SlideShare ist ein Scribd-Unternehmen logo
1 von 83
Managing Data in
Computational Edge Clouds
ACM SIGCOMM 2017 Workshop on Mobile Edge Communications
(MECOMM’2017), Los Angeles
Nitinder Mohan✼ Pengyuan Zhou✼ Keerthana Govindaraj△ Jussi Kangasharju✼
✼University of Helsinki △Robert Bosch GmbH
✼{firstname.lastname@helsinki.fi} △{keerthana.govindaraj@de.bosch.com}
Rise of Edge Clouds
Network
DatacenterEdge
Server
User
2
Edge Cloud: Small-scale server(s)
deployed at network edge to compute user
data
Motivation:
Rise of Edge Clouds
Network
DatacenterEdge
Server
User
Edge Cloud: Small-scale server(s)
deployed at network edge to compute user
data
Motivation:
üDecreased latency and network traffic
2
Rise of Edge Clouds
Network
Edge Cloud: Small-scale server(s)
deployed at network edge to compute user
data
Motivation:
üDecreased latency and network traffic
üComputing data of local relevance
2
Edge-Fog Cloud
Nitinder Mohan and Jussi Kangasharju. "Edge-fog cloud: A distributed cloud for internet of things computations." Cloudification of the Internet of Things (CIoT). IEEE, 2016.
• Voluntary compute devices
• 1-2 network hops from users
• Ad-hoc network connectivity
• Dedicated compute servers
• 5-6 network hops from users
• Dense network connectivity
• Permanent data archival
• No compute capability
3
Data Availability in Edge Cache
Query: Give me a route from point
A to B without road disruptions
4
Data Availability in Edge Cache
• Edge cloud proposals presume that the
required data for processing is pre-
available in Edge servers cache
• However, Edge server may request
Data Store for required data
Data Request: Any construction/
accidents on the map?
4
Data Availability in Edge Cache
• Edge cloud proposals presume that the
required data for processing is pre-
available in Edge servers cache
• However, Edge server may request
Data Store for required data
• The computation is delayed by a
round-trip-time to central Data Store
• Reduces benefits of employing
Edge clouds!
Data Request: Any construction/
accidents on the map?
Requirement: Caching strategy which predicts and stores
required data before computation 4
Edge Cloud Cache Grouping
Edge cloud caching mechanism
ØGroup edge resources into cache groups based on their locally
cached data
ØPredict required data for computation based on cache grouping
classification
ØProvide coherence on multiple copies of shared data throughout
a cache group
5
Resource Cache
Grouping
Grouping Algorithm
Task Deployment
Populate
Caches
Grouping
Classifiers
7
Grouping Algorithm
Deployment: Select {RC}i resources
for computing a task Tc
1. Algorithms such as LPCF* selects a
subset from available resources
which achieves least processing and
network cost for collaboration.
2. Several deployments for different
tasks can occur at same time.
Task Deployment
T1
T2
Nitinder Mohan and Jussi Kangasharju. "Edge-fog cloud: A distributed cloud for internet of things
computations." Cloudification of the Internet of Things (CIoT). IEEE, 2016.
8
Grouping Algorithm
1. At t=0, {RC}i does not have data
required for computation and it
requests the required data from
Data Store
2. The Data Store sends the required
data to requesting resources
populating their caches
Task Deployment
Populate Caches
9
Grouping Algorithm
1. At t=0, {RC}i does not have data
required for computation and it
requests the required data from
Data Store
2. The Data Store sends the required
data to requesting resources
populating their caches
Task Deployment
Populate Caches
9
Grouping Algorithm
1. The task Tc is classified in a
workload Wk where Wk is a set of
similar tasks
2. The resources {RC}i caching data
of workload Wk are grouped
together as cache group CGk
Task Deployment
Populate Caches
Grouping
Classifiers
W1
W2
10
Grouping Algorithm
1. The deployment algorithm prefers
CGk	∈Wk for task Ti	∈Wk
Task Deployment
Populate Caches
Grouping
Classifiers
W1
W2
T3
11
Grouping Algorithm
1. The deployment algorithm prefers
CGk	∈Wk for task Ti	∈Wk
2. If the resources in group are not
appropriate for deployment, the
algorithm can select other available
resources
Task Deployment
Populate Caches
Grouping
Classifiers
W1
W2
11
Grouping Algorithm
1. The deployment algorithm prefers
CGk	∈Wk for task Ti	∈Wk
2. If the resources in group are not
appropriate for deployment, the
algorithm can select other available
resources
3. These resources then join their
respective cache group
Task Deployment
Populate Caches
Grouping
Classifiers
W1
W2
11
Grouping Algorithm
1. The deployment algorithm prefers
CGk	∈Wk for task Ti	∈Wk
2. If the resources in group are not
appropriate for deployment, the
algorithm can select other available
resources
3. These resources then join their
respective cache group
4. A resource can be a part of more
than one group
Task Deployment
Populate Caches
Grouping
Classifiers
W1
W2
11
Group Communication
Requirements
1. Reduce unnecessary network traffic
2. Ensure consistent computation by avoiding
stale data copies
13
Resource Roles
1. Group Member
• Resource (RCx) part of a cache group
{CG}i
• The group member stores only data of
its group workload
14
Resource Roles
1. Group Member
• Resource (RCx) part of a cache group
{CG}i
• The group member stores only data of
its group workload
2. Group Leader
• Representative of a group
• Responsible for maintaining coherent
data within a group
• Has consistent connectivity with group
members
• Elected via distributed election algorithm
14
Communication Flow
The communication model is strictly pull-based to limit number of
messages in the system
The model operates over following assumptions:
1. The resources upload their updated data copy to Data Store
after every successful computation
2. Data upload is in-parallel to ongoing local computation and
does not impact the computation completion time
3. Data Store is persistent and immune to failures
4. Messages in the system are not lost 15
Cache Data Structures
1. Group member table
Maps locally cached data with
group leader
2. Group leader table
Ensures that group leader has
consistent state and address of
resource caching data
3. Data Store table
Maps data to groups
Data Name Leader Address Tag
Data Name Leader Address Classifier Type
Data Name Resource Address Tag Timestamp
16
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario	1:	Retrieving	free	data
17
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
Scenario	1:	Retrieving	free	data
17
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
Scenario	1:	Retrieving	free	data
17
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
GoTo: ’Resource 1’
Scenario	1:	Retrieving	free	data
17
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
GoTo: ’Resource 1’
Content ‘X’?
Scenario	1:	Retrieving	free	data
17
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
GoTo: ’Resource 1’
Content ‘X’?
Scenario	1:	Retrieving	free	data
17
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
GoTo: ’Resource 1’
Content ‘X’?
Content ‘X’
Scenario	1:	Retrieving	free	data
17
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
GoTo: ’Resource 1’
Content ‘X’?
Content ‘X’
Scenario	1:	Retrieving	free	data
Content ’X’
17
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
GoTo: ’Resource 1’
Content ‘X’?
Scenario	2: Resource	failure
18
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
GoTo: ’Resource 1’
Content ‘X’?
Unavailable!
Scenario	2: Resource	failure
18
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
GoTo: ’Resource 1’
Content ‘X’?
Timeout
Unavailable!
Scenario	2: Resource	failure
18
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
GoTo: ’Resource 1’
Content ‘X’?
Content ‘X’?
Timeout
Unavailable!
Scenario	2: Resource	failure
18
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
GoTo: ’Resource 1’
Content ‘X’?
Content ‘X’?
Timeout
Unavailable!
Scenario	2: Resource	failure
18
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
GoTo: ’Resource 1’
Content ‘X’?
Content ‘X’?
Timeout
Content ‘X’
Unavailable!
Scenario	2: Resource	failure
18
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
GoTo: ’Resource 1’
Content ‘X’?
Content ‘X’?
Timeout
Content ‘X’
Unavailable!
Scenario	2: Resource	failure
Content ’X’
18
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Lock
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Lock Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario	3:	Data	under	update
19
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Lock
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Lock Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario	3:	Data	under	update
Content ‘X’?
19
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Lock
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Lock Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario	3:	Data	under	update
Content ‘X’?
19
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Lock
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Lock Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Lock: ’Resource 1’
Scenario	3:	Data	under	update
Content ‘X’?
19
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Lock
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Lock Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Lock: ’Resource 1’
Scenario	3:	Data	under	update
Time Interval
Content ‘X’?
19
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Lock
Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Lock Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Lock: ’Resource 1’
Scenario	3:	Data	under	update
Time Interval
Content ‘X’?
19
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Lock Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Lock: ’Resource 1’
Scenario	3:	Data	under	update
