The Future of Fast Data Processing with fd.io VPP
f
d
.
i
o
i
s
t
h
e
f
u
t
u
r
e
Ed Warnicke
fd.io Foundation
1
T
h
e
F
u
t
u
r
e
o
f
I
O
fd.io Foundation
2
F
a
s
t
D
a
t
a
F
a
s
t
I
n
n
o
v
a
t
i
o
n
C
o
n
v
e
r
g
e
n
c
e
F
d
.
i
o
C
o
r
e
:
V
P
P
–
T
h
e
F
u
t
u
r
e
o
f
I
O
•
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
fd.io Foundation
3
V
P
P
i
n
t
h
e
O
v
e
r
a
l
l
S
t
a
c
k
fd.io Foundation
4
Hardware
Application Layer / App Server
VM/VIM Management Systems
Network Controller
Operating Systems
Data Plane Services
Orchestration
Network IO
V
P
P
F
e
a
t
u
r
e
S
u
m
m
a
r
y
–
F
u
l
l
y
P
r
o
g
r
a
m
m
a
b
l
e
fd.io Foundation
5
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
C
o
u
n
t
e
r
s
f
o
r
e
v
e
r
y
t
h
i
n
g
Mandatory Input Checks:
TTL expiration
header checksum
L2 length < IP length
ARP resolution/snooping
ARP proxy
I
P
v
4
/
I
P
v
6
I
P
v
4
GRE, MPLS-GRE, NSH-GRE,
VXLAN
IPSEC
DHCP client/proxy
CG NAT
I
P
v
6
Neighbor discovery
Router Advertisement
DHCPv6 Proxy
L2TPv3
Segment Routing
MAP/LW46 – IPv4aas
iOAM
M
P
L
S
MPLS-o-Ethernet –
Deep label stacks
supported
L
2
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
F
a
s
t
D
a
t
a
fd.io Foundation
6
•
FD.io
VPP data plane throughput not impacted by large
size of IPv6
FIB
•
VPP tested on
UCS 4-CPU-socket server with 4 of Intel “
Haswell
"
x86
-64
processors
E7-8890
v3
1
8
C 2.
5
GHz
•
24 Cores used – Another 48 cores can be used for other network services!
[
G
b
p
s
]
V
N
E
T
-
S
L
A
B
E
N
C
H
M
A
R
K
I
N
G
A
T
S
C
A
L
E
:
I
P
V
6
V
P
P
-
b
a
s
e
d
v
S
w
i
t
c
h
Phy-VS-Phy
Zero-packet-loss Throughput for 12 port 40GE, 24 cores, IPv6
[
M
p
p
s
]
VPP vSwitch IPv4 routed forwarding
FIB with 2 milion IPv6 entries
12x40GE (480GE) 64B frames
2
0
0
M
p
p
s
z
e
r
o
f
r
a
m
e
l
o
s
s
NIC and PCIe is the limit not VPP
VPP vSwitch IPv4 routed forwarding
FIB with 2 milion IPv6 entries
1
2
x
4
0
G
E
(
4
8
0
G
E
)
I
M
I
X
f
r
a
m
e
s
4
8
0
G
b
p
s
z
e
r
o
f
r
a
m
e
l
o
s
s
“Sky” is the limit not VPP
Zero-packet-loss Throughput for 12 port 40GE, 24 cores, IPv6
V
P
P
C
o
r
e
s
N
o
t
C
o
m
p
l
e
t
e
l
y
B
u
s
y
V
P
P
V
e
c
t
o
r
s
H
a
v
e
S
p
a
c
e
F
o
r
M
o
r
e
S
e
r
v
i
c
e
s
a
n
d
M
o
r
e
P
a
c
k
e
t
s
!
!
P
C
I
e
3
.
0
a
n
d
N
I
C
s
A
r
e
T
h
e
L
i
m
i
t
A
n
d
H
o
w
D
o
W
e
K
n
o
w
T
h
i
s
?
