Performance evaluation method and information processing device

Patent metadata

Field	Value
Publication number	US-9712404-B2
Application number	US-201414896114-A
Country	US
Kind code	B2
Filing date	Mar 7, 2014
Priority date	Mar 7, 2014
Publication date	Jul 18, 2017
Grant date	Jul 18, 2017

Abstract

In performance estimation when the configuration of a computer system is changed, the influence, on performance, of a temporary increase in load to a computer system due to reallocation of a process request allocated to a degenerated server to another server as a result of server degeneration is taken into consideration. In an embodiment, an information processing device which evaluates the performance of the computer system estimates the performance of the computer system when operation of a server included in the computer system is stopped based on the throughput required for the server at the time of operation stop and outputs the estimated performance of the computer system.

First claim

The invention claimed is:

1. An information processing device comprising:
a network interface communicatively coupled to a plurality of computer systems and a plurality of servers via a network, wherein the plurality of computer systems includes a first computer system and the plurality of servers includes a first server;
a memory; and
a processor communicatively coupled to the network interface and the memory;
wherein the memory stores:
configuration information indicating the configuration of each of the plurality of computer systems,
expected performance information indicating the expected performance of each of the plurality of computer systems according to the configuration information , and
load-performance correspondence information indicating a performance degradation amount of each of the plurality of computer systems according to a system configuration and a load;

wherein the processor detects an occurrence of a deterioration event caused by the first server stopping and in response to the occurrence of the deterioration event, the processor:
determines an operation stop time for the first server and an evaluation target period for the first computer system based on the occurrence of the deterioration event,
acquires, from the memory, the expected performance of the first computer system according to the configuration information for the first computer,
calculates a throughput required for the first server at the operation stop time,
calculates the performance degradation amount of the first computer system according to the throughput required for the first server and the load-performance correspondence information,
calculates a response time for a process request of the first computer system based on performance estimation information, wherein the performance estimation includes the expected performance and the performance degradation amount after the occurrence of the deterioration event,
calculates an additional time required, when the response time exceeds a predetermined threshold value, and
estimates an estimated performance of the first computer system based on the performance estimation information, the response time and the additional time required, and
automatically performs a configuration change to one or more of the plurality of servers based on the estimated performance of the first computer system to obtain a target value service level objective (SLO).

2. The information processing device according to claim 1 , wherein the processor further:
calculates the throughput required for the first server at the time of the operation stop as a load amount of the first computer system increasing with reception of retransmission of a request not processed by the first server in another server included in the plurality of servers, to thereby estimate the performance of the first computer system.

3. The information processing device according to claim 2 ,
wherein the processor further:
calculates the load according to the process request to the first computer system at the time of the operation stop,
acquires the configuration of the first computer system at the time of the operation stop from the configuration information, and
calculates a process request amount to the server based on the load according to the process request and the configuration of the first computer system.

4. The information processing device according to claim 1 ,
wherein, in the first computer system, a time-out period of the process request and a re-request policy for retransmitting a same process request when the process request is timed out are set, and
the processor further:
determines whether or not the response time in the evaluation target period exceeds the time-out period, and
when the response time exceeds the time-out period, outputs information indicating that the response time is likely to be increased due to an influence of the re-request.

5. A performance evaluation method comprising:
detecting, by a processor of an information processing device an occurrence of a deterioration event caused by a first server from a plurality of servers stopping;
determining, by the processor of the information processing device, an operation stop time of the first server and an evaluation target period for a first computer system based on the occurrence of the deterioration event;
acquiring, from a memory of the information processing device, an expected performance of the first computer system based configuration information stored in the memory for the first computer,
calculating, by the processor of the information processing device, a throughput required for the first server at the operation stop time,
calculating, by the processor of the information processing device, a performance degradation amount of the first computer system after the occurrence of the deterioration event according to the throughput required for the first server and load-performance correspondence information stored in the memory,
calculating, by the processor of the information processing device, a response time for a process request based on performance estimation information, wherein the performance estimation includes the expected performance and the performance degradation amount after the occurrence of the deterioration event,
calculating, by the processor of the information processing device, an additional time required, when the response time exceeds a predetermined threshold value, and
estimating, by the processor of the information processing device, an estimated-performance of the computer system based on the performance estimation information, the response time and the additional time required; and
automatically performs a configuration change to one or more of plurality of servers to obtain a target value service level objective (SLO) based on the estimated performance of the first computer system.

Assignees

• Hitachi Ltd

Inventors

• Ozaki Nobuaki
• Iizuka Daisuke

Classifications

• G06F9/505 Primary

  considering the load · CPC title

• G06F2201/81

  Threshold · CPC title

• H04L43/06

  Generation of reports · CPC title

• G06F2209/508

  Monitor · CPC title

• H04L41/147

  for predicting network behaviour · CPC title