JP3584891B2

JP3584891B2 - Data relocation method, data relocation method, and data relocation program

Info

Publication number: JP3584891B2
Application number: JP2001059535A
Authority: JP
Inventors: 寛之大谷
Original assignee: NEC Corp
Current assignee: NEC Corp
Priority date: 2001-03-05
Filing date: 2001-03-05
Publication date: 2004-11-04
Anticipated expiration: 2021-03-05
Also published as: JP2002259061A

Description

【０００１】
【発明の属する技術分野】
本発明は、データ再配置方式、データ再配置方法およびデータ再配置用プログラムに関し、特に、Ｉ／Ｏが集中する物理ディスク中の論理ディスクを効率的に再配置することができるデータ再配置方式、データ再配置方法およびデータ再配置用プログラムに関する。
【０００２】
【従来の技術】
従来、複数の物理ディスク上の領域を分割し、分割した領域をある単位でまとめて仮想化した複数の論理ディスクを構成するディスクアレイ装置において、特定の物理ディスクにＩ／Ｏ要求が集中すると、その物理ディスクに対する後続のＩ／Ｏ要求は先行のＩ／Ｏ要求が完了するまで待たされてレスポンスが悪化する。もし、このような性能低下が装置内の一部において発生すると、そこがボトルネックとなり、装置全体の性能が低下するという問題がある。
【０００３】
このような問題を解決する技術の一例が、特開２０００−２１５００７に記載されている。特開２０００−２１５００７では、物理ディスクのビジー値を測定し、物理ディスクのビジー率が利用者の指定した値を越えた場合、その物理ディスク上の論理ディスクを選び、その論理ディスク上のデータを他の物理ディスクに転送し、論理ディスクと物理ディスクとの対応関係を変更するような論理ディスクの再配置を実施する。これにより、一部の物理ディスクにＩ／Ｏが集中する状態を解消して、ディスクアレイ装置全体の性能を改善している。
【０００４】
【発明が解決しようとする課題】
上述した従来の技術は、論理ディスクの再配置を無駄に行っているため、余計なコストをかけるという問題点がある。その理由は、論理ディスクの再配置自体にＩ／Ｏが発生するため、一旦通常の運用を止めてから再配置を実施するにしても、その間は通常の運用が出来ない。一方、通常運用の背後で再配置処理を実行すると、再配置に伴うＩ／Ｏが通常運用のＩ／Ｏを妨げる結果となる。このように再配置処理は非常にコストのかかる処理であるため、不用意な再配置の実行は抑止しなければならない。ところが、単に物理ディスクのビジー率を比較して再配置の可否を判断する従来の再配置方式では、論理ディスクを構成する物理ディスクにおいて、同じ物理ディスク上の同じ論理ディスクに対してＩ／Ｏが集中するような状況を想定した場合、物理ディスクのビジー率は高いと検出される。しかし、その論理ディスクを別の物理ディスクに再配置しても、結局Ｉ／Ｏが集中する物理ディスクが他へ移動するだけでディスクアレイ装置全体の性能は向上せず、論理ディスクの再配置自体が無駄な処理となる。
【０００５】
本発明の目的は、上記の問題点を解決し、Ｉ／Ｏが集中する物理ディスク中の論理ディスクを効率的に再配置することができる論理ディスクの再配置方式を提供することにある。
【０００６】
【課題を解決するための手段】
本発明の第１のデータ再配置方式は、少なくとも１つ以上の論理ディスクを含む物理ディスクを少なくとも１つ以上接続したコンピュータシステムにおいて、排他アクセス時間が最大である論理ディスクをビジー時間が最小である物理ディスクに移動させる論理ディスク再配置手段を備えたことを特徴とする。
