Maintaining manageable utilization in a system to prevent excessive queuing of system requests

Patent metadata

Field	Value
Publication number	US-10649876-B2
Application number	US-201715492748-A
Country	US
Kind code	B2
Filing date	Apr 20, 2017
Priority date	Apr 20, 2017
Publication date	May 12, 2020
Grant date	May 12, 2020

Abstract

According to one embodiment, a computer-implemented method includes obtaining, during a measurement period, a set of utilization metrics for a system; determining, whether the average number of outstanding system requests of the system lies between a current and an actual value of the maximum concurrency; in response to determining that the average number of outstanding system requests of the system does not lie between the current and actual values of the maximum concurrency, not adjusting the maximum concurrency value; in response to determining that the average number of outstanding system requests of the system does lie between the current and actual values of the maximum concurrency, assigning a value of the maximum concurrency to be the value of the average number of outstanding system requests; accessing a utilization function; solving the utilization function; and managing performance problems of the system indicated by solving the utilization function.

First claim

What is claimed is:

1. A computer program product for maintaining manageable utilization in a device, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, wherein the computer readable storage medium is not a transitory signal per se, the program instructions readable and/or executable by the device, to cause the device to perform a method comprising:
obtaining, by the device, during a measurement period, a set of utilization metrics for a system, wherein the utilization metrics for the system includes at least an average number of outstanding system requests N, an average response time R of the system, and a maximum concurrency c, the maximum concurrency c being a maximum number of concurrent requests that the system is capable of processing without using a queue;
determining, by the device, whether the average number of outstanding system requests N of the system lies between a current estimated value of the maximum concurrency c and an actual value of the maximum concurrency c;
in response to determining that the average number of outstanding system requests N of the system does not lie between the current estimated value of the maximum concurrency c and the actual value of the maximum concurrency c, not adjusting, by the device, the maximum concurrency value c;
in response to determining that the average number of outstanding system requests N of the system does lie between the current estimated value of the maximum concurrency c and the actual value of the maximum concurrency c, assigning, by the device, a value of the maximum concurrency c to be the value of the average number of outstanding system requests N;
accessing, by the device, a utilization function which satisfies a black box equation for a utilization ρ value of the system;
solving, by the device, the utilization function, wherein the utilization function satisfies a predetermined fixed point relationship; and
managing, by the device, performance problems of the system indicated by the solving of the utilization function, wherein managing performance problems includes: identifying components of the system that exhibit a high utilization and/or instructing performance of an action on components of the system that exhibit a high utilization.

2. The computer program product of claim 1 , the program instructions readable and/or executable by the device, to cause the device to perform the method comprising determining, by the device, a tipping point N tip corresponding to the average number of outstanding system requests N; and using the determined tipping point N tip when determining an estimation of the utilization ρ value of the system.

3. The computer program product of claim 2 , wherein the utilization function is defined as: x=N/N+f(x), where:
f ( x )= c ((1−(γ−γ x c +βx ))/(1−(γ−γ x c +βx ) c )) for N≤N tip ,

f ( x )= c ((1−(β−γ x+αx c ) 1/c )/(1−(β−γ x+αx c ))) for N>N tip ,

wherein 0≤x≤1; α=c 2 /(c 2 +N 2 ); β=N 2 /(c 2 +N 2 ); γ=cN/(c 2 +N 2 ) and N tip =(c+1) (c−1)/c .

4. The computer program product of claim 3 , wherein an expression N/(L+δ+1) is an initial value used for the function f(x), where δ=c−N tip and L=max(N, N tip −1).

5. The computer program product of claim 4 , wherein an initial value for f(x) includes an expression selected from a group of expressions consisting of: min(N tip /(c+1), N/c); N/max(c, N+1+H c+2 −H 3 ) and NL/(L 2 +N+δN tip ), wherein H is a harmonic number.

6. The computer program product of claim 4 , wherein at least one of the coefficients α, β and γ are applied as trigonometric functions applied to a line that extends from an origin, wherein the line has a slope of N/c, and wherein

f
⁡

(
x
)

=

c
⁡

(

(

1
-

(

v
→

*

w
→

⁡

(
x
)

)

C

1
-
z

)

/

(

1
-

(

v
->

*

w
->

⁡

(
x
)

)

C

1
-
z

)

)

,

⁢

Z
=



_N
/

max
⁡

(

N
,
Ntip

)

⁢
_



Assignees

• IBM

Inventors

• Mcnutt Bruce
• Miller Vernon W

Classifications

• G06F9/5038

  considering the execution order of a plurality of tasks, e.g. taking priority or time dependency constraints into consideration (scheduling strategies G06F9/4881 and subgroups) · CPC title

• G06F11/3452 Primary

  Performance evaluation by statistical analysis · CPC title

• H04L41/5009

  Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF] · CPC title

• G06F11/3051

  Monitoring arrangements for monitoring the configuration of the computing system or of the computing system component, e.g. monitoring the presence of processing resources, peripherals, I/O links, software programs (verification or detection of system hardware configuration G06F11/2247) · CPC title

• G06F9/4881

  Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues · CPC title