US 8,621,080 B2Granted

Accurately predicting capacity requirements for information technology resources in physical, virtual and hybrid cloud environments

AssigneeGravitant, Inc.

InventorsIlyas Iyoob, Zehra Yalcin, Manish Modh, Mohammed Farooq

Life of this patent

Estimating the term…

Loading patent…

Abstract

A method, system and computer program product for accurately predicting capacity requirements for Information Technology (IT) resources in physical, virtual, cloud and hybrid cloud environments. The business demand is translated into IT application demand, which is translated into IT architecture layer demand. Resource groups are created for each IT application, IT architecture layer and resource type combination. The utilization for each IT resource group within the IT architecture layers is computed using the created resource groups, IT application demand and IT architecture layer demand. Under or over-utilized IT resource groups may be identified using the computed utilization for each IT resource group. Upon identifying such IT resources, an optimal resource profile (e.g., number of servers) for each resource group is computed. In this manner, the utilization of IT resources can be more accurately predicted thereby allowing the IT resources to be used more effectively without over-utilizing or underutilizing these resources.

Claims33

Dependent claims

A method for accurately predicting capacity requirements for information technology resources, the method comprising:

translating business demand into information technology application demand,
wherein said business demand is translated into said information technology application demand using the following algorithm:
Step
⁢
⁢
1
⁢
:
Initialize
⁢
⁢
t
=
-
1
,
⁢
I
=
ϕ
Step
⁢
⁢
2
⁢
:
t
=
t
+
1
i
=
0
,
I
=
I
+
{
i
}
B
⁢
D
^
mti
=
curVal
m
⁡
(
1
+
growth
m
)
t
⁢
(
1
+
eventImpact
emt
)
∀
m
∈
M
indep
Step
⁢
⁢
3
⁢
:
⁢
i
=
i
+
1
,
I
=
I
+
{
i
}
B
⁢
D
^
mti
=
(
∑
m
′
∈
M
⁢
bSlope
m
⁢
⁢
m
′
⁢
B
⁢
D
^
m
′
⁢
t
⁢
⁢
i
-
1
)
+
bIntercept
m
⁢
∀
m
∈
M
dep
If
⁢
⁢
∑
m
∈
M
⁢
B
⁢
D
^
mti
=
0
,
then
⁢
⁢
go
⁢
⁢
to
⁢
⁢
step
⁢
⁢
⁢
4.
⁢
⁢
Else
⁢
⁢
⁢
repeat
⁢
⁢
step
⁢
⁢
3
,
Step
⁢
⁢
4
⁢
:
⁢
BD
mt
=
∑
i
∈
I
⁢
B
⁢
D
^
mti
If
⁢
⁢
t
+
1
>

T

,
then
⁢
⁢
STOP
.
⁢
Else
⁢
⁢
go
⁢
⁢
to
⁢
⁢
⁢
Step
⁢
⁢
2
,
wherein said M is a set of demand metrics, wherein said Mindep is a subset of demand metrics that are independent, wherein said Mdep is a subset of demand metrics that are dependent on other demand metrics, wherein said I is a set of iterations, wherein said bSlopem,m′ is a slope parameter of said m with respect to said m′, wherein said bInterceptm is an intercept parameter of said m with respect to all metrics it depends on, wherein said curValm is a current or starting value of said m, wherein said growthm is a growth rate % per time period of said m, wherein said eventImpacte,m,t is a change % of said m due to an event e in said time t, wherein said B{circumflex over (D)}mti is demand type m generated in said time t in said iteration I, and wherein said BDm,t is a resulting demand of said m in said time t;
translating said information technology application demand into information technology architecture layer demand;
creating resource groups for each information technology application, information technology architecture layer and resource type combination, wherein said resource type is a type of information technology resource used to process said information technology application demand;
and
computing, by a processor, utilization for each information technology resource group within information technology architecture layers using said created resource groups, said information technology application demand and said information technology architecture layer demand.

231 more elements

A computer program product embodied in a non-transitory computer readable storage medium for accurately predicting capacity requirements for information technology resources, the computer program product comprising the programming instructions for:

