Cloud services are mid-way migrating from the traditional topic of unit VM/container/app migration to management of populations of VMs/containers/apps. Cloud Providers (CPs) today do not yet cooperate with Service Providers (SPs) by providing information on local performance and/or local execution context. On the other hand, CPs attempting to optimize performance for all their services would run into an acute complexity problem, where this presentation makes this case by discussing the Virtual Network Embedding (VNE) problem scaled to many concurrent services. One way to resolve the complexity problem is to allow services to manage their own populations, in a Do-It-Yourself (DiY) manner. The recently proposed Cloud Probing technology [1] does just that -- services actively probe e2e network performance of the cloud and optimize themselves. Another recent proposal in [2] discusses the next step -- the Local Hardware Awareness (LHA) technology through which CPs would cooperate with SPs in local discovery and would therefore improve DiY decisions. Among several examples, this presentation will discuss a cloud-based CDN [3] where SPs manage populations while clients aggregate content from several sources via concurrent streams -- referred to as the substream method in the literature on P2P streaming. This author realizes that the LHA technology breaks a fundamental rule in clouds -- blackboxing, but hopes that the case delivered in this presentation is sufficiently appealing to warrant a standardization effort in this direction.

2. .
Questions and Objectives
• (1a) How to steer the global network behavior resulting from
self-organization where agents have their own goals and actions? -- global
vs local/self optimization
• (1b) How to build and leverage the knowledge produced at the micro (agent) and
macro (network) levels? How to correlate and make sense out of
measurements and data from heterogeneous sources? -- e2e
performance heterogeneity
• (2) How to make the operations scale? To which extent are the new
techniques and solutions actionable, "computable"...? -- complexity problem
• (3) How can the management techniques leverage on the new properties and
features introduced by adaptive and cooperative agents, i.e. applying the
same techniques to the management plane itself. -- heterogeneity
problem
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 2/25
...
2/25

3. .
This Talk is about Cloud Networks
• a network of DCs is a cloud -- say, Amazon EC2 with 9 regions
• a network of clouds is a federation -- say, Amazon x 1dc + IDCF x
20dcs in Japan
◦ we already start hitting the heterogeneity wall, since there is no solution yet, no
such federations exist today, but some are planned 03
• .. moving further to network edge, we find Akamai with its 30k+ cloudlets
02 -- a class of its own
◦ heterogeneity problem? in hardware NO, but in e2e network performance STRONG YES
◦ but, Akamai has all the control, so, they manage to balance the load properly
• ... now, having incorporated all the devices at network edge you have the
fog cloud
• ... on a side note... why cloud networks? ... well, because most networking
soon will be done inside and/or by clouds
03 "Chameleon project" https://www.chameleoncloud.org (2014)
02 B.Frank+8 "Pushing CDN-ISP Collaboration to the Limit" ACM SIGCOMM Computer Communication Review (2013)
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 3/25
...
3/25

4. .
Cloudification Timeline and Scope
WiFi
Scale/sizeofcloudifieditems
Timeline
Mid-2015
Probably
should not
cloudify
Should
definitely
cloudify
Data
Centers
Racks IoV > VC
Desktop
APs,
Beacons
Smartphone
Sensor
IoT
Delegated
networking
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 4/25
...
4/25

5. .
One Cloud Service
.
VNE Problem
..
....stands for Virtual Network Embedding
APP
Cloud/DC
APP
APP
VM
Container
APP
Cloud/DC
APP
APP
…
• service can provide the
graph of its apps
• graph can include end users in
form of DC-user e2e
network performance
• federated cloud known its
actual topology and can use
VNE to map 04
04 M.Zhanikeev "A New VNE Method for More Responsive Networking in Many-to-Many Groups" 7th ICUFN (2015)
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 5/25
...
5/25

6. .
Many Cloud Services
.
VNE
..
....is the same, no difference
Cloud A Cloud B
DC
DC
DC
Virtual
Physical
Many
Service
topology
• for many services, you might
use a more detailed
physical topology
(rack-level?)
• otherwise, the same mapping
method
• many-service case is already
considered in literature 04
04 M.Zhanikeev "A New VNE Method for More Responsive Networking in Many-to-Many Groups" 7th ICUFN (2015)
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 6/25
...
6/25

7. .
OSPF, QoS Routing, VNE → complexity
• OSPF optimizes one parameter (cost/weight) for one physical graph, and is NP
Hard
• QoS Routing optimizes multiple parameters for one path, and is NP Hard
• VNE maps one virtual graph on top of a physical graph and is NP Hard-er?
• ... so, what happens when the global NOC tries to optimize many VNEs? NP
Hard-er-er-er?
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 7/25
...
7/25

