Difference between revisions of "VPP Usability Track"

From fd.io
Jump to: navigation, search
(Created page with "Creating VPP Usability Track")
 
Line 1: Line 1:
−
Creating VPP Usability Track
+
==<font color=red>WORK-IN-PROGRESS - This is a planning page for addressing VPP usability aspects</font>==
 +
 
 +
==Goals==
 +
# VPP working out of the box for most/all baseline use cases.
 +
# Profiled by target VPP consumers - use cases, users.
 +
#* Use1 - OPNFV/FDS
 +
#* Use2 - Programmable Virtual Forwarder
 +
#* Use3 - OpenStack NFVI
 +
#* Use4 - vSwitch for VMs
 +
 
 +
==VPP Use Case Requirements==
 +
 
 +
Functional requirements of target VPP consumers - wip - current snapshot:
 +
# Use1 - OPNFV/FDS - VXLAN+L2BD+vhost, VLAN+L2BD+vhost, BVI, VRF, IPv4, IPv6, SNAT, ACL/classifier
 +
# Use2 - Programmable Virtual Forwarder - VXLAN+L2BD+vhost, IPv4, IPv6, more TBC
 +
# Use3 - OpenStack NFVI - VXLAN+L2BD+vhost
 +
# Use4 - vSwitch for VMs - VLAN+L2BD+vhost, high-density VNF VMs (30VMs, 102vhost)
 +
 
 +
==User guide==
 +
Outline of user guide that needs to be produced:
 +
 
 +
# Installation per Linux distro - Ubuntu, Centos, RHEL, other - Ed, Sean
 +
#* Packaging
 +
#** https://wiki.fd.io/view/VPP/Installing_VPP_binaries_from_packages#CentOS_7.2_-_VPP_master_branch_RPMs_.28in_development.29
 +
#** sudo yum install vpp-python-api
 +
#* Manual install - ??
 +
#* Automated install - Ed, Sean
 +
#** see outline in email Ed, Sean "Re: 161026-vpp-usability-review - note", 27 October 2016 at 22:39:47 BST
 +
# Environment - MK, FB
 +
#* <complete the list of areas and combinations>
 +
#* Linux environment fundamentals
 +
#* SW Dependencies
 +
#** Linux kernel ver
 +
#** QEMU ver
 +
#** DPDK ver
 +
#* HW Dependencies
 +
#** x86_64 microarchitectures
 +
#** NICs
 +
# Initial configuration - MK, PM, DM for questions
 +
#* Startup configuration - startup.conf ...
 +
#* Interfaces
 +
# Optimizing VPP performance - MK, PM, PF&DM for questions
 +
#* Tuning performance
 +
#* VM and vhost-user considerations
 +
#* Useful host performance telemetry
 +
#* Useful VPP performance telemetry
 +
# Sample use cases - Chris Metz with team will work it
 +
#* L2 switching
 +
#** with VMs (vhost)
 +
#** with tunnels (vxlan, lisp-gpe)
 +
#** with security-groups
 +
#* IP routed forwarding
 +
#** with VMs (vhost)
 +
#** with tunnels (vxlan, lisp-gpe)
 +
#** with security-filters - iacl, cop-whitelist, cop-blacklist
 +
#* <add more>
 +
#* <structure differently?>
 +
# Doc Generation and Online Presentation
 +
#* Consider [http://readthedocs.org readthedocs] approach
 +
#** Based on [http://www.sphinx-doc.org/ sphinx] using [http://github.com/snide/sphinx_rtd_theme readthedocs theme]
 +
#** Used today by [http://www.dpdk.org/doc dpdk doc]
 +
 
 +
==Programmer's Guide==
 +
 
 +
Programmer's guide - KRB for TOC, OT to help
 +
# VPP API guide
 +
 
 +
==VPP Performance Considerations==
 +
 
 +
Here initial points to be addressed for optimizing VPP performance on specific compute HW configurations:
 +
# cpu core configuration and vpp thread mappings
 +
#* phy interfaces can be placed to thread/core
 +
#* vhost interfaces are round-robined - new feature for placement
 +
#** do after multi-queue patch by Pierre
 +
# vhost - qsz, cpu jitter, reconnect, interop qemu-virtio
 +
#* hard to install, doesn't work all the time, crashes
 +
# dpdk - performance with selection of NICs
 +
#* more detailed documentation about baremetal installations
 +
#** vs. just Vagrant
 +
#* dpdk new rls setup recommendations - per NIC basis
 +
#* dpdk.org is not publishing performance numbers
 +
#* csit can't address it - need manual tests and analysis
 +
# vpp self-diagnostics for optimal setup verification
 +
 
 +
==Minimal setup installation for vhost+VM connectivity==
 +
 
 +
Minimal setup installation for vhost+VM connectivity:
 +
# TRex
 +
# ansible scripts developed in CSIT
 +
# Consumability of CSIT-perf RF and python libraries
 +
 