translating business demand into information technology application demand,
wherein said business demand is translated into said information technology application demand using the following algorithm:
Step
⁢
⁢
1
⁢
:
Initialize
⁢
⁢
t
=
-
1
,
⁢
I
=
ϕ
Step
⁢
⁢
2
⁢
:
t
=
t
+
1
i
=
0
,
I
=
I
+
{
i
}
B
⁢
D
^
mti
=
curVal
m
⁡
(
1
+
growth
m
)
t
⁢
(
1
+
eventImpact
emt
)
∀
m
∈
M
indep
Step
⁢
⁢
3
⁢
:
⁢
i
=
i
+
1
,
I
=
I
+
{
i
}
B
⁢
D
^
mti
=
(
∑
m
′
∈
M
⁢
bSlope
m
⁢
⁢
m
′
⁢
B
⁢
D
^
m
′
⁢
t
⁢
⁢
i
-
1
)
+
bIntercept
m
⁢
∀
m
∈
M
dep
If
⁢
⁢
∑
m
∈
M
⁢
B
⁢
D
^
mti
=
0
,
then
⁢
⁢
go
⁢
⁢
to
⁢
⁢
step
⁢
⁢
⁢
4.
⁢
⁢
Else
⁢
⁢
⁢
repeat
⁢
⁢
step
⁢
⁢
3
,
Step
⁢
⁢
4
⁢
:
⁢
BD
mt
=
∑
i
∈
I
⁢
B
⁢
D
^
mti
If
⁢
⁢
t
+
1
>

T

,
then
⁢
⁢
STOP
.
⁢
Else
⁢
⁢
go
⁢
⁢
to
⁢
⁢
⁢
Step
⁢
⁢
2
,
wherein said M is a set of demand metrics, wherein said Mindep is a subset of demand metrics that are independent, wherein said Mdep is a subset of demand metrics that are dependent on other demand metrics, wherein said I is a set of iterations, wherein said bSlopem,m′ is a slope parameter of said m with respect to said m′, wherein said bInterceptm is an intercept parameter of said m with respect to all metrics it depends on, wherein said curValm is a current or starting value of said m, wherein said growthm is a growth rate % per time period of said m, wherein said eventImpacte,m,t is a change % of said m due to an event e in said time t, wherein said B{circumflex over (D)}mti is demand type m generated in said time t in said iteration I, and wherein said BDm,t is a resulting demand of said m in said time t;
translating said information technology application demand into information technology architecture layer demand;
creating resource groups for each information technology application, information technology architecture layer and resource type combination, wherein said resource type is a type of information technology resource used to process said information technology application demand;
and
computing utilization for each information technology resource group within information technology architecture layers using said created resource groups, said information technology application demand and said information technology architecture layer demand.

231 more elements

A system, comprising:

a memory unit for storing a computer program for accurately predicting capacity requirements for information technology resources;
and
a processor coupled to said memory unit, wherein said processor, responsive to said computer program, comprises:
circuitry for translating business demand into information technology application demand,
wherein said business demand is translated into said information technology application demand using the following algorithm:
Step
⁢
⁢
1
⁢
:
Initialize
⁢
⁢
t
=
-
1
,
⁢
I
=
ϕ
Step
⁢
⁢
2
⁢
:
t
=
t
+
1
i
=
0
,
I
=
I
+
{
i
}
B
⁢
D
^
mti
=
curVal
m
⁡
(
1
+
growth
m
)
t
⁢
(
1
+
eventImpact
emt
)
∀
m
∈
M
indep
Step
⁢
⁢
3
⁢
:
⁢
i
=
i
+
1
,
I
=
I
+
{
i
}
B
⁢
D
^
mti
=
(
∑
m
′
∈
M
⁢
bSlope
m
⁢
⁢
m
′
⁢
B
⁢
D
^
m
′
⁢
t
⁢
⁢
i
-
1
)
+
bIntercept
m
⁢
∀
m
∈
M
dep
If
⁢
⁢
∑
m
∈
M
⁢
B
⁢
D
^
mti
=
0
,
then
⁢
⁢
go
⁢
⁢
to
⁢
⁢
step
⁢
⁢
⁢
4.
⁢
⁢
Else
⁢
⁢
⁢
repeat
⁢
⁢
step
⁢
⁢
3
,
Step
⁢
⁢
4
⁢
:
⁢
BD
mt
=
∑
i
∈
I
⁢
B
⁢
D
^
mti
If
⁢
⁢
t
+
1
>

T

,
then
⁢
⁢
STOP
.
⁢
Else
⁢
⁢
go
⁢
⁢
to
⁢
⁢
⁢
Step
⁢
⁢
2
,
wherein said M is a set of demand metrics, wherein said Mindep is a subset of demand metrics that are independent, wherein said Mdep is a subset of demand metrics that are dependent on other demand metrics, wherein said I is a set of iterations, wherein said bSlopem,m′ is a slope parameter of said m with respect to said m′, wherein said bInterceptm is an intercept parameter of said m with respect to all metrics it depends on, wherein said curValm is a current or starting value of said m, wherein said growthm is a growth rate % per time period of said m, wherein said eventImpacte,m,t is a change % of said m due to an event e in said time t, wherein said B{circumflex over (D)}mti is demand type m generated in said time t in said iteration I, and wherein said BDm,t is a resulting demand of said m in said time t;
circuitry for translating said information technology application demand into information technology architecture layer demand;
circuitry for creating resource groups for each information technology application, information technology architecture layer and resource type combination, wherein said resource type is a type of information technology resource used to process said information technology application demand;
and
circuitry for computing utilization for each information technology resource group within information technology architecture layers using said created resource groups, said information technology application demand and said information technology architecture layer demand.