【０００７】
本発明の第２のデータ再配置方式は、第１のデータ再配置方式において、前記論理ディスク毎の排他アクセス時間を調べる論理ディスク排他アクセス時間監視手段と、前記物理ディスク毎のビジー時間を調べる物理ディスクビジー時間監視手段とを備え、前記論理ディスク再配置手段は、前記論理ディスク排他アクセス時間監視手段が調べた前記論理ディスク毎の排他アクセス時間と、前記物理ディスクビジー時間監視手段が調べた前記物理ディスク毎のビジー時間とにより、排他アクセス時間が最大である論理ディスクをビジー時間が最小である物理ディスクに移動させることを特徴とする。
【０００８】
本発明の第３のデータ再配置方式は、第２のデータ再配置方式において、前記物理ディスクのそれぞれについて、ビジー状態となっている時間の累積値を示すビジー時間と、ビジー状態となった時刻を示すビジー状態開始時間とを記憶するとともに、前記物理ディスクのそれぞれに属する前記論理ディスクのそれぞれについて、処理待ちまたは処理中の状態にあるＩＯ要求の件数を示すＩＯカウンタと、排他アクセス状態となった時刻を示す排他アクセス開始時刻と、排他アクセス状態となっている時間の累積値を示す排他アクセス時間とを記憶するＩＯカウンタテーブルを備え、前記論理ディスク排他アクセス時間監視手段は、少なくとも論理ディスク番号を含むＩＯ開始通知を受信した場合は、前記ＩＯカウンタテーブルの当該論理ディスク番号に対応する論理ディスクのＩＯカウンタを１だけインクリメントするとともに、前記ＩＯカウンタが２以上であるか否かを確認し、前記ＩＯカウンタが２より小さいときには、前記論理ディスクが所属する物理ディスク内の他の論理ディスクのＩＯカウンタが全てゼロであるかを確認し、前記ＩＯカウンタの全てがゼロではないときには、前記ＩＯカウンタテーブル内の前記論理ディスクの排他アクセス開始時刻に現在時刻を記録するとともに、ＩＯカウンタが１以上である他の論理ディスクの数を調べ、一つだけある場合には、当該他の論理ディスクの排他アクセス開始時刻に現在時刻を記録し、少なくとも論理ディスク番号を含むＩＯ完了通知を受信した場合には、前記ＩＯカウンタテーブルを参照して、当該論理ディスク番号に対応する論理ディスクのＩＯカウンタが２以上であるか否かを確認し、前記ＩＯカウンタが２より小さいときには、前記論理ディスクが所属する物理ディスク内の他の論理ディスクのＩＯカウンタが全てゼロであるかを確認し、前記ＩＯカウンタの全てがゼロではないときは、現在時刻から前記ＩＯカウンタテーブル内の前記論理ディスクの排他アクセス開始時刻を減算した値を前記論理ディスクの排他アクセス時間に加算するとともに、ＩＯカウンタが１以上である他の論理ディスクの数を調べ、一つだけある場合には、現在時刻から前記ＩＯカウンタテーブル内の当該他の論理ディスクの排他アクセス開始時刻を減算した値を当該他の論理ディスクの排他アクセス時間に加算し、前記論理ディスクのＩＯカウンタを１だけデクリメントし、物理ディスクビジー時間監視手段は、少なくとも物理ディスク番号を含むＩＯ開始通知を受信した場合は、前記ＩＯカウンタテーブルを参照して、当該物理ディスク番号に対応する物理ディスクに属する全ての論理ディスクのＩＯカウンタがゼロであるかを確認し、当該ＩＯカウンタが全てゼロである場合には、前記ＩＯカウンタテーブル内の前記物理ディスクのビジー状態開始時刻に現在時刻を記録し、少なくとも物理ディスク番号を含むＩＯ完了通知を受信した場合には、前記ＩＯカウンタテーブルを参照して、当該物理ディスク番号に対応する物理ディスクに属する論理ディスクのＩＯカウンタが全てゼロであるかを確認し、当該ＩＯカウンタが全てゼロである場合には、現在時刻から前記ＩＯカウンタテーブル内の前記物理ディスクのビジー状態開始時刻を減算した値を前