Time Interval
Content ‘X’?
Free
19
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Lock Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Lock: ’Resource 1’
Scenario	3:	Data	under	update
Time Interval
Free: Content ‘X’
Content ‘X’?
Free
19
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Lock Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Lock: ’Resource 1’
Scenario	3:	Data	under	update
Time Interval
Free: Content ‘X’
Content ‘X’?
Free
19
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Lock: ’Resource 1’
Scenario	3:	Data	under	update
Time Interval
Free: Content ‘X’
Content ‘X’?
Free Free
19
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
Lock: ’Resource 1’
Scenario	3:	Data	under	update
Time Interval
Free: Content ‘X’
Content ‘X’?
Free Free
19
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
Lock: ’Resource 1’
Scenario	3:	Data	under	update
Time Interval
Free: Content ‘X’
Content ‘X’?
GoTo: ’Resource 1’
Free Free
19
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
Lock: ’Resource 1’
Content ‘X’?
Scenario	3:	Data	under	update
Time Interval
Free: Content ‘X’
Content ‘X’?
GoTo: ’Resource 1’
Free Free
19
Content Retrival in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A Lock
Resource 2
Member Table
Content ’X’ (stale)
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Content ‘X’?
Lock: ’Resource 1’
Content ‘X’?
Content ‘X’
Scenario	3:	Data	under	update
Time Interval
Free: Content ‘X’
Content ‘X’?
GoTo: ’Resource 1’
Free Free
19
Content Update in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario:	Update	on	Content	‘X’	by	Resource	2
20
Content Update in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario:	Update	on	Content	‘X’	by	Resource	2
Resource 2: Retrieve updated Content ‘X’
20
Content Update in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’
Content ‘Q’
Leader A
Leader A
Free
Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario:	Update	on	Content	‘X’	by	Resource	2
Resource 2: Retrieve updated Content ‘X’
20
Content Update in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’
Content ‘Q’
Leader A
Leader A Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario:	Update	on	Content	‘X’	by	Resource	2
Resource 2: Retrieve updated Content ‘X’
Lock
20
Content Update in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’
Content ‘Q’
Leader A
Leader A Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario:	Update	on	Content	‘X’	by	Resource	2
Resource 2: Retrieve updated Content ‘X’
Lock: Content ‘X’
Lock
20
Content Update in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’
Content ‘Q’
Leader A
Leader A Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Free Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario:	Update	on	Content	‘X’	by	Resource	2
Resource 2: Retrieve updated Content ‘X’
Lock: Content ‘X’
Lock
20
Content Update in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’
Content ‘Q’
Leader A
Leader A Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario:	Update	on	Content	‘X’	by	Resource	2
Resource 2: Retrieve updated Content ‘X’
Lock: Content ‘X’
Lock Lock
20
Resource 2
Content Update in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’
Content ‘Q’
Leader A
Leader A Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario:	Update	on	Content	‘X’	by	Resource	2
Resource 2: Retrieve updated Content ‘X’
Lock: Content ‘X’
Processing
Lock Lock
20
Resource 2
Content Update in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’
Content ‘Q’
Leader A
Leader A Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario:	Update	on	Content	‘X’	by	Resource	2
Resource 2: Retrieve updated Content ‘X’
Lock: Content ‘X’
Processing
Free Lock
20
Resource 2
Content Update in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’
Content ‘Q’
Leader A
Leader A Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Tx
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario:	Update	on	Content	‘X’	by	Resource	2
Resource 2: Retrieve updated Content ‘X’
Lock: Content ‘X’
Processing
Free: Content ‘X’
Lock Lock
20
Resource 2
Content Update in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’
Content ‘Q’
Leader A
Leader A Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Tx1
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario:	Update	on	Content	‘X’	by	Resource	2
Resource 2: Retrieve updated Content ‘X’
Lock: Content ‘X’
Processing
Free: Content ‘X’
Lock Free
20
Resource 2
Content Update in Group
Resource 1
Member Table
Content ’X’
Content ‘Y’
Leader A
Leader A
Free
Lock
Resource 2
Member Table
Content ’X’
Content ‘Q’
Leader A
Leader A Free
Leader A
Leader Table
Content ’X’
Content ‘Y’
Tx1
Ty
Content ‘Q’ Free
Lock
Tq
Resource 1
Resource 2Content ‘Z’ Leader B Free
Data Store
Data Store Table
Content ’X’
Content ‘Y’
Leader A
Leader B
Wi
WjContent ‘Z’
Content ‘Q’
Leader A
Leader A
Wi
Wi
Scenario:	Update	on	Content	‘X’	by	Resource	2
Resource 2: Retrieve updated Content ‘X’
Lock: Content ‘X’
Processing
Free: Content ‘X’
Content ‘X’
Lock Free
20
Resource 2
Requirements
1. Reduce unnecessary network traffic
Ø Communication model encourages data sharing
within a group and removes a single point-of-failure
2. Ensure consistent computation by avoiding
stale data copies
Ø Model ensures causal coherence on shared data
21
Evaluation
Evaluation
Evaluated on Icarus simulator*
Parameters:
1. 320 Edge and Fog nodes + Data Store
2. Fog node acts as group leader
3. 96*104 content items divided in 32 workloads
4. Resource cache size capacity is 10% of overall content size
*Lorenzo Saino, Ioannis Psaras, and George Pavlou. "ICARUS: a caching simulator for information centric networking (ICN)." Proceedings of the 7th International
ICST conference on Simulation Tools and Techniques. ICST (Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering), 2014. 23
Evaluation
Q1. Does resource grouping offer any benefits to edge
clouds?
Q2. If yes, what is the appropriate size of a group?
24
Q1. Does resource grouping offer any
benefits to edge clouds?
a) Effect on cache hit rate
0.02 0.04 0.06 0.08 0.1
Cache Size
0
0.2
0.4
0.6
0.8
1
CacheHitRate
4-group
4-no_group
32-group
32-no_group
25
Q1. Does resource grouping offer any
benefits to edge clouds?
a) Effect on cache hit rate
0.02 0.04 0.06 0.08 0.1
Cache Size
0
0.2
0.4
0.6
0.8
1
CacheHitRate
4-group
4-no_group
32-group
32-no_group
≈2	times	increase	in	hit	rate
25
Q1. Does resource grouping offer any
benefits to edge clouds?
b) Effect on latency
0.02 0.04 0.06 0.08 0.1
Cache Size
20
30
40
50
60
70
80
Latency(ms)
4-group
4-no_group
32-group
32-no_group
26
Q1. Does resource grouping offer any
benefits to edge clouds?
b) Effect on latency
0.02 0.04 0.06 0.08 0.1
Cache Size
20
30
40
50
60
70
80
Latency(ms)
4-group
4-no_group
32-group
32-no_group
45%	decrease	in	latency
26
Q2. If yes, what is the appropriate size
of a group?
0 25 50 75 100 125 150
Number of Groups
0.2
0.4
0.6
0.8
1
CacheHitRate
4 workloads
8 workloads
16 workloads
32 workloads
0 25 50 75 100 125 150
Number of Groups
20
30
40
50
Latency(ms)
4 workloads
8 workloads
16 workloads
32 workloads
Maximum gains are achieved when
number of cache groups = number of workloads
27
Conclusion
1. Edge resource grouping strategy manages content in Edge caches
2. Strategy classifies content based on resource’s deployed
workloads and cached content
3. The 3-step iterative algorithm maps future deployments on
resources with required data locally cached
4. Our evaluations via simulation show
i. Direct correlation between grouping and increased system
performance
ii. Number of cache groups and number of assigned workloads
28
Thank You!
Questions?
nitinder.mohan@helsinki.fi
CleanSky - EU FP7 Marie Curie Initial Training Network
Backup Slides
Industry 4.0 on Edge
Industry 4.0 proposes automation in
manufacturing
1. Robots such as Bosch APAS can
automate tasks and work along side
humans while ensuring safety
2. IoT sensor data is used in augmented
reality for collaborative support and
manufacturing
Requirement: Time-critical and fast data
processing
Industry 4.0 on Edge
Industry 4.0 proposes automation in
manufacturing
1. Robots such as Bosch APAS can
automate tasks and work along side
humans while ensuring safety
2. IoT sensor data is used in augmented
reality for collaborative support and
manufacturing
Requirement: Time-critical and fast data
processing
Solution: Implement industrial Edge clouds
Network
Datacenter
Edge
Servers
Grouping Classifier
1. Location: location of place while
constructing AR and 3D models
2. Relevance: laptop and cellphone
manufacturing tasks as casing
3. Time: sensor data generation and
‘freshness’ time
4. Warning signals: broadcast of
information-only or safety-related
signals
Q1. Does resource grouping offer any
benefits to edge clouds?
c) Effect of cache size
0.02 0.04 0.06 0.08 0.1
Cache Size
20
30
40
50
60
70
80
Latency(ms)
4-group
4-no_group
32-group
32-no_group
0.02 0.04 0.06 0.08 0.1
Cache Size
0
0.2
0.4
0.6
0.8
1
CacheHitRate
4-group
4-no_group
32-group
32-no_group
Gains decrease as we increase cache size!
1. Grouping works best when cache sizes are limited and cannot store all content
2. With larger caches, resources can locally store all content!
Q1. Does resource grouping offer any
benefits to edge clouds?
d) Effect of number of workloads
0.02 0.04 0.06 0.08 0.1
Cache Size
20
30
40
50
60
70
80
Latency(ms)
4-group
4-no_group
32-group
32-no_group
0.02 0.04 0.06 0.08 0.1
Cache Size
0
0.2
0.4
0.6
0.8
1
CacheHitRate
4-group
4-no_group
32-group
32-no_group
Gains decrease on increasing number of workloads!
Ø More workloads ≈ more unique content requests which cannot be satisfied by local
caches
Workloads is a uniform distribution of similar content requests