8. .
Solution to the Complexity Problem
.
The Simple Solution is...
..
.... Do-it-Yourself, that is, to let services self-optimize
• however....
◦ it is one thing to self optimize in a non-cooperative environment
◦ ... and quite another when clouds offer tools/envs/platforms that
help self-optimize better
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 8/25
...
8/25

9. .
Non-Cooperative Way
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 9/25
...
9/25

10. .
Step 1: Probe your Cloud
• visualization of active probing
across 8 Amazon regions 05
california
ireland
oregon
saopaulo
singapore
sydney
tokyo virginia
same DC
large bulk transfer
california
ireland
oregon
saopaulo
singapore
sydney
virginia
DC - other DC
e2e delay
05 M.Zhanikeev "Performance Management of Cloud Populations via Cloud Probing" IPSJ JIP (related to http://goo.gl/f3QpiW) (2015)
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 10/25
...
10/25

11. .
Step 2: Optimize your service network
• change DCs for some members of your population, gradually optimize 05
Pooler model
0 100 200 300 400 500
Ordered list of values
0
20
40
60
80
100
Completiontime(s)
Do nothing
Optimize
05 M.Zhanikeev "Performance Management of Cloud Populations via Cloud Probing" IPSJ JIP (related to http://goo.gl/f3QpiW) (2015)
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 11/25
...
11/25

12. .
Cooperative Clouds
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 12/25
...
12/25

13. .
Traditional vs Akamai vs LHAP
Local
hardware
VirtualMachine
Cloud
Platform
(Xen)
App
App
App
…Multitenant
environment
Physical Machine (PM)
Tools
API
Layer
Control
Layer
Local
Hardware
Awareness
Platform
(LHAP)
…
VirtualMachine
App
Tools
…
VirtualMachine
App
VirtualMachine
App
App
App
…
Global network
(to Internet)
In-box
network
Fig1a: Conventional Platform
…
MiniOS
App
SSD
Operating
System
VirtualMachine
App
App
App
…
Multitenancy
impossible
Physical Machine (PM)
Fig1b: Mini Cache
Cloud
Platform
(Xen)
Operating
System
Global network
(to Internet)
HDD
Local
Network
Global network
(to Internet)
Physical Machine (PM)
Fig1c: LHAP (proposed)
Cloud
Platform
(Xen)
Operating
System
DtL:
Direct to Local
Local
hardware
SSD
HDD
02 B.Frank+8 "Pushing CDN-ISP Collaboration to the Limit" ACM SIGCOMM Computer Communication Review (2013)
06 M.Zhanikeev "A Cloud Visitation Platform to Facilitate Cloud Federation and Fog Computing" IEEE Computer (2015)
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 13/25
...
13/25

14. .
Akamai MiniCache Technology
Local
hardware
Cloud
Platform
(Xen)
Multitenant
environment
Physical Machine (PM)
…
VirtualMachine
App
VirtualMachineApp
App
App
…
Global network
(to Internet)
Fig1a: Conventional Platform
…
MiniOS
App
SSD
Operating
System
VirtualMachine
App
App
App
…
Multitenancy
impossible
Physical Machine (PM)
Fig1b: Mini Cache
Cloud
Platform
(Xen)
Operating
System
Global network
(to Internet)
HDD
• depends on MiniOS, part of Xen
• MiniCache allows MiniOS-based apps
to use SSD/HDD connected to that PM
• MiniCache (storage) and ClickOS (network)
are similar technologies, the letter is used
for SDN
• demerits
1. restricted to PM
2. storage only
3. no concurrency control or load balancing
02 B.Frank+8 "Pushing CDN-ISP Collaboration to the Limit" ACM SIGCOMM Computer Communication Review (2013)
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 14/25
...
14/25

15. .
Local Hardware Awareness (LHA)
VirtualMachine
App
App
App
…
Tools
API
Layer
Control
Layer
Local
Hardware
Awareness
Platform
(LHAP)
…
VirtualMachine
App
Tools
In-box
network
Local
Network
Global network
(to Internet)
Physical Machine (PM)
Fig1c: LHAP (proposed)
Cloud
Platform
(Xen)
Operating
System
DtL:
Direct to Local
Local
hardware
SSD
HDD
• not restricted to resource type, any
resource
◦ storage, sensors, Hadoop/MapReduce jobs, active
probing, ...
• not resricted to same PM, API Layer knows
the resources offered for discovery
• Tools in VMs (and containers) interface with
LHAP
• LHAP is optional -- legacy VMs will not know
it's there
06 M.Zhanikeev "A Cloud Visitation Platform to Facilitate Cloud Federation and Fog Computing" IEEE Computer (2015)
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 15/25
...
15/25