S
i
m
p
l
e
s
–
A
W
e
l
l
E
n
g
i
n
e
e
r
e
d
T
e
l
e
m
e
t
r
y
I
n
L
i
n
u
x
a
n
d
V
P
P
T
e
l
l
s
U
s
S
o
========
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
V
P
P
C
o
r
e
s
N
o
t
C
o
m
p
l
e
t
e
l
y
B
u
s
y
V
P
P
V
e
c
t
o
r
s
H
a
v
e
S
p
a
c
e
F
o
r
M
o
r
e
S
e
r
v
i
c
e
s
a
n
d
M
o
r
e
P
a
c
k
e
t
s
!
!
P
C
I
e
3
.
0
a
n
d
N
I
C
s
A
r
e
T
h
e
L
i
m
i
t
A
n
d
H
o
w
D
o
W
e
K
n
o
w
T
h
i
s
?
S
i
m
p
l
e
s
–
A
W
e
l
l
E
n
g
i
n
e
e
r
e
d
T
e
l
e
m
e
t
r
y
I
n
L
i
n
u
x
a
n
d
V
P
P
T
e
l
l
s
U
s
S
o
========
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
VPP average vector size below shows 23-to-27
This indicates VPP program worker threads are not busy
Busy VPP worker threads should be showing 255
This means that VPP worker threads operate at 10% capacity
It’s like driving 1,000hp car at 100hp power – lots of space
for adding (service) acceleration and (sevice) speed.
V
P
P
C
o
r
e
s
N
o
t
C
o
m
p
l
e
t
e
l
y
B
u
s
y
V
P
P
V
e
c
t
o
r
s
H
a
v
e
S
p
a
c
e
F
o
r
M
o
r
e
S
e
r
v
i
c
e
s
a
n
d
M
o
r
e
P
a
c
k
e
t
s
!
!
P
C
I
e
3
.
0
a
n
d
N
I
C
s
A
r
e
T
h
e
L
i
m
i
t
A
n
d
H
o
w
D
o
W
e
K
n
o
w
T
h
i
s
?
S
i
m
p
l
e
s
–
A
W
e
l
l
E
n
g
i
n
e
e
r
e
d
T
e
l
e
m
e
t
r
y
I
n
L
i
n
u
x
a
n
d
V
P
P
T
e
l
l
s
U
s
S
o
========
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
VPP average vector size below shows 23-to-27
This indicates VPP program worker threads are not busy
Busy VPP worker threads should be showing 255
This means that VPP worker threads operate at 10% capacity
It’s like driving 1,000bhp car at 100bhp power – lots of space
for adding (service) acceleration and (sevice) speed.
VPP is also counting the cycles-per-packet (CPP)
We know exactly what feature, service, packet processing activity is using the CPU cores
We can engineer, we can capacity plan, we can automate service placement
We can scale across many many CPU cores and computers
And AUTOMATE it easily – as it is after all just SOFTWARE
F
a
s
t
I
n
n
o
v
a
t
i
o
n
fd.io Foundation
11
F
a
s
t
I
n
n
o
v
a
t
i
o
n
-
M
o
d
u
l
a
r
i
t
y
E
n
a
b
l
i
n
g
F
l
e
x
i
b
l
e
P
l
u
g
i
n
s
Plugins == Subprojects
Plugins can:
•
Introduce new graph nodes
•
Rearrange packet processing graph
•
Can be built independently of VPP source tree
•
Can be added at runtime (drop into plugin
directory)
•
All in user space
Enabling:
•
Ability to take advantage of diverse hardware
when present
•
Support for multiple processor architectures (x86,
ARM, PPC)
•
Few dependencies on the OS (clib) allowing
easier ports to other Oses/Env
…
C
o
n
v
e
r
g
e
n
c
e
fd.io Foundation
13
F
a
s
t
D
a
t
a
S
c
o
p
e
•
Fast Data Scope:
•
IO
•
Hardware/vHardware <-> cores/threads
•
Processing
•
Classify
•
Transform
•
Prioritize
•
Forward
•
Terminate
•
Management Agents
•
Control/manage IO/Processing
fd.io Foundation
14
IO
Processing
Management Agent
Bare Metal/VM/Container
N
e
x
t
S
t
e
p
s
–
G
e
t
I
n
v
o
l
v
e
d
We invite you to Participate in
fd.io
•
Attend “fd.io Intro” talk Wed 4:10-4:50pm in Grand Ballroom B
•
Attend “fd.io Meetup” Wed 5:00pm in Napa
•
Attend Packet Processed Storage in a Software Defined World Thu 3:15pm in Grand Ballroom F
•
Register for fd.io Training/Hackfest April 4-7, 2016 in Santa Clara
•
Get the Code, Build the Code, Run the Code
•
Read/Watch the Tutorials
•
Join the Mailing Lists
•
Join the IRC Channels
•
Explore the wiki
fd.io Foundation
15
"Explore the future of fast data processing through the innovative fd.io VPP technology. VPP stands as a high-performance packet processing platform running on commodity CPUs. It leverages DPDK for optimal data plane management and boasts fully programmable features like IPv4/IPv6 support, MPLS-GRE, VXLAN, IPSEC, DHCP, and more. With VPP in the overall stack, application layers, server management, network controllers, and data plane services benefit from efficient packet processing capabilities, making it a versatile solution for various networking applications."