Weitere ähnliche Inhalte

Was ist angesagt?

3 ijece 1 ed iqbal qc
3 ijece 1 ed iqbal qc3 ijece 1 ed iqbal qc
3 ijece 1 ed iqbal qcIAESIJEECS
 
An efficient, secure deduplication data storing in cloud storage environment
An efficient, secure deduplication data storing in cloud storage environmentAn efficient, secure deduplication data storing in cloud storage environment
An efficient, secure deduplication data storing in cloud storage environmenteSAT Journals
 
Secure Mining of Association Rules in Horizontally Distributed Databases
Secure Mining of Association Rules in Horizontally Distributed DatabasesSecure Mining of Association Rules in Horizontally Distributed Databases
Secure Mining of Association Rules in Horizontally Distributed DatabasesIJSRD
 
Privacy Preserving Multi-keyword Top-K Search based on Cosine Similarity Clus...
Privacy Preserving Multi-keyword Top-K Search based on Cosine Similarity Clus...Privacy Preserving Multi-keyword Top-K Search based on Cosine Similarity Clus...
Privacy Preserving Multi-keyword Top-K Search based on Cosine Similarity Clus...IRJET Journal
 
An effective classification approach for big data with parallel generalized H...
An effective classification approach for big data with parallel generalized H...An effective classification approach for big data with parallel generalized H...
An effective classification approach for big data with parallel generalized H...riyaniaes
 
Introduction: Linked Data and the Semantic Web
Introduction: Linked Data and the Semantic WebIntroduction: Linked Data and the Semantic Web
Introduction: Linked Data and the Semantic WebNikolaos Konstantinou
 
IRJET- Multiple Keyword Search over Encrypted Cloud Data
IRJET- Multiple Keyword Search over Encrypted Cloud DataIRJET- Multiple Keyword Search over Encrypted Cloud Data
IRJET- Multiple Keyword Search over Encrypted Cloud DataIRJET Journal
 
Deploying Linked Open Data: Methodologies and Software Tools
Deploying Linked Open Data: Methodologies and Software ToolsDeploying Linked Open Data: Methodologies and Software Tools
Deploying Linked Open Data: Methodologies and Software ToolsNikolaos Konstantinou
 
Integrity Privacy to Public Auditing for Shared Data in Cloud Computing
Integrity Privacy to Public Auditing for Shared Data in Cloud ComputingIntegrity Privacy to Public Auditing for Shared Data in Cloud Computing
Integrity Privacy to Public Auditing for Shared Data in Cloud ComputingIJERA Editor
 
Materializing the Web of Linked Data
Materializing the Web of Linked DataMaterializing the Web of Linked Data
Materializing the Web of Linked DataNikolaos Konstantinou
 
A Survey on Approaches for Frequent Item Set Mining on Apache Hadoop
A Survey on Approaches for Frequent Item Set Mining on Apache HadoopA Survey on Approaches for Frequent Item Set Mining on Apache Hadoop
A Survey on Approaches for Frequent Item Set Mining on Apache HadoopIJTET Journal
 
A hybrid cloud approach for secure authorized deduplication
A hybrid cloud approach for secure authorized deduplicationA hybrid cloud approach for secure authorized deduplication
A hybrid cloud approach for secure authorized deduplicationTmks Infotech
 
DataXDay - Exploring graphs: looking for communities & leaders
DataXDay - Exploring graphs: looking for communities & leadersDataXDay - Exploring graphs: looking for communities & leaders
DataXDay - Exploring graphs: looking for communities & leadersDataXDay Conference by Xebia
 
Key-Aggregate Searchable Encryption (KASE) for Group Data Sharing via Cloud S...
Key-Aggregate Searchable Encryption (KASE) for Group Data Sharing via Cloud S...Key-Aggregate Searchable Encryption (KASE) for Group Data Sharing via Cloud S...
Key-Aggregate Searchable Encryption (KASE) for Group Data Sharing via Cloud S...1crore projects
 
A Novel Additive Order Protocol in Cloud Storage and Avoiding the Trapdoors
A Novel Additive Order Protocol in Cloud Storage and Avoiding the TrapdoorsA Novel Additive Order Protocol in Cloud Storage and Avoiding the Trapdoors
A Novel Additive Order Protocol in Cloud Storage and Avoiding the TrapdoorsIRJET Journal
 
Duplicate File Analyzer using N-layer Hash and Hash Table
Duplicate File Analyzer using N-layer Hash and Hash TableDuplicate File Analyzer using N-layer Hash and Hash Table
Duplicate File Analyzer using N-layer Hash and Hash TableAM Publications
 
Turn InSecure And High Speed Intra-Cloud and Inter-Cloud Communication
Turn InSecure And High Speed Intra-Cloud and Inter-Cloud CommunicationTurn InSecure And High Speed Intra-Cloud and Inter-Cloud Communication
Turn InSecure And High Speed Intra-Cloud and Inter-Cloud CommunicationRichard Jung
 
Cloud-based Data Stream Processing
Cloud-based Data Stream ProcessingCloud-based Data Stream Processing
Cloud-based Data Stream ProcessingZbigniew Jerzak
 

Was ist angesagt? (20)

3 ijece 1 ed iqbal qc
3 ijece 1 ed iqbal qc3 ijece 1 ed iqbal qc
3 ijece 1 ed iqbal qc
 
An efficient, secure deduplication data storing in cloud storage environment
An efficient, secure deduplication data storing in cloud storage environmentAn efficient, secure deduplication data storing in cloud storage environment
An efficient, secure deduplication data storing in cloud storage environment
 
Secure Mining of Association Rules in Horizontally Distributed Databases
Secure Mining of Association Rules in Horizontally Distributed DatabasesSecure Mining of Association Rules in Horizontally Distributed Databases
Secure Mining of Association Rules in Horizontally Distributed Databases
 
Privacy Preserving Multi-keyword Top-K Search based on Cosine Similarity Clus...
Privacy Preserving Multi-keyword Top-K Search based on Cosine Similarity Clus...Privacy Preserving Multi-keyword Top-K Search based on Cosine Similarity Clus...
Privacy Preserving Multi-keyword Top-K Search based on Cosine Similarity Clus...
 
An effective classification approach for big data with parallel generalized H...
An effective classification approach for big data with parallel generalized H...An effective classification approach for big data with parallel generalized H...
An effective classification approach for big data with parallel generalized H...
 
