundation, Apache Foundation, OpenStack Foundation, Eclipse Foundation

• Reward the “right type of open” projects – active and open communities

• Projects that produce utility for the industry vs “dead code repository”

• Support projects that compliment the standards development

• Example: OpenDaylight (Linux Foundation) – driving YANG modeling into IETF and other OSS

• Example: OPNFV (Linux Foundation) – NFV Platform, leveraging ETSI NFV architecture & specs

Embrace “Good” Open Source