Uploaded on Nov 27, 2024 | 0 Views
Download Presentation
Please find below an Image/Link to download the presentation.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.
You are allowed to download the files provided on this website for personal or commercial use, subject to the condition that they are used lawfully. All files are the property of their respective owners.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.
E N D
Presentation Transcript
fd.io is the future fd.io is the future Ed Warnicke fd.io Foundation 1
The Future of IO The Future of IO Fast Data Fast Data Fast Innovation Fast Innovation Convergence Convergence fd.io Foundation 2
Fd.io Core: VPP Fd.io Core: VPP The Future of IO The Future of IO VPP is a rapid packet processing development platform for highly performing network applications. Bare Metal/VM/Container It runs on commodity CPUs and leverages DPDK Data Plane Management Agent It creates a vector of packet indices and processes them using a directed graph of nodes resulting in a highly performant solution. Packet Processing Network IO Runs as a Linux user-space application Ships as part of both embedded & server products, in volume Active development since 2002 fd.io Foundation 3
VPP in the Overall Stack VPP in the Overall Stack Application Layer / App Server VM/VIM Management Systems Orchestration Network Controller Operating Systems Data Plane Services VPP Network IO Packet Processing Hardware fd.io Foundation 4
VPP Feature Summary VPP Feature Summary Fully Programmable IPv4/IPv6 GRE, MPLS-GRE, NSH-GRE, VXLAN IPSEC DHCP client/proxy CG NAT Fully Programmable IPv4 L2 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 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 IPv6 Neighbor discovery Router Advertisement DHCPv6 Proxy L2TPv3 Segment Routing MAP/LW46 IPv4aas iOAM MPLS MPLS-o-Ethernet Deep label stacks supported fd.io Foundation 5
Fast Data Fast Data fd.io Foundation 6
VNET-SLA BENCHMARKING AT SCALE: IPV6 VPP-based vSwitch Zero-packet-loss Throughput for 12 port 40GE, 24 cores, IPv6 Phy-VS-Phy Zero-packet-loss Throughput for 12 port 40GE, 24 cores, IPv6 [Gbps] [Mpps] 500.0 250.0 450.0 400.0 200.0 350.0 300.0 150.0 250.0 100.0 200.0 150.0 50.0 100.0 50.0 1518B 1518B 0.0 0.0 IMIX IMIX 12 routes 12 routes 1k routes 1k routes 64B 100k routes500k routes 64B 100k routes 500k routes 1M routes 1M routes 2M routes 2M routes VPP vSwitch IPv4 routed forwarding FIB with 2 milion IPv6 entries VPP vSwitch IPv4 routed forwarding FIB with 2 milion IPv6 entries 12x40GE (480GE) 64B frames FD.io VPP data plane throughput not impacted by large size of IPv6 FIB VPP tested on UCS 4-CPU-socket server with 4 of Intel Haswell" x86-64 processors E7-8890v3 18C 2.5GHz 12x40GE (480GE) IMIX frames 200Mpps zero frame loss 480Gbps zero frame loss 24 Cores used Another 48 cores can be used for other network services! NIC and PCIe is the limit not VPP Sky is the limit not VPP