Introduction: Linked Data and the Semantic Web
Introduction: Linked Data and the Semantic WebIntroduction: Linked Data and the Semantic Web
Introduction: Linked Data and the Semantic Web
 
IRJET- Multiple Keyword Search over Encrypted Cloud Data
IRJET- Multiple Keyword Search over Encrypted Cloud DataIRJET- Multiple Keyword Search over Encrypted Cloud Data
IRJET- Multiple Keyword Search over Encrypted Cloud Data
 
A1803010105
A1803010105A1803010105
A1803010105
 
Deploying Linked Open Data: Methodologies and Software Tools
Deploying Linked Open Data: Methodologies and Software ToolsDeploying Linked Open Data: Methodologies and Software Tools
Deploying Linked Open Data: Methodologies and Software Tools
 
Integrity Privacy to Public Auditing for Shared Data in Cloud Computing
Integrity Privacy to Public Auditing for Shared Data in Cloud ComputingIntegrity Privacy to Public Auditing for Shared Data in Cloud Computing
Integrity Privacy to Public Auditing for Shared Data in Cloud Computing
 
Materializing the Web of Linked Data
Materializing the Web of Linked DataMaterializing the Web of Linked Data
Materializing the Web of Linked Data
 
A Survey on Approaches for Frequent Item Set Mining on Apache Hadoop
A Survey on Approaches for Frequent Item Set Mining on Apache HadoopA Survey on Approaches for Frequent Item Set Mining on Apache Hadoop
A Survey on Approaches for Frequent Item Set Mining on Apache Hadoop
 
A hybrid cloud approach for secure authorized deduplication
A hybrid cloud approach for secure authorized deduplicationA hybrid cloud approach for secure authorized deduplication
A hybrid cloud approach for secure authorized deduplication
 
DataXDay - Exploring graphs: looking for communities & leaders
DataXDay - Exploring graphs: looking for communities & leadersDataXDay - Exploring graphs: looking for communities & leaders
DataXDay - Exploring graphs: looking for communities & leaders
 
Key-Aggregate Searchable Encryption (KASE) for Group Data Sharing via Cloud S...
Key-Aggregate Searchable Encryption (KASE) for Group Data Sharing via Cloud S...Key-Aggregate Searchable Encryption (KASE) for Group Data Sharing via Cloud S...
Key-Aggregate Searchable Encryption (KASE) for Group Data Sharing via Cloud S...
 
A Novel Additive Order Protocol in Cloud Storage and Avoiding the Trapdoors
A Novel Additive Order Protocol in Cloud Storage and Avoiding the TrapdoorsA Novel Additive Order Protocol in Cloud Storage and Avoiding the Trapdoors
A Novel Additive Order Protocol in Cloud Storage and Avoiding the Trapdoors
 
Dq36708711
Dq36708711Dq36708711
Dq36708711
 
Duplicate File Analyzer using N-layer Hash and Hash Table
Duplicate File Analyzer using N-layer Hash and Hash TableDuplicate File Analyzer using N-layer Hash and Hash Table
Duplicate File Analyzer using N-layer Hash and Hash Table
 
Turn InSecure And High Speed Intra-Cloud and Inter-Cloud Communication
Turn InSecure And High Speed Intra-Cloud and Inter-Cloud CommunicationTurn InSecure And High Speed Intra-Cloud and Inter-Cloud Communication
Turn InSecure And High Speed Intra-Cloud and Inter-Cloud Communication
 
Cloud-based Data Stream Processing
Cloud-based Data Stream ProcessingCloud-based Data Stream Processing
Cloud-based Data Stream Processing
 

Ähnlich wie Managing data in computational edge clouds

Big Data in a neurophysiology research lab… what?
Big Data in a neurophysiology research lab… what?Big Data in a neurophysiology research lab… what?
Big Data in a neurophysiology research lab… what?J On The Beach
 
Distributed Framework for Data Mining As a Service on Private Cloud
Distributed Framework for Data Mining As a Service on Private CloudDistributed Framework for Data Mining As a Service on Private Cloud
Distributed Framework for Data Mining As a Service on Private CloudIJERA Editor
 
International Refereed Journal of Engineering and Science (IRJES)
International Refereed Journal of Engineering and Science (IRJES)International Refereed Journal of Engineering and Science (IRJES)
International Refereed Journal of Engineering and Science (IRJES)irjes
 
F233842
F233842F233842
F233842irjes
 
The Materials Data Facility: A Distributed Model for the Materials Data Commu...
The Materials Data Facility: A Distributed Model for the Materials Data Commu...The Materials Data Facility: A Distributed Model for the Materials Data Commu...
The Materials Data Facility: A Distributed Model for the Materials Data Commu...Ben Blaiszik
 
Apache Kafka and the Data Mesh | Ben Stopford and Michael Noll, Confluent
Apache Kafka and the Data Mesh | Ben Stopford and Michael Noll, ConfluentApache Kafka and the Data Mesh | Ben Stopford and Michael Noll, Confluent
Apache Kafka and the Data Mesh | Ben Stopford and Michael Noll, ConfluentHostedbyConfluent
 
Dynamic Resource Allocation Algorithm using Containers
Dynamic Resource Allocation Algorithm using ContainersDynamic Resource Allocation Algorithm using Containers
Dynamic Resource Allocation Algorithm using ContainersIRJET Journal
 
2015 04 bio it world
2015 04 bio it world2015 04 bio it world
2015 04 bio it worldChris Dwan
 
What is a Data Commons and Why Should You Care?
What is a Data Commons and Why Should You Care? What is a Data Commons and Why Should You Care?
What is a Data Commons and Why Should You Care? Robert Grossman
 
A Gen3 Perspective of Disparate Data
A Gen3 Perspective of Disparate DataA Gen3 Perspective of Disparate Data
A Gen3 Perspective of Disparate DataRobert Grossman
 
Microtask Crowdsourcing Applications for Linked Data
Microtask Crowdsourcing Applications for Linked DataMicrotask Crowdsourcing Applications for Linked Data
Microtask Crowdsourcing Applications for Linked DataEUCLID project
 
Deep Learning Based Real-Time DNS DDoS Detection System
Deep Learning Based Real-Time DNS DDoS Detection SystemDeep Learning Based Real-Time DNS DDoS Detection System
Deep Learning Based Real-Time DNS DDoS Detection SystemSeungjoo Kim
 
Data Infrastructure at LinkedIn
Data Infrastructure at LinkedInData Infrastructure at LinkedIn
Data Infrastructure at LinkedInAmy W. Tang
 
Towards an Incremental Schema-level Index for Distributed Linked Open Data G...
Towards an Incremental Schema-level Index  for Distributed Linked Open Data G...Towards an Incremental Schema-level Index  for Distributed Linked Open Data G...
Towards an Incremental Schema-level Index for Distributed Linked Open Data G...Till Blume
 
Dynamic Data Center concept
Dynamic Data Center concept  Dynamic Data Center concept
Dynamic Data Center concept Miha Ahronovitz
 
Efficient frequent pattern mining in distributed system
Efficient frequent pattern mining in distributed systemEfficient frequent pattern mining in distributed system
Efficient frequent pattern mining in distributed systemSaurav Kumar
 
Big Data in Distributed Analytics,Cybersecurity And Digital Forensics
Big Data in Distributed Analytics,Cybersecurity And Digital ForensicsBig Data in Distributed Analytics,Cybersecurity And Digital Forensics
Big Data in Distributed Analytics,Cybersecurity And Digital ForensicsSherinMariamReji05
 

Ähnlich wie Managing data in computational edge clouds (20)

Big Data in a neurophysiology research lab… what?
Big Data in a neurophysiology research lab… what?Big Data in a neurophysiology research lab… what?
Big Data in a neurophysiology research lab… what?
 
Distributed Framework for Data Mining As a Service on Private Cloud
Distributed Framework for Data Mining As a Service on Private CloudDistributed Framework for Data Mining As a Service on Private Cloud
Distributed Framework for Data Mining As a Service on Private Cloud
 
International Refereed Journal of Engineering and Science (IRJES)
International Refereed Journal of Engineering and Science (IRJES)International Refereed Journal of Engineering and Science (IRJES)
International Refereed Journal of Engineering and Science (IRJES)
 
F233842
F233842F233842
F233842
 
The Materials Data Facility: A Distributed Model for the Materials Data Commu...
The Materials Data Facility: A Distributed Model for the Materials Data Commu...The Materials Data Facility: A Distributed Model for the Materials Data Commu...
The Materials Data Facility: A Distributed Model for the Materials Data Commu...
 
Apache Kafka and the Data Mesh | Ben Stopford and Michael Noll, Confluent
Apache Kafka and the Data Mesh | Ben Stopford and Michael Noll, ConfluentApache Kafka and the Data Mesh | Ben Stopford and Michael Noll, Confluent
Apache Kafka and the Data Mesh | Ben Stopford and Michael Noll, Confluent
 
Dynamic Resource Allocation Algorithm using Containers
Dynamic Resource Allocation Algorithm using ContainersDynamic Resource Allocation Algorithm using Containers
Dynamic Resource Allocation Algorithm using Containers
 
Shikha fdp 62_14july2017
Shikha fdp 62_14july2017Shikha fdp 62_14july2017
Shikha fdp 62_14july2017
 
Thoughts on Cybersecurity
Thoughts on CybersecurityThoughts on Cybersecurity
Thoughts on Cybersecurity
 
2015 04 bio it world
2015 04 bio it world2015 04 bio it world
2015 04 bio it world
 
What is a Data Commons and Why Should You Care?
What is a Data Commons and Why Should You Care? What is a Data Commons and Why Should You Care?
What is a Data Commons and Why Should You Care?
 
