Managing Transfer-Based Datacenter Connections

Adrian Asensio; Luis Velasco

doi:10.1364/JOCN.6.000660

Journal of Optical Communications and Networking
Vol. 6,
Issue 7,
pp. 660-669
(2014)
•https://doi.org/10.1364/JOCN.6.000660

Managing Transfer-Based Datacenter Connections

Adrian Asensio and Luis Velasco

Not Accessible

Your library or personal account may give you access

Get PDF
Email
Share
Get Citation
Copy Citation Text
Adrian Asensio and Luis Velasco, "Managing Transfer-Based Datacenter Connections," J. Opt. Commun. Netw. 6, 660-669 (2014)

Export Citation
- BibTex
- Endnote (RIS)
- HTML
- Plain Text
Citation alert
Save article

Abstract

Creating datacenter federations allows balancing workloads accommodating spikes in cloud demand while reducing the extensive energy consumption of datacenters. This allows datacenter operators to reduce cost and increase their revenue from using under-utilized resources. To implement those elastic operations, datacenters need to be connected through a network providing huge capacity and guaranteeing adequate resource availability when it is required. Those features can be covered by the flexgrid optical technology provided with an intelligent control plane. For the latter, in this paper we rely on the application-based network operations (ABNO) architecture currently under standardization in the IETF. To increase the availability of the desired capacity, a carrier software defined network (SDN) is placed on top of ABNO. On this architecture, cloud resource managers request data transferences using an application-oriented semantic, which includes data volume and required completion time. Those requests are transformed into connection requests by the carrier SDN. To increase resource availability for incoming requests, the carrier SDN can perform elastic operations on already established connections supporting transferences provided that the committed completion time is ensured. Requests are routed and resource allocation is scheduled, which becomes the routing and scheduled spectrum allocation (RSSA) problem. An integer linear programming model is proposed, and an algorithm for solving the RSSA problem in realistic scenarios is designed. Results showing remarkable gains in the amount of data transported motivate its use for inter-datacenter traffic.

Full Article | PDF Article

More Like This

Towards a carrier SDN: an example for elastic inter-datacenter connectivity

L. Velasco, A. Asensio, J.L. Berral, A. Castro, and V. López
Opt. Express 22(1) 55-61 (2014)

Cost-Optimized Joint Resource Allocation in Grids/Clouds With Multilayer Optical Network Architecture

Pan Yi, Hui Ding, and Byrav Ramamurthy
J. Opt. Commun. Netw. 6(10) 911-924 (2014)

Design of Light-Tree Based Optical Inter-Datacenter Networks

Rongping Lin, Moshe Zukerman, Gangxiang Shen, and Wen-De Zhong
J. Opt. Commun. Netw. 5(12) 1443-1455 (2013)

Cited By

You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Figures (7)

You do not have subscription access to this journal. Figure files are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Tables (6)

You do not have subscription access to this journal. Article tables are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Equations (7)

You do not have subscription access to this journal. Equations are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

			Time to Solve		Solution
$\| R \|$	max ( $\| A_{0} \|$ )	max ( $\| A_{1} \|$ )	ILP	Alg.	ILP	Alg.
7	691	3409	10.3 s	5 ms	1	1
8	2173	7203	15.4 s	13 ms	1	1
11	1442	3693	19.8 s	9 ms	1	1
14	1592	5845	25.9 s	31 ms	1	1
15	1447	4571	26.8 s	6 ms	1	1
17	1463	5381	19.8 s	27 ms	1	2

		TEL	BT	DT
Gain at 1% unserved bitrate		12.2%	17.8%	30.9%
Transference time in seconds (RSSA)	Max	3444	3523	3543
Transference time in seconds (RSSA)	Avg	1180	1092	1048
Transference time in seconds (RSA)	Max	2167	1822	1518
Transference time in seconds (RSA)	Avg	1085	912	761

		TEL	BT	DT
# of spectrum reductions	Max	3.5	3.6	3.9
# of spectrum reductions	Avg.	1.09	1.13	1.18
# of spectrum expansions	Max	4.1	4.8	5.6
# of spectrum expansions	Avg.	1.18	1.27	1.37
% of connections experiencing elastic operations	Red.	9.3	18.8	24.4
% of connections experiencing elastic operations	Exp.	23.6	43.5	53.4

