SDNs: hot topics, evolution & research opportunities

1. SDNs: hot topics, evolution
& research opportunities
Diego Kreutz
kreutz@lasige.di.fc.ul.pt
LASIGE - Large-Scale Informatics Systems Laboratory
FCUL - Faculty of Sciences of University of Lisbon

2. 2!NavTalks, October, 2012!
Roadmap
1. HotSDN & HotNets hot topics
2. SDNs: evolution & improvements
3. On SDN controllers
4. Final notes & pointers

3. 3!NavTalks, October, 2012!
Roadmap
1. HotSDN & HotNets hot topics
2. SDNs: evolution & improvements
3. On SDN controllers
4. Final notes & pointers

4. 4!NavTalks, October, 2012!
HotSDN & HotNets 2012
(hot topics & open research/engineering)
Ø Troble Shooting and Debugging
Ø Delivering assured SDNs
§ Make sure they work correctly
Ø Look more at operators practices
Ø Think more about OpenFlow-like protocols
Ø High level abstractions to put intelligence
into the network
Ø OpenFlow-based WANs are becoming a
reality

5. 5!NavTalks, October, 2012!
HotSDN & HotNets 2012
(hot topics & open research/engineering)
Ø Seriously engineering (e.g. performance)
Ø Improve SDNs with things from security and
reliability areas
Ø Reduce the gap between research and
practical stuff
Ø Think again (“call to arms”)! SDNs are not
yet there.
§ We are only on chapter 1. There are at least 9
more chapters to go.

6. 6!NavTalks, October, 2012!
HotSDN & HotNets 2012
(examples of papers/ideas)
Paper: Where is the Debugger for my Software-Defined
Network?
Ø Faced problem: debug the network while
developing an in-network load balancer
§ Bug 1: no match rule in the middle of the network
§ Bug 2: servers at wrong location
§ Bug 3: servers that do not connect to clients
Ø Solution: ndb (similar to gdb)
§ Allow online debugging
§ Use small postcards to reconstruct and debug
§ First primitives: breakpoint and backtrace

7. 7!NavTalks, October, 2012!
HotSDN & HotNets 2012
(examples of papers/ideas)
Paper: Hey, You Darned Counters! Get Off My ASIC!
Ø Faced problem: counters are a big issue in
hardware design and development
§ It is not easy and cheap to put specific counters inside
ASICs
§ OpenFlow has many counters (per flow/rule: received
packets, received bytes, duration, …)
Ø Solution: Software-Defined Counters (SDCs)
§ Use off-the-shelf CPUs
§ Let the software/CPUs deal with the counters
§ End up with: flexibility, scalability and less hw costs

8. 8!NavTalks, October, 2012!
HotSDN & HotNets 2012
(examples of papers/ideas)
Paper: Programming Your Network at Run-time for Big Data
Applications
Ø Problem: waste of networking resources
§ Applications are not optimized for the network
§ The network is not optimized for the applications
Ø Solution: NaaS for specific applications (e.g.
Hadoop) inside datacenters
§ Goal: optimize app performance and net utilization
§ Applications can request resources
§ Applications can drive the network use
§ Routing is done based on each app’s requirements

9. 9!NavTalks, October, 2012!
HotSDN & HotNets 2012
(examples of papers/ideas)
Paper: Towards Programmable Enterprise WLANs with Odin
Ø Problem: association decisions are made by
clients and not the infrastructure
§ Handoffs are expensive
§ Things such as load distribution are yet not feasible
Ø Solution: light virtual APs (LVAP) with Odin
§ Framework do empower network operators
v No changes on 802.11 standard
v One LVAP abstraction per client (unique BSSID)
v For the app, a user is attached to an “Ethernet port”
§ Applications: seamless mobility, load balancing, …
v Re-association is done by the management software

10. 10!NavTalks, October, 2012!
HotSDN & HotNets 2012
(towards future Internet architectures)
Papers:
Ø Fabric: A Retrospective on Evolving SDN
§ Borrowing ideas from MPLS (simple cores/data planes)
§ Separation: fabric and edge elements
§ Two OpenFlows? One for edge and one for core/fabric?
Ø Software-Defined Internet Architecture: Decoupling
Architecture from Infrastructure
§ Key elements: MPLS, SDNs, Middleboxes, Software Forwarding
§ Core routers/backbones: any technology (Fabric-like)
§ Edge routers: Software-Defined Routers (SDRs?)
v Software-based forwarding and control (e.g. OpenFlow)