VPP Cores Not Completely Busy VPP Vectors Have Space For More Services and More Packets!! PCIe 3.0 and NICs Are The Limit And How Do We Know This? Simples A Well Engineered Telemetry In Linux and VPP Tells Us So ======== TC5 TC5.0 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 TenGigabitEtherneta/0/1-output TenGigabitEtherneta/0/1-tx cop-input dpdk-input ip4-cop-whitelist ip4-input ip4-lookup ip4-rewrite-transit --------------- 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 TenGigabitEthernet88/0/0-outpu TenGigabitEthernet88/0/0-tx cop-input dpdk-input ip4-cop-whitelist ip4-input ip4-lookup ip4-rewrite-transit 120ge.vpp.24t24pc.ip4.cop 120ge.2pnic.6nic.rss2.vpp.24t24pc.ip4.cop State active active active polling active active active active Calls Vectors 211054648 211054648 211054648 211054648 211054648 211054648 211054648 211054648 Suspends Clocks Vectors/Call 9003498 9003498 9003498 45658750 9003498 9003498 9003498 9003498 0 0 0 0 0 0 0 0 1.63e1 7.94e1 2.23e1 1.52e2 4.34e1 4.98e1 6.25e1 3.43e1 23.44 23.44 23.44 4.62 23.44 23.44 23.44 23.44 State active active active polling active active active active Calls Vectors 211055503 211055503 211055503 211055503 211055503 211055503 211055503 211055503 Suspends Clocks Vectors/Call 7805705 7805705 7805705 46628961 7805705 7805705 7805705 7805705 0 0 0 0 0 0 0 0 1.54e1 7.75e1 2.12e1 1.60e2 4.35e1 4.86e1 6.02e1 3.36e1 27.04 27.04 27.04 4.53 27.04 27.04 27.04 27.04
VPP Cores Not Completely Busy VPP Vectors Have Space For More Services and More Packets!! PCIe 3.0 and NICs Are The Limit And How Do We Know This? Simples A Well Engineered Telemetry In Linux and VPP Tells Us So ======== TC5 TC5.0 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 TenGigabitEtherneta/0/1-output TenGigabitEtherneta/0/1-tx cop-input dpdk-input ip4-cop-whitelist ip4-input ip4-lookup ip4-rewrite-transit --------------- 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 TenGigabitEthernet88/0/0-outpu TenGigabitEthernet88/0/0-tx cop-input dpdk-input ip4-cop-whitelist ip4-input ip4-lookup ip4-rewrite-transit VPP average vector size below shows 23-to-27 This indicates VPP program worker threads are not busy Busy VPP worker threads should be showing 255 This means that VPP worker threads operate at 10% capacity 120ge.vpp.24t24pc.ip4.cop 120ge.2pnic.6nic.rss2.vpp.24t24pc.ip4.cop It s like driving 1,000hp car at 100hp power lots of space for adding (service) acceleration and (sevice) speed. State active active active polling active active active active Calls Vectors 211054648 211054648 211054648 211054648 211054648 211054648 211054648 211054648 Suspends Clocks Vectors/Call 9003498 9003498 9003498 45658750 9003498 9003498 9003498 9003498 0 0 0 0 0 0 0 0 1.63e1 7.94e1 2.23e1 1.52e2 4.34e1 4.98e1 6.25e1 3.43e1 23.44 23.44 23.44 4.62 23.44 23.44 23.44 23.44 State active active active polling active active active active Calls Vectors 211055503 211055503 211055503 211055503 211055503 211055503 211055503 211055503 Suspends Clocks Vectors/Call 7805705 7805705 7805705 46628961 7805705 7805705 7805705 7805705 0 0 0 0 0 0 0 0 1.54e1 7.75e1 2.12e1 1.60e2 4.35e1 4.86e1 6.02e1 3.36e1 27.04 27.04 27.04 4.53 27.04 27.04 27.04 27.04