記物理ディスクのビジー時間に加算し、少なくとも論理ディスク番号を含む論理ディスクへのＩＯ要求を受信すると、当該論理ディスク番号を少なくとも含む前記ＩＯ開始通知を前記論理ディスク開始手段に送信するとともに、当該論理ディスクが所属する物理ディスクの物理ディスク番号を少なくとも含むＩＯ開始通知を前記物理ディスクビジー時間監視手段に送信し、前記物理ディスクのアクセス後、前記物理ディスク番号を少なくとも含むＩＯ完了通知を前記物理ディスクビジー時間監視手段に送信するとともに、前記論理ディスク番号を少なくとも含むＩＯ完了通知を前記論理ディスク開始手段に送信する論理ディスク管理手段を備え、前記論理ディスク再配置手段は、前記ＩＯカウンタテーブルを参照して排他アクセス時間が最大である論理ディスクを調べるとともに、ビジー時間が最小である物理ディスクを調べ、当該論理ディスクを当該物理ディスクに移動させることを特徴とする。
【０００９】
本発明の第４のデータ再配置方式は、第３のデータ再配置方式において、前記物理ディスクと前記論理ディスクの対応関係を記憶する論理ディスク管理テーブルを備え、前記論理ディスク排他アクセス時間監視手段は、前記論理ディスク管理テーブルを参照して、前記論理ディスクが所属する物理ディスク調べ、前記物理ディスクビジー時間監視手段は、前記論理ディスク管理テーブルを参照して、前記物理ディスクに属する論理ディスクを調べ、前記論理ディスク再配置手段は、排他アクセス時間が最大である論理ディスクをビジー時間が最小である物理ディスクに移動させた後に、前記論理ディスク管理テーブル内の当該論理ディスクと当該物理ディスクとの対応関係を変更することを特徴とする。
【００１０】
本発明の第１のデータ再配置方法は、少なくとも１つ以上の論理ディスクを含む物理ディスクを少なくとも１つ以上接続したコンピュータシステムにおいて、コンピュータが、排他アクセス時間が最大である論理ディスクをビジー時間が最小である物理ディスクに移動させる論理ディスク再配置ステップを含むことを特徴とする。
【００１１】
本発明の第２のデータ再配置方法は、第１のデータ再配置方法において、前記コンピュータが、前記論理ディスク毎の排他アクセス時間を調べる論理ディスク排他アクセス時間監視ステップと、前記コンピュータが、前記物理ディスク毎のビジー時間を調べる物理ディスクビジー時間監視ステップとを含み、前記論理ディスク再配置ステップは、前記論理ディスク排他アクセス時間監視ステップで調べた前記論理ディスク毎の排他アクセス時間と、前記物理ディスクビジー時間監視ステップで調べた前記物理ディスク毎のビジー時間とにより、排他アクセス時間が最大である論理ディスクをビジー時間が最小である物理ディスクに移動させることを特徴とする。
【００１２】
本発明の第３のデータ再配置方法は、第２のデータ再配置方法において、前記論理ディスク排他アクセス時間監視ステップは、少なくとも論理ディスク番号を含むＩＯ開始通知を受信した場合は、前記物理ディスクのそれぞれについて、ビジー状態となっている時間の累積値を示すビジー時間と、ビジー状態となった時刻を示すビジー状態開始時間とを記憶するとともに、前記物理ディスクのそれぞれに属する前記論理ディスクのそれぞれについて、処理待ちまたは処理中の状態にあるＩＯ要求の件数を示すＩＯカウンタと、排他アクセス状態となった時刻を示す排他アクセス開始時刻と、排他アクセス状態となっている時間の累積値を示す排他アクセス時間とを記憶するＩＯカウンタテーブルの当該論理ディスク番号に対応する論理ディスクのＩＯカウンタを１だけインクリメントするとともに、前記ＩＯカウンタが２以上であるか否かを確認し、前記ＩＯカウンタが２より小さいときには、前記論理ディスクが所属する物理ディスク内の他の論理ディスクのＩＯカウンタが全てゼロであるかを確認し、前記ＩＯカウンタの全てがゼロではないときには、前記ＩＯカウンタテーブル内の前記論理ディスクの排他アクセス