234 more elements

A method for accurately predicting capacity requirements for information technology resources, the method comprising:

forecasting business demand by defining a set of planned or unplanned events and determining their impact on demand;
determining the relationship between said business demand and information technology demand;
translating said business demand to said information technology demand at information technology architecture layers,
wherein said business demand is translated into said information technology demand using the following algorithm:
Step
⁢
⁢
1
⁢
:
Initialize
⁢
⁢
t
=
-
1
,
⁢
I
=
ϕ
Step
⁢
⁢
2
⁢
:
t
=
t
+
1
i
=
0
,
I
=
I
+
{
i
}
B
⁢
D
^
mti
=
curVal
m
⁡
(
1
+
growth
m
)
t
⁢
(
1
+
eventImpact
emt
)
∀
m
∈
M
indep
Step
⁢
⁢
3
⁢
:
⁢
i
=
i
+
1
,
I
=
I
+
{
i
}
B
⁢
D
^
mti
=
(
∑
m
′
∈
M
⁢
bSlope
m
⁢
⁢
m
′
⁢
B
⁢
D
^
m
′
⁢
t
⁢
⁢
i
-
1
)
+
bIntercept
m
⁢
∀
m
∈
M
dep
If
⁢
⁢
∑
m
∈
M
⁢
B
⁢
D
^
mti
=
0
,
then
⁢
⁢
go
⁢
⁢
to
⁢
⁢
step
⁢
⁢
⁢
4.
⁢
⁢
Else
⁢
⁢
⁢
repeat
⁢
⁢
step
⁢
⁢
3
,
Step
⁢
⁢
4
⁢
:
⁢
BD
mt
=
∑
i
∈
I
⁢
B
⁢
D
^
mti
If
⁢
⁢
t
+
1
>

T

,
then
⁢
⁢
STOP
.
⁢
Else
⁢
⁢
go
⁢
⁢
to
⁢
⁢
⁢
Step
⁢
⁢
2
,
wherein said M is a set of demand metrics, wherein said Mindep is a subset of demand metrics that are independent, wherein said Mdep is a subset of demand metrics that are dependent on other demand metrics, wherein said I is a set of iterations, wherein said bSlopem,m′ is a slope parameter of said m with respect to said m′, wherein said bInterceptm is an intercept parameter of said m with respect to all metrics it depends on, wherein said curValm is a current or starting value of said m, wherein said growthm is a growth rate % per time period of said m, wherein said eventImpacte,m,t is a change % of said m due to an event e in said time t, wherein said B{circumflex over (D)}mti is demand type m generated in said time t in said iteration I, and wherein said BDm,t is a resulting demand of said m in said time t;
computing estimated utilization of each of said information technology architecture layers based on said information technology demand and based on an estimate of capacity at said information technology architecture layers;
identifying any capacity gaps in said information technology resources;
simulating said capacity gaps;
generating resolution options to said capacity gaps;
and
selecting, by a processor, a subset of said resolution options to minimize costs.

235 more elements

forecasting business demand by defining a set of planned or unplanned events and determining their impact on demand;
determining the relationship between said business demand and information technology demand;
translating said business demand to said information technology demand at information technology architecture layers,
wherein said business demand is translated into said information technology demand using the following algorithm:
Step
⁢
⁢
1
⁢
:
Initialize
⁢
⁢
t
=
-
1
,
⁢
I
=
ϕ
Step
⁢
⁢
2
⁢
:
t
=
t
+
1
i
=
0
,
I
=
I
+
{
i
}
B
⁢
D
^
mti
=
curVal
m
⁡
(
1
+
growth
m
)
t
⁢
(
1
+
eventImpact
emt
)
∀
m
∈
M
indep
Step
⁢
⁢
3
⁢
:
⁢
i
=
i
+
1
,
I
=
I
+
{
i
}
B
⁢
D
^
mti
=
(
∑
m
′
∈
M
⁢
bSlope
m
⁢
⁢
m
′
⁢
B
⁢
D
^
m
′
⁢
t
⁢
⁢
i
-
1
)
+
bIntercept
m
⁢
∀
m
∈
M
dep
If
⁢
⁢
∑
m
∈
M
⁢
B
⁢
D
^
mti
=
0
,
then
⁢
⁢
go
⁢
⁢
to
⁢
⁢
step
⁢
⁢
⁢
4.
⁢
⁢
Else
⁢
⁢
⁢
repeat
⁢
⁢
step
⁢
⁢
3
,
Step
⁢
⁢
4
⁢
:
⁢
BD
mt
=
∑
i
∈
I
⁢
B
⁢
D
^
mti
If
⁢
⁢
t
+
1
>