 +
==Important test areas - MK==
 +
 
 +
Listing of important areas to increase VPP test coverage:
 +
# negative tests - add/remove interfaces/routes/MAC entries
 +
# box-full tests NIC-NIC
 +
# box-full tests NIC-VM-NIC
 +
# stress-tests - add/remove VMs
 +
# negative weird setups - mixed L2BD, IRB, IPv4, IPv6 forwarding
 +
# soak-tests
 +
# tap devices - add/remove
 +
# negative stress/weird tests
 +
 
 +
==VPP Diagnostics and telemetry==
 +
 
 +
Following VPP diagnostics and telemetry aspects need to be addressed:
 +
# live health and performance metrics
 +
#* compute HW
 +
#* Linux kernel
 +
#* VM guest
 +
# collectd + influxdb + grafana
 +
#* VPP interface counters
 +
#* VPP vector size

Revision as of 21:41, 6 November 2016

WORK-IN-PROGRESS - This is a planning page for addressing VPP usability aspects

Goals

  1. VPP working out of the box for most/all baseline use cases.
  2. Profiled by target VPP consumers - use cases, users.
    • Use1 - OPNFV/FDS
    • Use2 - Programmable Virtual Forwarder
    • Use3 - OpenStack NFVI
    • Use4 - vSwitch for VMs

VPP Use Case Requirements

Functional requirements of target VPP consumers - wip - current snapshot:

  1. Use1 - OPNFV/FDS - VXLAN+L2BD+vhost, VLAN+L2BD+vhost, BVI, VRF, IPv4, IPv6, SNAT, ACL/classifier
  2. Use2 - Programmable Virtual Forwarder - VXLAN+L2BD+vhost, IPv4, IPv6, more TBC
  3. Use3 - OpenStack NFVI - VXLAN+L2BD+vhost
  4. Use4 - vSwitch for VMs - VLAN+L2BD+vhost, high-density VNF VMs (30VMs, 102vhost)

User guide

Outline of user guide that needs to be produced:

  1. Installation per Linux distro - Ubuntu, Centos, RHEL, other - Ed, Sean
  2. Environment - MK, FB
    • <complete the list of areas and combinations>
    • Linux environment fundamentals
    • SW Dependencies
      • Linux kernel ver
      • QEMU ver
      • DPDK ver
    • HW Dependencies
      • x86_64 microarchitectures
      • NICs
  3. Initial configuration - MK, PM, DM for questions
    • Startup configuration - startup.conf ...
    • Interfaces
  4. Optimizing VPP performance - MK, PM, PF&DM for questions
    • Tuning performance
    • VM and vhost-user considerations
    • Useful host performance telemetry
    • Useful VPP performance telemetry
  5. Sample use cases - Chris Metz with team will work it
    • L2 switching
      • with VMs (vhost)
      • with tunnels (vxlan, lisp-gpe)
      • with security-groups
    • IP routed forwarding
      • with VMs (vhost)
      • with tunnels (vxlan, lisp-gpe)
      • with security-filters - iacl, cop-whitelist, cop-blacklist
    • <add more>
    • <structure differently?>
  6. Doc Generation and Online Presentation

Programmer's Guide

Programmer's guide - KRB for TOC, OT to help

  1. VPP API guide

VPP Performance Considerations

Here initial points to be addressed for optimizing VPP performance on specific compute HW configurations:

  1. cpu core configuration and vpp thread mappings
    • phy interfaces can be placed to thread/core
    • vhost interfaces are round-robined - new feature for placement
      • do after multi-queue patch by Pierre
  2. vhost - qsz, cpu jitter, reconnect, interop qemu-virtio
    • hard to install, doesn't work all the time, crashes
  3. dpdk - performance with selection of NICs
    • more detailed documentation about baremetal installations
      • vs. just Vagrant
    • dpdk new rls setup recommendations - per NIC basis
    • dpdk.org is not publishing performance numbers
    • csit can't address it - need manual tests and analysis
  4. vpp self-diagnostics for optimal setup verification

Minimal setup installation for vhost+VM connectivity

Minimal setup installation for vhost+VM connectivity:

  1. TRex
  2. ansible scripts developed in CSIT
  3. Consumability of CSIT-perf RF and python libraries

Important test areas - MK

Listing of important areas to increase VPP test coverage:

  1. negative tests - add/remove interfaces/routes/MAC entries
  2. box-full tests NIC-NIC
  3. box-full tests NIC-VM-NIC
  4. stress-tests - add/remove VMs
  5. negative weird setups - mixed L2BD, IRB, IPv4, IPv6 forwarding
  6. soak-tests
  7. tap devices - add/remove
  8. negative stress/weird tests

VPP Diagnostics and telemetry

Following VPP diagnostics and telemetry aspects need to be addressed:

  1. live health and performance metrics
    • compute HW
    • Linux kernel
    • VM guest
  2. collectd + influxdb + grafana
    • VPP interface counters
    • VPP vector size
Retrieved from "https://wiki.fd.io/index.php?title=VPP_Usability_Track&oldid=3705"