開始時刻に現在時刻を記録するとともに、ＩＯカウンタが１以上である他の論理ディスクの数を調べ、一つだけある場合には、当該他の論理ディスクの排他アクセス開始時刻に現在時刻を記録し、少なくとも論理ディスク番号を含むＩＯ完了通知を受信した場合には、前記ＩＯカウンタテーブルを参照して、当該論理ディスク番号に対応する論理ディスクのＩＯカウンタが２以上であるか否かを確認し、前記ＩＯカウンタが２より小さいときには、前記論理ディスクが所属する物理ディスク内の他の論理ディスクのＩＯカウンタが全てゼロであるかを確認し、前記ＩＯカウンタの全てがゼロではないときは、現在時刻から前記ＩＯカウンタテーブル内の前記論理ディスクの排他アクセス開始時刻を減算した値を前記論理ディスクの排他アクセス時間に加算するとともに、ＩＯカウンタが１以上である他の論理ディスクの数を調べ、一つだけある場合には、現在時刻から前記ＩＯカウンタテーブル内の当該他の論理ディスクの排他アクセス開始時刻を減算した値を当該他の論理ディスクの排他アクセス時間に加算し、前記論理ディスクのＩＯカウンタを１だけデクリメントし、前記物理ディスクビジー時間監視ステップは、少なくとも物理ディスク番号を含むＩＯ開始通知を受信した場合は、前記ＩＯカウンタテーブルを参照して、当該物理ディスク番号に対応する物理ディスクに属する全ての論理ディスクのＩＯカウンタがゼロであるかを確認し、当該ＩＯカウンタが全てゼロである場合には、前記ＩＯカウンタテーブル内の前記物理ディスクのビジー状態開始時刻に現在時刻を記録し、少なくとも物理ディスク番号を含むＩＯ完了通知を受信した場合には、前記ＩＯカウンタテーブルを参照して、当該物理ディスク番号に対応する物理ディスクに属する論理ディスクのＩＯカウンタが全てゼロであるかを確認し、当該ＩＯカウンタが全てゼロである場合には、現在時刻から前記ＩＯカウンタテーブル内の前記物理ディスクのビジー状態開始時刻を減算した値を前記物理ディスクのビジー時間に加算し、少なくとも論理ディスク番号を含む論理ディスクへのＩＯ要求を受信すると、当該論理ディスク番号を少なくとも含む前記ＩＯ開始通知を前記論理ディスク開始ステップに送信するとともに、当該論理ディスクが所属する物理ディスクの物理ディスク番号を少なくとも含むＩＯ開始通知を前記物理ディスクビジー時間監視ステップに送信し、前記物理ディスクのアクセス後、前記物理ディスク番号を少なくとも含むＩＯ完了通知を前記物理ディスクビジー時間監視ステップに送信するとともに、前記論理ディスク番号を少なくとも含むＩＯ完了通知を前記論理ディスク開始ステップに送信する論理ディスク管理ステップを含み、前記論理ディスク再配置ステップは、前記ＩＯカウンタテーブルを参照して排他アクセス時間が最大である論理ディスクを調べるとともに、ビジー時間が最小である物理ディスクを調べ、当該論理ディスクを当該物理ディスクに移動させることを特徴とする。
【００１３】
本発明の第４のデータ再配置方法は、第３のデータ再配置方法において、前記論理ディスク排他アクセス時間監視ステップは、前記物理ディスクと前記論理ディスクの対応関係を記憶する論理ディスク管理テーブルを参照して、前記論理ディスクが所属する物理ディスク調べ、前記物理ディスクビジー時間監視ステップは、前記論理ディスク管理テーブルを参照して、前記物理ディスクに属する論理ディスクを調べ、前記論理ディスク再配置ステップは、排他アクセス時間が最大である論理ディスクをビジー時間が最小である物理ディスクに移動させた後に、前記論理ディスク管理テーブル内の当該論理ディスクと当該物理ディスクとの対応関係を変更することを特徴とする。
【００１４】
本発明の第１のプログラムは、少なくとも１つ以上の論理ディスクを含む物理ディスクを少なくとも１つ以上接続したコンピュータシステムにおいて、排他アクセス時間が最大である論理ディスクをビジー時間が最小である物理ディスクに移動させる論理ディスク再配置処理をコンピュータに実行させることを特徴とする。