T

,
then
⁢
⁢
STOP
.
⁢
Else
⁢
⁢
go
⁢
⁢
to
⁢
⁢
⁢
Step
⁢
⁢
2
,
wherein said M is a set of demand metrics, wherein said Mindep is a subset of demand metrics that are independent, wherein said Mdep is a subset of demand metrics that are dependent on other demand metrics, wherein said I is a set of iterations, wherein said bSlopem,m′ is a slope parameter of said m with respect to said m′, wherein said bInterceptm is an intercept parameter of said m with respect to all metrics it depends on, wherein said curValm is a current or starting value of said m, wherein said growthm is a growth rate % per time period of said m, wherein said eventImpacte,m,t is a change % of said m due to an event e in said time t, wherein said B{circumflex over (D)}mti is demand type m generated in said time t in said iteration I, and wherein said BDm,t is a resulting demand of said m in said time t;
computing estimated utilization of each of said information technology architecture layers based on said information technology demand and based on an estimate of capacity at said information technology architecture layers;
identifying any capacity gaps in said information technology resources;
simulating said capacity gaps;
generating resolution options to said capacity gaps;
and
selecting a subset of said resolution options to minimize costs.

235 more elements

A system, comprising:

a memory unit for storing a computer program for accurately predicting capacity requirements for information technology resources;
and
a processor coupled to said memory unit, wherein said processor, responsive to said computer program, comprises:
circuitry for forecasting business demand by defining a set of planned or unplanned events and determining their impact on demand;
circuitry for determining the relationship between said business demand and information technology demand;
circuitry for translating said business demand to said information technology demand at information technology architecture layers,
wherein said business demand is translated into said information technology demand using the following algorithm:
Step
⁢
⁢
1
⁢
:
Initialize
⁢
⁢
t
=
-
1
,
⁢
I
=
ϕ
Step
⁢
⁢
2
⁢
:
t
=
t
+
1
i
=
0
,
I
=
I
+
{
i
}
B
⁢
D
^
mti
=
curVal
m
⁡
(
1
+
growth
m
)
t
⁢
(
1
+
eventImpact
emt
)
∀
m
∈
M
indep
Step
⁢
⁢
3
⁢
:
⁢
i
=
i
+
1
,
I
=
I
+
{
i
}
B
⁢
D
^
mti
=
(
∑
m
′
∈
M
⁢
bSlope
m
⁢
⁢
m
′
⁢
B
⁢
D
^
m
′
⁢
t
⁢
⁢
i
-
1
)
+
bIntercept
m
⁢
∀
m
∈
M
dep
If
⁢
⁢
∑
m
∈
M
⁢
B
⁢
D
^
mti
=
0
,
then
⁢
⁢
go
⁢
⁢
to
⁢
⁢
step
⁢
⁢
⁢
4.
⁢
⁢
Else
⁢
⁢
⁢
repeat
⁢
⁢
step
⁢
⁢
3
,
Step
⁢
⁢
4
⁢
:
⁢
BD
mt
=
∑
i
∈
I
⁢
B
⁢
D
^
mti
If
⁢
⁢
t
+
1
>

T

,
then
⁢
⁢
STOP
.
⁢
Else
⁢
⁢
go
⁢
⁢
to
⁢
⁢
⁢
Step
⁢
⁢
2
,
wherein said M is a set of demand metrics, wherein said Mindep is a subset of demand metrics that are independent, wherein said Mdep is a subset of demand metrics that are dependent on other demand metrics, wherein said I is a set of iterations, wherein said bSlopem,m′ is a slope parameter of said m with respect to said m′, wherein said bInterceptm is an intercept parameter of said m with respect to all metrics it depends on, wherein said curValm is a current or starting value of said m, wherein said growthm is a growth rate % per time period of said m, wherein said eventImpacte,m,t is a change % of said m due to an event e in said time t, wherein said B{circumflex over (D)}mti is demand type m generated in said time t in said iteration I, and wherein said BDm,t is a resulting demand of said m in said time t;
circuitry for computing, by a processor, estimated utilization of each of said information technology architecture layers based on said information technology demand and based on an estimate of capacity at said information technology architecture layers;
circuitry for identifying any capacity gaps in said information technology resources;
circuitry for simulating said capacity gaps;
circuitry for generating resolution options to said capacity gaps;
and
circuitry for selecting a subset of said resolution options to minimize costs.

238 more elements

Dependencies are inferred from the claim text.