16. .
Example Application : CDN
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 16/25
...
16/25

17. .
CDN on Federated Clouds
Large
Cloud
Small
Cloud
Small
Cloud
CDN
Manager
CDN
Local clients
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 17/25
...
17/25

18. .
Cloud Streaming in the Wild
VM
population
Current
Sources
Service
Provider
(SP)
Tracker
Service
Provider
(SP)
Parent
peers
P2P
streaming
Cloud
streaming
ClientClient
07 M.Zhanikeev "Multi-Source Stream Aggregation in the Cloud" ISBN 978-1-118-57521-5, Wiley (2014)
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 18/25
...
18/25

19. .
Caching options with LHAP
• (don't laugh but) caching methods have been lost on cloud-based CDN -- a true
clean slate
• yet, LHAP makes it possible to add additional caches
• note: Cache 1 and Cache 2 are completely different in nature -- VMs
are short-lived
CDN
Main
many
Manager
Storage
Large
Federation,
fog cloudsVM
Cache 1 Cache 2
Manage
content
CDN Traffic
(End Users)
many
Cloud
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 19/25
...
19/25

20. .
Back to Networks: Link with NGN
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 20/25
...
20/25

21. .
Key Viewpoints
1. traditional NGN in a recent book 08
◦ virtualization is key (SDN, NFV), among other reasons for supporting mobility
2. my own viewpoint is different
◦ because of hotspot traffic (bulk transfer = BigData = hotspots), e2e circuits
(cut-through) are key 09
◦ circuits are made in a distributed/social/sensing manner 10 close to the Y.mnm 11
story
◦ as a guy who was working on a distributed network of active probes 12 (now cloud
boxes), the Y.mnm story is close to mine
08 T.Janevski "NGN Architectures, Protocols and Services" Wiley (2014)
09 M.Zhanikeev "The Next Generation of Networks is all about Hotspot Distributions and Cut-Through Circuits" http://goo.gl/a16JtD (2015)
10 M.Zhankeev "A City Traffic Model for Optical Circuit Switching in Data Centers" http://goo.gl/5z2V6F (2015)
11 "Management of Performance Measurement for NGN" ITU-T Y.2173, before that Y.mnm (2008)
12 M.Zhanikeev "A Home Gateway Box with Meter, Probe and L2 QoS Policy Edge" IEEE COMPSAC (also http://goo.gl/9graXy) (2013)
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 21/25
...
21/25

22. .
The Y.mnm Story : Active Probes
• 2 parts: active probing and
distributed management
11
• between OSPF and QoS
Routing, closer to the
latter, but that was before
VNE was defined
• active probing is its own
world of research 13
• among smaller projects
(smartphones, etc.)
Google-backed M-Lab is the
biggest
11 "Management of Performance Measurement for NGN" ITU-T Y.2173, before that Y.mnm (2008)
13 Y.Tanaka, M.Zhanikeev "Active Network Measurement: Theory, Methods, and Tools" ITU Association of Japan, ISBN 978-4916128072 (2009)
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 22/25
...
22/25

23. .
IETF Liason with Y.mnm
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 23/25
...
23/25

24. .
Wrapup
• complexity wise, VNE formulation is close to reality, but is NP
Hard-er-er-er
◦ effective heuristics are unlikely
• the Do-it-Yourself (DiY) Approach can resolve the
complexity problem
◦ non-cooperative form: networks optimize themselves (example: Cloud Probing
05
◦ assisted form: clouds/networks can implement a Local Awareness
Feature (like LHAP 06) to improve self-optimization
• the Assisted DiY also leads to better coordination across competing
service networks
05 M.Zhanikeev "Performance Management of Cloud Populations via Cloud Probing" IPSJ JIP (related to http://goo.gl/f3QpiW) (2015)
06 M.Zhanikeev "A Cloud Visitation Platform to Facilitate Cloud Federation and Fog Computing" IEEE Computer (2015)
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 24/25
...
24/25

25. .
That’s all, thank you ...
M.Zhanikeev -- maratishe@gmail.com Population Management in Clouds is a Do-It-Yourself Technology -- bit.do/*** 25/25
...
25/25