A Gen3 Perspective of Disparate Data
A Gen3 Perspective of Disparate DataA Gen3 Perspective of Disparate Data
A Gen3 Perspective of Disparate Data
 
Microtask Crowdsourcing Applications for Linked Data
Microtask Crowdsourcing Applications for Linked DataMicrotask Crowdsourcing Applications for Linked Data
Microtask Crowdsourcing Applications for Linked Data
 
Deep Learning Based Real-Time DNS DDoS Detection System
Deep Learning Based Real-Time DNS DDoS Detection SystemDeep Learning Based Real-Time DNS DDoS Detection System
Deep Learning Based Real-Time DNS DDoS Detection System
 
Data Infrastructure at LinkedIn
Data Infrastructure at LinkedInData Infrastructure at LinkedIn
Data Infrastructure at LinkedIn
 
Towards an Incremental Schema-level Index for Distributed Linked Open Data G...
Towards an Incremental Schema-level Index  for Distributed Linked Open Data G...Towards an Incremental Schema-level Index  for Distributed Linked Open Data G...
Towards an Incremental Schema-level Index for Distributed Linked Open Data G...
 
Dynamic Data Center concept
Dynamic Data Center concept  Dynamic Data Center concept
Dynamic Data Center concept
 
Efficient frequent pattern mining in distributed system
Efficient frequent pattern mining in distributed systemEfficient frequent pattern mining in distributed system
Efficient frequent pattern mining in distributed system
 
Big Data in Distributed Analytics,Cybersecurity And Digital Forensics
Big Data in Distributed Analytics,Cybersecurity And Digital ForensicsBig Data in Distributed Analytics,Cybersecurity And Digital Forensics
Big Data in Distributed Analytics,Cybersecurity And Digital Forensics
 
Distributed Deep Learning + others for Spark Meetup
Distributed Deep Learning + others for Spark MeetupDistributed Deep Learning + others for Spark Meetup
Distributed Deep Learning + others for Spark Meetup
 

Mehr von Nitinder Mohan

Pruning Edge Research with Latency Shears
Pruning Edge Research with Latency ShearsPruning Edge Research with Latency Shears
Pruning Edge Research with Latency ShearsNitinder Mohan
 
Edge Computing Platforms and Protocols - Ph.D. thesis
Edge Computing Platforms and Protocols - Ph.D. thesisEdge Computing Platforms and Protocols - Ph.D. thesis
Edge Computing Platforms and Protocols - Ph.D. thesisNitinder Mohan
 
DeCloud: Truthful Decentralized Double Auction for Edge Clouds
DeCloud: Truthful Decentralized Double Auction for Edge CloudsDeCloud: Truthful Decentralized Double Auction for Edge Clouds
DeCloud: Truthful Decentralized Double Auction for Edge CloudsNitinder Mohan
 
Big Data Frameworks: A primer on Apache Spark and MapReduce
Big Data Frameworks: A primer on Apache Spark and MapReduceBig Data Frameworks: A primer on Apache Spark and MapReduce
Big Data Frameworks: A primer on Apache Spark and MapReduceNitinder Mohan
 
Open Infrastructure for Edge: A Distributed Ledger Outlook
Open Infrastructure for Edge: A Distributed Ledger OutlookOpen Infrastructure for Edge: A Distributed Ledger Outlook
Open Infrastructure for Edge: A Distributed Ledger OutlookNitinder Mohan
 
ExEC: Elastic Extensible Edge Cloud
ExEC: Elastic Extensible Edge Cloud ExEC: Elastic Extensible Edge Cloud
ExEC: Elastic Extensible Edge Cloud Nitinder Mohan
 
ICON: Intelligent Container Overlays
ICON: Intelligent Container OverlaysICON: Intelligent Container Overlays
ICON: Intelligent Container OverlaysNitinder Mohan
 
Redesigning MPTCP in Edge clouds
Redesigning MPTCP in Edge cloudsRedesigning MPTCP in Edge clouds
Redesigning MPTCP in Edge cloudsNitinder Mohan
 
Anveshak: Placing Edge Servers In The Wild
Anveshak: Placing Edge Servers In The WildAnveshak: Placing Edge Servers In The Wild
Anveshak: Placing Edge Servers In The WildNitinder Mohan
 
MUTE: Multi-Tier Edge networks
MUTE: Multi-Tier Edge networksMUTE: Multi-Tier Edge networks
MUTE: Multi-Tier Edge networksNitinder Mohan
 
Edge-Fog Cloud: Scaling IoT computations on the edge
Edge-Fog Cloud: Scaling IoT computations on the edgeEdge-Fog Cloud: Scaling IoT computations on the edge
Edge-Fog Cloud: Scaling IoT computations on the edgeNitinder Mohan
 

Mehr von Nitinder Mohan (12)

Pruning Edge Research with Latency Shears
Pruning Edge Research with Latency ShearsPruning Edge Research with Latency Shears
Pruning Edge Research with Latency Shears
 
Edge Computing Platforms and Protocols - Ph.D. thesis
Edge Computing Platforms and Protocols - Ph.D. thesisEdge Computing Platforms and Protocols - Ph.D. thesis
Edge Computing Platforms and Protocols - Ph.D. thesis
 
DeCloud: Truthful Decentralized Double Auction for Edge Clouds
DeCloud: Truthful Decentralized Double Auction for Edge CloudsDeCloud: Truthful Decentralized Double Auction for Edge Clouds
DeCloud: Truthful Decentralized Double Auction for Edge Clouds
 
Big Data Frameworks: A primer on Apache Spark and MapReduce
Big Data Frameworks: A primer on Apache Spark and MapReduceBig Data Frameworks: A primer on Apache Spark and MapReduce
Big Data Frameworks: A primer on Apache Spark and MapReduce
 
Open Infrastructure for Edge: A Distributed Ledger Outlook
Open Infrastructure for Edge: A Distributed Ledger OutlookOpen Infrastructure for Edge: A Distributed Ledger Outlook
Open Infrastructure for Edge: A Distributed Ledger Outlook
 
ExEC: Elastic Extensible Edge Cloud
ExEC: Elastic Extensible Edge Cloud ExEC: Elastic Extensible Edge Cloud
ExEC: Elastic Extensible Edge Cloud
 
ICON: Intelligent Container Overlays
ICON: Intelligent Container OverlaysICON: Intelligent Container Overlays
ICON: Intelligent Container Overlays
 
Redesigning MPTCP in Edge clouds
Redesigning MPTCP in Edge cloudsRedesigning MPTCP in Edge clouds
Redesigning MPTCP in Edge clouds
 
Anveshak: Placing Edge Servers In The Wild
Anveshak: Placing Edge Servers In The WildAnveshak: Placing Edge Servers In The Wild
Anveshak: Placing Edge Servers In The Wild
 
MUTE: Multi-Tier Edge networks
MUTE: Multi-Tier Edge networksMUTE: Multi-Tier Edge networks
MUTE: Multi-Tier Edge networks
 
Edge-Fog Cloud: Scaling IoT computations on the edge
Edge-Fog Cloud: Scaling IoT computations on the edgeEdge-Fog Cloud: Scaling IoT computations on the edge
Edge-Fog Cloud: Scaling IoT computations on the edge
 
Edge-Fog Cloud
Edge-Fog CloudEdge-Fog Cloud
Edge-Fog Cloud
 

Kürzlich hochgeladen

ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITYISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITYKayeClaireEstoconing
 
AMERICAN LANGUAGE HUB_Level2_Student'sBook_Answerkey.pdf
AMERICAN LANGUAGE HUB_Level2_Student'sBook_Answerkey.pdfAMERICAN LANGUAGE HUB_Level2_Student'sBook_Answerkey.pdf
AMERICAN LANGUAGE HUB_Level2_Student'sBook_Answerkey.pdfphamnguyenenglishnb
 
Proudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptxProudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptxthorishapillay1
 
Inclusivity Essentials_ Creating Accessible Websites for Nonprofits .pdf
Inclusivity Essentials_ Creating Accessible Websites for Nonprofits .pdfInclusivity Essentials_ Creating Accessible Websites for Nonprofits .pdf
Inclusivity Essentials_ Creating Accessible Websites for Nonprofits .pdfTechSoup
 
How to Add Barcode on PDF Report in Odoo 17
How to Add Barcode on PDF Report in Odoo 17How to Add Barcode on PDF Report in Odoo 17
How to Add Barcode on PDF Report in Odoo 17Celine George
 
How to do quick user assign in kanban in Odoo 17 ERP
How to do quick user assign in kanban in Odoo 17 ERPHow to do quick user assign in kanban in Odoo 17 ERP
How to do quick user assign in kanban in Odoo 17 ERPCeline George
 
ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...
ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...
ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...JhezDiaz1
 
Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)Mark Reed
 
THEORIES OF ORGANIZATION-PUBLIC ADMINISTRATION
THEORIES OF ORGANIZATION-PUBLIC ADMINISTRATIONTHEORIES OF ORGANIZATION-PUBLIC ADMINISTRATION
THEORIES OF ORGANIZATION-PUBLIC ADMINISTRATIONHumphrey A Beña
 
INTRODUCTION TO CATHOLIC CHRISTOLOGY.pptx
INTRODUCTION TO CATHOLIC CHRISTOLOGY.pptxINTRODUCTION TO CATHOLIC CHRISTOLOGY.pptx
INTRODUCTION TO CATHOLIC CHRISTOLOGY.pptxHumphrey A Beña
 
ECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptx
ECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptxECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptx
ECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptxiammrhaywood
 
Earth Day Presentation wow hello nice great
Earth Day Presentation wow hello nice greatEarth Day Presentation wow hello nice great
Earth Day Presentation wow hello nice greatYousafMalik24
 
Grade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdf
Grade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdfGrade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdf
Grade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdfJemuel Francisco
 
Field Attribute Index Feature in Odoo 17
Field Attribute Index Feature in Odoo 17Field Attribute Index Feature in Odoo 17
Field Attribute Index Feature in Odoo 17Celine George
 
AUDIENCE THEORY -CULTIVATION THEORY - GERBNER.pptx
AUDIENCE THEORY -CULTIVATION THEORY -  GERBNER.pptxAUDIENCE THEORY -CULTIVATION THEORY -  GERBNER.pptx
AUDIENCE THEORY -CULTIVATION THEORY - GERBNER.pptxiammrhaywood
 
ANG SEKTOR NG agrikultura.pptx QUARTER 4
ANG SEKTOR NG agrikultura.pptx QUARTER 4ANG SEKTOR NG agrikultura.pptx QUARTER 4
ANG SEKTOR NG agrikultura.pptx QUARTER 4MiaBumagat1
 
Incoming and Outgoing Shipments in 3 STEPS Using Odoo 17
Incoming and Outgoing Shipments in 3 STEPS Using Odoo 17Incoming and Outgoing Shipments in 3 STEPS Using Odoo 17
Incoming and Outgoing Shipments in 3 STEPS Using Odoo 17Celine George
 
USPS® Forced Meter Migration - How to Know if Your Postage Meter Will Soon be...
USPS® Forced Meter Migration - How to Know if Your Postage Meter Will Soon be...USPS® Forced Meter Migration - How to Know if Your Postage Meter Will Soon be...
USPS® Forced Meter Migration - How to Know if Your Postage Meter Will Soon be...Postal Advocate Inc.
 

Kürzlich hochgeladen (20)

ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITYISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
ISYU TUNGKOL SA SEKSWLADIDA (ISSUE ABOUT SEXUALITY
 
AMERICAN LANGUAGE HUB_Level2_Student'sBook_Answerkey.pdf
AMERICAN LANGUAGE HUB_Level2_Student'sBook_Answerkey.pdfAMERICAN LANGUAGE HUB_Level2_Student'sBook_Answerkey.pdf
AMERICAN LANGUAGE HUB_Level2_Student'sBook_Answerkey.pdf
 
Proudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptxProudly South Africa powerpoint Thorisha.pptx
Proudly South Africa powerpoint Thorisha.pptx
 
Inclusivity Essentials_ Creating Accessible Websites for Nonprofits .pdf
Inclusivity Essentials_ Creating Accessible Websites for Nonprofits .pdfInclusivity Essentials_ Creating Accessible Websites for Nonprofits .pdf
Inclusivity Essentials_ Creating Accessible Websites for Nonprofits .pdf
 
How to Add Barcode on PDF Report in Odoo 17
How to Add Barcode on PDF Report in Odoo 17How to Add Barcode on PDF Report in Odoo 17
How to Add Barcode on PDF Report in Odoo 17
 
How to do quick user assign in kanban in Odoo 17 ERP
How to do quick user assign in kanban in Odoo 17 ERPHow to do quick user assign in kanban in Odoo 17 ERP
How to do quick user assign in kanban in Odoo 17 ERP
 
ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...
ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...
ENGLISH 7_Q4_LESSON 2_ Employing a Variety of Strategies for Effective Interp...
 
Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)Influencing policy (training slides from Fast Track Impact)
Influencing policy (training slides from Fast Track Impact)
 
THEORIES OF ORGANIZATION-PUBLIC ADMINISTRATION
THEORIES OF ORGANIZATION-PUBLIC ADMINISTRATIONTHEORIES OF ORGANIZATION-PUBLIC ADMINISTRATION
THEORIES OF ORGANIZATION-PUBLIC ADMINISTRATION
 
INTRODUCTION TO CATHOLIC CHRISTOLOGY.pptx
INTRODUCTION TO CATHOLIC CHRISTOLOGY.pptxINTRODUCTION TO CATHOLIC CHRISTOLOGY.pptx
INTRODUCTION TO CATHOLIC CHRISTOLOGY.pptx
 
ECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptx
ECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptxECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptx
ECONOMIC CONTEXT - PAPER 1 Q3: NEWSPAPERS.pptx
 
Earth Day Presentation wow hello nice great
Earth Day Presentation wow hello nice greatEarth Day Presentation wow hello nice great
Earth Day Presentation wow hello nice great
 
YOUVE GOT EMAIL_FINALS_EL_DORADO_2024.pptx
YOUVE GOT EMAIL_FINALS_EL_DORADO_2024.pptxYOUVE GOT EMAIL_FINALS_EL_DORADO_2024.pptx
YOUVE GOT EMAIL_FINALS_EL_DORADO_2024.pptx
 
Grade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdf
Grade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdfGrade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdf
Grade 9 Quarter 4 Dll Grade 9 Quarter 4 DLL.pdf
 
Field Attribute Index Feature in Odoo 17
Field Attribute Index Feature in Odoo 17Field Attribute Index Feature in Odoo 17
Field Attribute Index Feature in Odoo 17
 
AUDIENCE THEORY -CULTIVATION THEORY - GERBNER.pptx
AUDIENCE THEORY -CULTIVATION THEORY -  GERBNER.pptxAUDIENCE THEORY -CULTIVATION THEORY -  GERBNER.pptx
AUDIENCE THEORY -CULTIVATION THEORY - GERBNER.pptx
 
ANG SEKTOR NG agrikultura.pptx QUARTER 4
ANG SEKTOR NG agrikultura.pptx QUARTER 4ANG SEKTOR NG agrikultura.pptx QUARTER 4
ANG SEKTOR NG agrikultura.pptx QUARTER 4
 
Incoming and Outgoing Shipments in 3 STEPS Using Odoo 17
Incoming and Outgoing Shipments in 3 STEPS Using Odoo 17Incoming and Outgoing Shipments in 3 STEPS Using Odoo 17
Incoming and Outgoing Shipments in 3 STEPS Using Odoo 17
 
Raw materials used in Herbal Cosmetics.pptx
Raw materials used in Herbal Cosmetics.pptxRaw materials used in Herbal Cosmetics.pptx
Raw materials used in Herbal Cosmetics.pptx
 
USPS® Forced Meter Migration - How to Know if Your Postage Meter Will Soon be...
USPS® Forced Meter Migration - How to Know if Your Postage Meter Will Soon be...USPS® Forced Meter Migration - How to Know if Your Postage Meter Will Soon be...
USPS® Forced Meter Migration - How to Know if Your Postage Meter Will Soon be...
 