11. 11!NavTalks, October, 2012!
Roadmap
1. HotSDN and HotNets hot topics
2. SDNs: evolution & improvements
3. On SDN controllers
4. Final notes & pointers

12. 12!NavTalks, October, 2012!
SDNs: evolution
(from LANs to end user NaaS)
1. LANs: Ethane - Taking Control of the Enterprise
2. WANs (Wi-Fi): Towards Programmable Enterprise WLANs With Odin
3. WANs: Inter-Datacenter WAN with centralized TE using SDN and OpenFlow
4. Backbones: Fabric: A Retrospective on Evolving SDN
5. Internet architecture: Software-Defined Internet Architecture:
Decoupling Architecture from Infrastructure
6. Cellular Networks: Toward Software-Defined Cellular Networks
7. xDSL (and related) Networks: ???
8. NaaS (LAN): Programming Your Network at Run-time for Big Data
Applications
9. NaaS (WAN/LAN): ???

13. 13!NavTalks, October, 2012!
SDNs: evolution

14. 14!NavTalks 2012!
SDNs: evolution
(on future NaaS to end users)
MPLS GMPLS
Optical(L1)
SaaS
Users
PaaS
Users
IaaS
Users
SaaS
Provider
IaaS
Provider
Content
Provider
PaaS
Provider
Wireless
Broadband

15. 15!NavTalks 2012!
SDNs: improvements
(scalability: from 30K to 1M flows/s)
Challenge: 100 edge switches = 10M flows/s
Ø Original NOX: 30K flows/s
Ø Maestro: 300K & 600K flows/s
§ Scales linearly (?) with the number of machines
v 10M / 300K = 34 Maestro instances
Ø NOX-MT: 1M flows/s
§ 1.8M flows/s for 32/64 edge switches (105 MACs/switch)
§ 1.0M flows/s for 256 edge switches
v 10M / 1M = 10 NOX-MT instances
Ø Kandoo: ? flows/s
§ Two-layers of controllers (local and root)
§ Can easily scale to N (?) on-demand edge switches

16. 16!NavTalks 2012!
SDNs: improvements
(scalability: from 30K to 1M flows/s)
Challenge: 100 edge switches = 10M flows/s
Ø Original NOX: 30K flows/s
Ø Maestro: 300K & 600K flows/s
§ Scales linearly (?) with the number of machines
v 10M / 300K = 34 Maestro instances
Ø NOX-MT: 1M flows/s
§ 1.8M flows/s for 32/64 edge switches (105 MACs/switch)
§ 1.0M flows/s for 256 edge switches
v 10M / 1M = 10 NOX-MT instances
Ø Kandoo: ? flows/s
§ Two-layers of controllers (local and root)
§ Can easily scale to N (?) on-demand edge switches

17. 17!NavTalks 2012!
SDNs: improvements
(scalability: fighting with the counters)
Ø DevoFlow
§ Reduce control traffic (updates of counters)
Ø Software-Defined Counters (SDCs)
§ Solve the counters’ ASIC design problem
Ø DIFANE
§ Reduce control overhead (counters)
Ø Software-Defined Cell Networks (SDCNs?)
§ Reduce control overhead (counters)
§ Use specific TCAMs/tables for DPI
Ø Kandoo
§ Scalable controllers (reduce the monitoring overhead)

18. 18!NavTalks 2012!
SDNs: improvements
(scalability: fighting with the counters)
Ø DevoFlow
§ Reduce control traffic (updates of counters)
Ø Software-Defined Counters (SDCs)
§ Solve the counters’ ASIC design problem
Ø DIFANE
§ Reduce control overhead (counters)
Ø Software-Defined Cell Networks (SDCNs?)
§ Reduce control overhead (counters)
§ Use specific TCAMs/tables for DPI
Ø Kandoo
§ Scalable controllers (reduce the monitoring overhead)

19. 19!NavTalks 2012!
Roadmap
1. HotSDN and HotNets hot topics
2. SDNs: evolution & improvements
3. On SDN controllers
4. Final notes & pointers

20. 20!NavTalks 2012!
On SDN controllers
(the OpenFlow example)
… … ……
Data plane tier
Controller tier
Application tier
Controller
(Server Software)Controller
(Server Software)

21. 21!NavTalks 2012!
On SDN controllers
(software is now king: take care of it)
… … ……
Controller
(Server Software)Controller
(Server Software)
Challenge:
modular,
updatable and
secure software
Challenges:
modular, hot-
updatable and
secure software
Challenges:
modular, hot-
updatable,
secure and
reliable software