VPP Cores Not Completely Busy VPP Vectors Have Space For More Services and More Packets!! PCIe 3.0 and NICs Are The Limit And How Do We Know This? Simples A Well Engineered Telemetry In Linux and VPP Tells Us So ======== TC5 TC5.0 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 TenGigabitEtherneta/0/1-output TenGigabitEtherneta/0/1-tx cop-input dpdk-input ip4-cop-whitelist ip4-input ip4-lookup ip4-rewrite-transit --------------- 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 TenGigabitEthernet88/0/0-outpu TenGigabitEthernet88/0/0-tx cop-input dpdk-input ip4-cop-whitelist ip4-input ip4-lookup ip4-rewrite-transit VPP average vector size below shows 23-to-27 This indicates VPP program worker threads are not busy Busy VPP worker threads should be showing 255 This means that VPP worker threads operate at 10% capacity 120ge.vpp.24t24pc.ip4.cop 120ge.2pnic.6nic.rss2.vpp.24t24pc.ip4.cop It s like driving 1,000bhp car at 100bhp power lots of space for adding (service) acceleration and (sevice) speed. State active active active polling active active active active We can scale across many many CPU cores and computers And AUTOMATE it easily as it is after all just SOFTWARE Calls Vectors 211054648 211054648 211054648 211054648 211054648 211054648 211054648 211054648 Suspends Clocks Vectors/Call 9003498 9003498 9003498 45658750 9003498 9003498 9003498 9003498 0 0 0 0 0 0 0 0 1.63e1 7.94e1 2.23e1 1.52e2 4.34e1 4.98e1 6.25e1 3.43e1 23.44 23.44 23.44 4.62 23.44 23.44 23.44 23.44 VPP is also counting the cycles-per-packet (CPP) We know exactly what feature, service, packet processing activity is using the CPU cores We can engineer, we can capacity plan, we can automate service placement State active active active polling active active active active Calls Vectors 211055503 211055503 211055503 211055503 211055503 211055503 211055503 211055503 Suspends Clocks Vectors/Call 7805705 7805705 7805705 46628961 7805705 7805705 7805705 7805705 0 0 0 0 0 0 0 0 1.54e1 7.75e1 2.12e1 1.60e2 4.35e1 4.86e1 6.02e1 3.36e1 27.04 27.04 27.04 4.53 27.04 27.04 27.04 27.04
Fast Innovation Fast Innovation fd.io Foundation 11
Fast Innovation - Modularity Enabling Flexible Plugins Plugins == Subprojects Plugins can: Introduce new graph nodes Rearrange packet processing graph Can be built independently of VPP source tree Can be added at runtime (drop into plugin directory) All in user space Enabling: Ability to take advantage of diverse hardware when present Support for multiple processor architectures (x86, ARM, PPC) Few dependencies on the OS (clib) allowing easier ports to other Oses/Env Packet vector ethernet-input Plug-in to enable new HW input Nodes mpls-ethernet-input ip4input llc-input ip6-input arp-input ip6-lookup Plug-in to create new nodes ip6-rewrite-transmit ip6-local Custom-B Custom-A
Convergence Convergence fd.io Foundation 13
Fast Data Scope Fast Data Scope Bare Metal/VM/Container Fast Data Scope: IO Processing Classify Transform Prioritize Forward Terminate Management Agents Control/manage IO/Processing Hardware/vHardware <-> cores/threads Management Agent Processing IO fd.io Foundation 14
Next Steps Next Steps Get Involved We invite you to Participate in fd.io Get Involved Attend fd.io Intro talk Wed 4:10-4:50pm in Grand Ballroom B Attend fd.io Meetup Wed 5:00pm in Napa Attend Packet Processed Storage in a Software Defined World Thu 3:15pm in Grand Ballroom F Register for fd.io Training/Hackfest April 4-7, 2016 in Santa Clara Get the Code, Build the Code, Run the Code Read/Watch the Tutorials Join the Mailing Lists Join the IRC Channels Explore the wiki fd.io Foundation 15