Managing data in computational edge clouds

  • 1. Managing Data in Computational Edge Clouds ACM SIGCOMM 2017 Workshop on Mobile Edge Communications (MECOMM’2017), Los Angeles Nitinder Mohan✼ Pengyuan Zhou✼ Keerthana Govindaraj△ Jussi Kangasharju✼ ✼University of Helsinki △Robert Bosch GmbH ✼{firstname.lastname@helsinki.fi} △{keerthana.govindaraj@de.bosch.com}
  • 2. Rise of Edge Clouds Network DatacenterEdge Server User 2 Edge Cloud: Small-scale server(s) deployed at network edge to compute user data Motivation:
  • 3. Rise of Edge Clouds Network DatacenterEdge Server User Edge Cloud: Small-scale server(s) deployed at network edge to compute user data Motivation: üDecreased latency and network traffic 2
  • 4. Rise of Edge Clouds Network Edge Cloud: Small-scale server(s) deployed at network edge to compute user data Motivation: üDecreased latency and network traffic üComputing data of local relevance 2
  • 5. Edge-Fog Cloud Nitinder Mohan and Jussi Kangasharju. "Edge-fog cloud: A distributed cloud for internet of things computations." Cloudification of the Internet of Things (CIoT). IEEE, 2016. • Voluntary compute devices • 1-2 network hops from users • Ad-hoc network connectivity • Dedicated compute servers • 5-6 network hops from users • Dense network connectivity • Permanent data archival • No compute capability 3
  • 6. Data Availability in Edge Cache Query: Give me a route from point A to B without road disruptions 4
  • 7. Data Availability in Edge Cache • Edge cloud proposals presume that the required data for processing is pre- available in Edge servers cache • However, Edge server may request Data Store for required data Data Request: Any construction/ accidents on the map? 4
  • 8. Data Availability in Edge Cache • Edge cloud proposals presume that the required data for processing is pre- available in Edge servers cache • However, Edge server may request Data Store for required data • The computation is delayed by a round-trip-time to central Data Store • Reduces benefits of employing Edge clouds! Data Request: Any construction/ accidents on the map? Requirement: Caching strategy which predicts and stores required data before computation 4
  • 9. Edge Cloud Cache Grouping Edge cloud caching mechanism ØGroup edge resources into cache groups based on their locally cached data ØPredict required data for computation based on cache grouping classification ØProvide coherence on multiple copies of shared data throughout a cache group 5
  • 12. Grouping Algorithm Deployment: Select {RC}i resources for computing a task Tc 1. Algorithms such as LPCF* selects a subset from available resources which achieves least processing and network cost for collaboration. 2. Several deployments for different tasks can occur at same time. Task Deployment T1 T2 Nitinder Mohan and Jussi Kangasharju. "Edge-fog cloud: A distributed cloud for internet of things computations." Cloudification of the Internet of Things (CIoT). IEEE, 2016. 8
  • 13. Grouping Algorithm 1. At t=0, {RC}i does not have data required for computation and it requests the required data from Data Store 2. The Data Store sends the required data to requesting resources populating their caches Task Deployment Populate Caches 9
  • 14. Grouping Algorithm 1. At t=0, {RC}i does not have data required for computation and it requests the required data from Data Store 2. The Data Store sends the required data to requesting resources populating their caches Task Deployment Populate Caches 9
  • 15. Grouping Algorithm 1. The task Tc is classified in a workload Wk where Wk is a set of similar tasks 2. The resources {RC}i caching data of workload Wk are grouped together as cache group CGk Task Deployment Populate Caches Grouping Classifiers W1 W2 10
  • 16. Grouping Algorithm 1. The deployment algorithm prefers CGk ∈Wk for task Ti ∈Wk Task Deployment Populate Caches Grouping Classifiers W1 W2 T3 11
  • 17. Grouping Algorithm 1. The deployment algorithm prefers CGk ∈Wk for task Ti ∈Wk 2. If the resources in group are not appropriate for deployment, the algorithm can select other available resources Task Deployment Populate Caches Grouping Classifiers W1 W2 11
  • 18. Grouping Algorithm 1. The deployment algorithm prefers CGk ∈Wk for task Ti ∈Wk 2. If the resources in group are not appropriate for deployment, the algorithm can select other available resources 3. These resources then join their respective cache group Task Deployment Populate Caches Grouping Classifiers W1 W2 11
  • 19. Grouping Algorithm 1. The deployment algorithm prefers CGk ∈Wk for task Ti ∈Wk 2. If the resources in group are not appropriate for deployment, the algorithm can select other available resources 3. These resources then join their respective cache group 4. A resource can be a part of more than one group Task Deployment Populate Caches Grouping Classifiers W1 W2 11
  • 21. Requirements 1. Reduce unnecessary network traffic 2. Ensure consistent computation by avoiding stale data copies 13
  • 22. Resource Roles 1. Group Member • Resource (RCx) part of a cache group {CG}i • The group member stores only data of its group workload 14
  • 23. Resource Roles 1. Group Member • Resource (RCx) part of a cache group {CG}i • The group member stores only data of its group workload 2. Group Leader • Representative of a group • Responsible for maintaining coherent data within a group • Has consistent connectivity with group members • Elected via distributed election algorithm 14
  • 24. Communication Flow The communication model is strictly pull-based to limit number of messages in the system The model operates over following assumptions: 1. The resources upload their updated data copy to Data Store after every successful computation 2. Data upload is in-parallel to ongoing local computation and does not impact the computation completion time 3. Data Store is persistent and immune to failures 4. Messages in the system are not lost 15
  • 25. Cache Data Structures 1. Group member table Maps locally cached data with group leader 2. Group leader table Ensures that group leader has consistent state and address of resource caching data 3. Data Store table Maps data to groups Data Name Leader Address Tag Data Name Leader Address Classifier Type Data Name Resource Address Tag Timestamp 16
  • 26. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario 1: Retrieving free data 17
  • 27. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? Scenario 1: Retrieving free data 17
  • 28. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? Scenario 1: Retrieving free data 17
  • 29. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? GoTo: ’Resource 1’ Scenario 1: Retrieving free data 17
  • 30. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? GoTo: ’Resource 1’ Content ‘X’? Scenario 1: Retrieving free data 17
  • 31. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? GoTo: ’Resource 1’ Content ‘X’? Scenario 1: Retrieving free data 17
  • 32. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? GoTo: ’Resource 1’ Content ‘X’? Content ‘X’ Scenario 1: Retrieving free data 17
  • 33. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? GoTo: ’Resource 1’ Content ‘X’? Content ‘X’ Scenario 1: Retrieving free data Content ’X’ 17
  • 34. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? GoTo: ’Resource 1’ Content ‘X’? Scenario 2: Resource failure 18
  • 35. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? GoTo: ’Resource 1’ Content ‘X’? Unavailable! Scenario 2: Resource failure 18
  • 36. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? GoTo: ’Resource 1’ Content ‘X’? Timeout Unavailable! Scenario 2: Resource failure 18
  • 37. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? GoTo: ’Resource 1’ Content ‘X’? Content ‘X’? Timeout Unavailable! Scenario 2: Resource failure 18
  • 38. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? GoTo: ’Resource 1’ Content ‘X’? Content ‘X’? Timeout Unavailable! Scenario 2: Resource failure 18
  • 39. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? GoTo: ’Resource 1’ Content ‘X’? Content ‘X’? Timeout Content ‘X’ Unavailable! Scenario 2: Resource failure 18
  • 40. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? GoTo: ’Resource 1’ Content ‘X’? Content ‘X’? Timeout Content ‘X’ Unavailable! Scenario 2: Resource failure Content ’X’ 18
  • 41. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Lock Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Lock Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario 3: Data under update 19
  • 42. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Lock Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Lock Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario 3: Data under update Content ‘X’? 19
  • 43. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Lock Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Lock Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario 3: Data under update Content ‘X’? 19
  • 44. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Lock Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Lock Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Lock: ’Resource 1’ Scenario 3: Data under update Content ‘X’? 19
  • 45. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Lock Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Lock Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Lock: ’Resource 1’ Scenario 3: Data under update Time Interval Content ‘X’? 19
  • 46. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Lock Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Lock Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Lock: ’Resource 1’ Scenario 3: Data under update Time Interval Content ‘X’? 19
  • 47. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Lock Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Lock: ’Resource 1’ Scenario 3: Data under update Time Interval Content ‘X’? Free 19