			Time to Solve		Solution
$\| R \|$	max ( $\| A_{0} \|$ )	max ( $\| A_{1} \|$ )	ILP	Alg.	ILP	Alg.
7	691	3409	10.3 s	5 ms	1	1
8	2173	7203	15.4 s	13 ms	1	1
11	1442	3693	19.8 s	9 ms	1	1
14	1592	5845	25.9 s	31 ms	1	1
15	1447	4571	26.8 s	6 ms	1	1
17	1463	5381	19.8 s	27 ms	1	2

		TEL	BT	DT
Gain at 1% unserved bitrate		12.2%	17.8%	30.9%
Transference time in seconds (RSSA)	Max	3444	3523	3543
Transference time in seconds (RSSA)	Avg	1180	1092	1048
Transference time in seconds (RSA)	Max	2167	1822	1518
Transference time in seconds (RSA)	Avg	1085	912	761

INPUT: $r$
OUTPUT: $A_{0}$ , $A_{1}$ , $β_{01}$
1: $A_{0} \leftarrow getRectangles (r)$
2: $A_{1} \leftarrow Ø$
3: for each $a_{0} \in A_{0}$ do
4: $A_{1} \leftarrow A_{1} \cup getRectangles (r, a_{0})$
5: for each $a_{0} \in A_{0}$ do
6: for each $a_{1} \in A_{1}$ do
7: $β_{01} \leftarrow compute Feasibility (r, a_{0}, a_{1})$
8: return $A_{0}$ , $A_{1}$ , $β_{01}$

INPUT: $o$ , $d$ , $v$ , $c$
OUTPUT: route, slot
1: $b r \leftarrow getMinBitrate (v, c)$
2: ${route, slot} \leftarrow computeRSA (o, d, b r, s T E D)$
3: if $route \neq \emptyset$ then
4: $slot \leftarrow getMaxSA (route, s T E D)$
5: return {route, slot}
6: $P \leftarrow computeKSP (o, d)$
7: for each $p \in P$ do
8: $s w \leftarrow getSlotWidth (p.length, b r)$
9: $S \leftarrow getFreeSlots (p, s T E D)$
10: sort( $S$ , slot width, DESC)
11: for each $s \in S$ do
12: $Q \leftarrow getAdjacentPaths (p, s, s T E D)$
13: $R \leftarrow findSlot (Q, s, s w, s T E D)$
14: if $R \neq \emptyset$ then
15: doReschedule ( $R$ )
16: $slot \leftarrow getSA (p, b r, s T E D)$
17: return ${p, slot}$
18: for each $p \in P$ do
19: $s w \leftarrow getSlotWidth (p.length, b r)$
20: $S \leftarrow getFeasibleBorderPathSlots (p, s T E D)$
21: sort( $S$ , slot width, DESC)
22: repeat from line 11 to line 17
23: return ${\emptyset, 0}$

INPUT: $R$ , $t^{0}$
1: for each $r \in R$ do
2: if $t_{r}^{1} = t^{0}$ then
3: doReSchedule ( $r$ )
4: $sort (R, t_{r} - t^{0}, ASC)$
5: for each $r \in R$ do
6: expandMax ( $r$ )

INPUT: $r$
OUTPUT: $A_{0}$ , $A_{1}$ , $β_{01}$
1: $A_{0} \leftarrow getRectangles (r)$
2: $A_{1} \leftarrow Ø$
3: for each $a_{0} \in A_{0}$ do
4: $A_{1} \leftarrow A_{1} \cup getRectangles (r, a_{0})$
5: for each $a_{0} \in A_{0}$ do
6: for each $a_{1} \in A_{1}$ do
7: $β_{01} \leftarrow compute Feasibility (r, a_{0}, a_{1})$
8: return $A_{0}$ , $A_{1}$ , $β_{01}$

Abstract

Cited By

Figures (7)

Tables (6)

Equations (7)

Journal of Optical Communications and Networking