22. 22!NavTalks 2012!
On SDN controllers
(inside a controller)
Raw-
Packet
Task
Queue
Input
Stage
Output
Stage
Discovery
Intradomain
Routing
Routing
Table
Authenti-
cation
Route
Flow
Periodic
Timer Discovery
Updates
Topology
Changes
LLDP
Packets
Flow
Requests
Flow
Requests
Config
Msgs
LLDP
PacketsTriggers
Controller (Server Software)

23. 23!NavTalks 2012!
On SDN controllers
(simple app/policies using NOX with FML)
FML (Flow-based Management Language)
Basic format:
allow(Us, Hs, As, Ut, Ht, At, Prot, Req) <= …
Policy P3
allow(Flow) <= Prot = arp
allow(Flow) <= Prot = dhcp
allow(Flow) <= Ht = auth_server ^ Prot = http
allow(Flow) <= Hs = auth_server ^ Prot = http
Policy P2
httpRedirect(Flow, 307, auth_server) <= Us = unknown ^ Prot = http
Policy P1
deny(Flow) <= Us = unknown
Deny all flows
from non-
authenticated
users

24. 24!NavTalks 2012!
On SDN controllers
SDN
Controller
Interdomain
routing
Quagga =>
RouteFlow

25. 25!NavTalks 2012!
On SDN controllers
(challenges)
Ø Secure & Resilient Controllers
Ø Examples of open issues
§ Distributed controllers (e.g. Onix, HyperFlow)
v Only weak data consistency
v Only crash faults are tolerated
§ Centralized controllers (e.g. NOX, Foodlight,
Beacon, Maestro)
v Single points of failure
§ Common issues (lack of)
v Assured strong data and operation integrity
v Support for tolerating Byzantine behavior

26. 26!NavTalks 2012!
Ø Redundancy
§ e.g. controller clusters (e.g. Onix)
Ø Efficient fail-over mechanisms
Ø Distribution taking into consideration:
§ Strong data consistency mechanisms
§ Techniques for strong mutual authentication,
integrity and confidentiality
Ø Fault and Intrusion Tolerant (FIT)
Controllers
Ø …
On SDN controllers
(things to do/evaluate/improve)

27. 27!NavTalks 2012!
Roadmap
1. HotSDN and HotNets hot topics
2. SDNs: evolution & improvements
3. On SDN controllers
4. Final notes & pointers

28. 28!NavTalks 2012!
Controllers out there
(ready to use tools)
Name Lang Platforms License Link
NOX C++/Python Linux GPL http://www.noxrepo.org/
Floodlight Java Any Apache http://floodlight.openflowhub.org/
Maestro Java
Win,
Linux, Mac
LGPL
http://code.google.com/p/maestro-
platform/
Beacon Java
Linux,
Mac, Win
GPL &
FOSS
http://openflow.stanford.edu/
display/Beacon
Trema C/Ruby Linux GPL2 http://trema.github.com/trema/
NodeFlow JavaScript http://garyberger.net/?p=537
RouteFlow Linux Apache
https://sites.google.com/site/
routeflow/
FlowVisor Java
https://openflow.stanford.edu/
display/DOCS/Flowvisor
See more examples at http://yuba.stanford.edu/~casado/of-sw.html

29. 29!NavTalks 2012!
Ø Flexibility is the main turn key in SDNs
Ø Scalability/performance issues are solvable
§ Ongoing and future engineering & research tasks
Ø Software is the king!
§ Going back to software-based forwarding
§ Network virtualization is going wild
Ø Keep it simple! Decoupling infra from arch.
§ Very simple and efficient data planes (ASICs, …)
§ Very simple abstractions (designs & programming)
§ Advanced control planes & software apps
Final Notes
(try to keep in mind)

30. 30!NavTalks 2012!
Ø ONRC Reseach (Stanford, Berkeley, ON.LAB)
§ http://onrc.stanford.edu/research.html
Ø SDNCentral
§ http://www.sdncentral.com/
Ø Conferences & Workshops
§ Hot-Nets, Hot-SDN, Hot-ICE, …
§ SIGCOMM, Usenix NSDI, …
Ø Some tech stuff
§ OpenFlow: http://www.openflow.org/
§ OpenvSwitch: http://openvswitch.org/
§ …
Some pointers
(cool projects, publications & tech stuff)