  • 48. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Lock Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Lock: ’Resource 1’ Scenario 3: Data under update Time Interval Free: Content ‘X’ Content ‘X’? Free 19
  • 49. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Lock Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Lock: ’Resource 1’ Scenario 3: Data under update Time Interval Free: Content ‘X’ Content ‘X’? Free 19
  • 50. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Lock: ’Resource 1’ Scenario 3: Data under update Time Interval Free: Content ‘X’ Content ‘X’? Free Free 19
  • 51. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? Lock: ’Resource 1’ Scenario 3: Data under update Time Interval Free: Content ‘X’ Content ‘X’? Free Free 19
  • 52. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? Lock: ’Resource 1’ Scenario 3: Data under update Time Interval Free: Content ‘X’ Content ‘X’? GoTo: ’Resource 1’ Free Free 19
  • 53. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? Lock: ’Resource 1’ Content ‘X’? Scenario 3: Data under update Time Interval Free: Content ‘X’ Content ‘X’? GoTo: ’Resource 1’ Free Free 19
  • 54. Content Retrival in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Lock Resource 2 Member Table Content ’X’ (stale) Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Content ‘X’? Lock: ’Resource 1’ Content ‘X’? Content ‘X’ Scenario 3: Data under update Time Interval Free: Content ‘X’ Content ‘X’? GoTo: ’Resource 1’ Free Free 19
  • 55. Content Update in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario: Update on Content ‘X’ by Resource 2 20
  • 56. Content Update in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario: Update on Content ‘X’ by Resource 2 Resource 2: Retrieve updated Content ‘X’ 20
  • 57. Content Update in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ Content ‘Q’ Leader A Leader A Free Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario: Update on Content ‘X’ by Resource 2 Resource 2: Retrieve updated Content ‘X’ 20
  • 58. Content Update in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ Content ‘Q’ Leader A Leader A Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario: Update on Content ‘X’ by Resource 2 Resource 2: Retrieve updated Content ‘X’ Lock 20
  • 59. Content Update in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ Content ‘Q’ Leader A Leader A Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario: Update on Content ‘X’ by Resource 2 Resource 2: Retrieve updated Content ‘X’ Lock: Content ‘X’ Lock 20
  • 60. Content Update in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ Content ‘Q’ Leader A Leader A Free Leader A Leader Table Content ’X’ Content ‘Y’ Free Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario: Update on Content ‘X’ by Resource 2 Resource 2: Retrieve updated Content ‘X’ Lock: Content ‘X’ Lock 20
  • 61. Content Update in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ Content ‘Q’ Leader A Leader A Free Leader A Leader Table Content ’X’ Content ‘Y’ Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario: Update on Content ‘X’ by Resource 2 Resource 2: Retrieve updated Content ‘X’ Lock: Content ‘X’ Lock Lock 20 Resource 2
  • 62. Content Update in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ Content ‘Q’ Leader A Leader A Free Leader A Leader Table Content ’X’ Content ‘Y’ Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario: Update on Content ‘X’ by Resource 2 Resource 2: Retrieve updated Content ‘X’ Lock: Content ‘X’ Processing Lock Lock 20 Resource 2
  • 63. Content Update in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ Content ‘Q’ Leader A Leader A Free Leader A Leader Table Content ’X’ Content ‘Y’ Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario: Update on Content ‘X’ by Resource 2 Resource 2: Retrieve updated Content ‘X’ Lock: Content ‘X’ Processing Free Lock 20 Resource 2
  • 64. Content Update in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ Content ‘Q’ Leader A Leader A Free Leader A Leader Table Content ’X’ Content ‘Y’ Tx Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario: Update on Content ‘X’ by Resource 2 Resource 2: Retrieve updated Content ‘X’ Lock: Content ‘X’ Processing Free: Content ‘X’ Lock Lock 20 Resource 2
  • 65. Content Update in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ Content ‘Q’ Leader A Leader A Free Leader A Leader Table Content ’X’ Content ‘Y’ Tx1 Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario: Update on Content ‘X’ by Resource 2 Resource 2: Retrieve updated Content ‘X’ Lock: Content ‘X’ Processing Free: Content ‘X’ Lock Free 20 Resource 2
  • 66. Content Update in Group Resource 1 Member Table Content ’X’ Content ‘Y’ Leader A Leader A Free Lock Resource 2 Member Table Content ’X’ Content ‘Q’ Leader A Leader A Free Leader A Leader Table Content ’X’ Content ‘Y’ Tx1 Ty Content ‘Q’ Free Lock Tq Resource 1 Resource 2Content ‘Z’ Leader B Free Data Store Data Store Table Content ’X’ Content ‘Y’ Leader A Leader B Wi WjContent ‘Z’ Content ‘Q’ Leader A Leader A Wi Wi Scenario: Update on Content ‘X’ by Resource 2 Resource 2: Retrieve updated Content ‘X’ Lock: Content ‘X’ Processing Free: Content ‘X’ Content ‘X’ Lock Free 20 Resource 2
  • 67. Requirements 1. Reduce unnecessary network traffic Ø Communication model encourages data sharing within a group and removes a single point-of-failure 2. Ensure consistent computation by avoiding stale data copies Ø Model ensures causal coherence on shared data 21
  • 69. Evaluation Evaluated on Icarus simulator* Parameters: 1. 320 Edge and Fog nodes + Data Store 2. Fog node acts as group leader 3. 96*104 content items divided in 32 workloads 4. Resource cache size capacity is 10% of overall content size *Lorenzo Saino, Ioannis Psaras, and George Pavlou. "ICARUS: a caching simulator for information centric networking (ICN)." Proceedings of the 7th International ICST conference on Simulation Tools and Techniques. ICST (Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering), 2014. 23
  • 70. Evaluation Q1. Does resource grouping offer any benefits to edge clouds? Q2. If yes, what is the appropriate size of a group? 24
  • 71. Q1. Does resource grouping offer any benefits to edge clouds? a) Effect on cache hit rate 0.02 0.04 0.06 0.08 0.1 Cache Size 0 0.2 0.4 0.6 0.8 1 CacheHitRate 4-group 4-no_group 32-group 32-no_group 25
  • 72. Q1. Does resource grouping offer any benefits to edge clouds? a) Effect on cache hit rate 0.02 0.04 0.06 0.08 0.1 Cache Size 0 0.2 0.4 0.6 0.8 1 CacheHitRate 4-group 4-no_group 32-group 32-no_group ≈2 times increase in hit rate 25
  • 73. Q1. Does resource grouping offer any benefits to edge clouds? b) Effect on latency 0.02 0.04 0.06 0.08 0.1 Cache Size 20 30 40 50 60 70 80 Latency(ms) 4-group 4-no_group 32-group 32-no_group 26
  • 74. Q1. Does resource grouping offer any benefits to edge clouds? b) Effect on latency 0.02 0.04 0.06 0.08 0.1 Cache Size 20 30 40 50 60 70 80 Latency(ms) 4-group 4-no_group 32-group 32-no_group 45% decrease in latency 26
  • 75. Q2. If yes, what is the appropriate size of a group? 0 25 50 75 100 125 150 Number of Groups 0.2 0.4 0.6 0.8 1 CacheHitRate 4 workloads 8 workloads 16 workloads 32 workloads 0 25 50 75 100 125 150 Number of Groups 20 30 40 50 Latency(ms) 4 workloads 8 workloads 16 workloads 32 workloads Maximum gains are achieved when number of cache groups = number of workloads 27
  • 76. Conclusion 1. Edge resource grouping strategy manages content in Edge caches 2. Strategy classifies content based on resource’s deployed workloads and cached content 3. The 3-step iterative algorithm maps future deployments on resources with required data locally cached 4. Our evaluations via simulation show i. Direct correlation between grouping and increased system performance ii. Number of cache groups and number of assigned workloads 28
  • 77. Thank You! Questions? nitinder.mohan@helsinki.fi CleanSky - EU FP7 Marie Curie Initial Training Network
  • 79. Industry 4.0 on Edge Industry 4.0 proposes automation in manufacturing 1. Robots such as Bosch APAS can automate tasks and work along side humans while ensuring safety 2. IoT sensor data is used in augmented reality for collaborative support and manufacturing Requirement: Time-critical and fast data processing
  • 80. Industry 4.0 on Edge Industry 4.0 proposes automation in manufacturing 1. Robots such as Bosch APAS can automate tasks and work along side humans while ensuring safety 2. IoT sensor data is used in augmented reality for collaborative support and manufacturing Requirement: Time-critical and fast data processing Solution: Implement industrial Edge clouds Network Datacenter Edge Servers
  • 81. Grouping Classifier 1. Location: location of place while constructing AR and 3D models 2. Relevance: laptop and cellphone manufacturing tasks as casing 3. Time: sensor data generation and ‘freshness’ time 4. Warning signals: broadcast of information-only or safety-related signals
  • 82. Q1. Does resource grouping offer any benefits to edge clouds? c) Effect of cache size 0.02 0.04 0.06 0.08 0.1 Cache Size 20 30 40 50 60 70 80 Latency(ms) 4-group 4-no_group 32-group 32-no_group 0.02 0.04 0.06 0.08 0.1 Cache Size 0 0.2 0.4 0.6 0.8 1 CacheHitRate 4-group 4-no_group 32-group 32-no_group Gains decrease as we increase cache size! 1. Grouping works best when cache sizes are limited and cannot store all content 2. With larger caches, resources can locally store all content!
  • 83. Q1. Does resource grouping offer any benefits to edge clouds? d) Effect of number of workloads 0.02 0.04 0.06 0.08 0.1 Cache Size 20 30 40 50 60 70 80 Latency(ms) 4-group 4-no_group 32-group 32-no_group 0.02 0.04 0.06 0.08 0.1 Cache Size 0 0.2 0.4 0.6 0.8 1 CacheHitRate 4-group 4-no_group 32-group 32-no_group Gains decrease on increasing number of workloads! Ø More workloads ≈ more unique content requests which cannot be satisfied by local caches Workloads is a uniform distribution of similar content requests