【００１５】
本発明の第２のプログラムは、第１のプログラムにおいて、前記論理ディスク毎の排他アクセス時間を調べる論理ディスク排他アクセス時間監視処理と、前記物理ディスク毎のビジー時間を調べる物理ディスクビジー時間監視処理とを含み、前記論理ディスク再配置処理は、前記論理ディスク排他アクセス時間監視処理で調べた前記論理ディスク毎の排他アクセス時間と、前記物理ディスクビジー時間監視処理で調べた前記物理ディスク毎のビジー時間とにより、排他アクセス時間が最大である論理ディスクをビジー時間が最小である物理ディスクに移動させることを特徴とする。
【００１６】
本発明の第３のプログラムは、第２のプログラムにおいて、前記論理ディスク排他アクセス時間監視処理は、少なくとも論理ディスク番号を含むＩＯ開始通知を受信した場合は、前記物理ディスクのそれぞれについて、ビジー状態となっている時間の累積値を示すビジー時間と、ビジー状態となった時刻を示すビジー状態開始時間とを記憶するとともに、前記物理ディスクのそれぞれに属する前記論理ディスクのそれぞれについて、処理待ちまたは処理中の状態にあるＩＯ要求の件数を示すＩＯカウンタと、排他アクセス状態となった時刻を示す排他アクセス開始時刻と、排他アクセス状態となっている時間の累積値を示す排他アクセス時間とを記憶するＩＯカウンタテーブルの当該論理ディスク番号に対応する論理ディスクのＩＯカウンタを１だけインクリメントするとともに、前記ＩＯカウンタが２以上であるか否かを確認し、前記ＩＯカウンタが２より小さいときには、前記論理ディスクが所属する物理ディスク内の他の論理ディスクのＩＯカウンタが全てゼロであるかを確認し、前記ＩＯカウンタの全てがゼロではないときには、前記ＩＯカウンタテーブル内の前記論理ディスクの排他アクセス開始時刻に現在時刻を記録するとともに、ＩＯカウンタが１以上である他の論理ディスクの数を調べ、一つだけある場合には、当該他の論理ディスクの排他アクセス開始時刻に現在時刻を記録し、少なくとも論理ディスク番号を含むＩＯ完了通知を受信した場合には、前記ＩＯカウンタテーブルを参照して、当該論理ディスク番号に対応する論理ディスクのＩＯカウンタが２以上であるか否かを確認し、前記ＩＯカウンタが２より小さいときには、前記論理ディスクが所属する物理ディスク内の他の論理ディスクのＩＯカウンタが全てゼロであるかを確認し、前記ＩＯカウンタの全てがゼロではないときは、現在時刻から前記ＩＯカウンタテーブル内の前記論理ディスクの排他アクセス開始時刻を減算した値を前記論理ディスクの排他アクセス時間に加算するとともに、ＩＯカウンタが１以上である他の論理ディスクの数を調べ、一つだけある場合には、現在時刻から前記ＩＯカウンタテーブル内の当該他の論理ディスクの排他アクセス開始時刻を減算した値を当該他の論理ディスクの排他アクセス時間に加算し、前記論理ディスクのＩＯカウンタを１だけデクリメントし、前記物理ディスクビジー時間監視処理は、少なくとも物理ディスク番号を含むＩＯ開始通知を受信した場合は、前記ＩＯカウンタテーブルを参照して、当該物理ディスク番号に対応する物理ディスクに属する全ての論理ディスクのＩＯカウンタがゼロであるかを確認し、当該ＩＯカウンタが全てゼロである場合には、前記ＩＯカウンタテーブル内の前記物理ディスクのビジー状態開始時刻に現在時刻を記録し、少なくとも物理ディスク番号を含むＩＯ完了通知を受信した場合には、前記ＩＯカウンタテーブルを参照して、当該物理ディスク番号に対応する物理ディスクに属する論理ディスクのＩＯカウンタが全てゼロであるかを確認し、当該ＩＯカウンタが全てゼロである場合には、現在時刻から前記ＩＯカウンタテーブル内の前記物理ディスクのビジー状態開始時刻を減算した値を前記物理ディスクのビジー時間に加算し、少なくとも論理ディスク番号を含む論理ディスクへのＩＯ要求を受信すると、当該論理ディスク番号を少なくとも含む前記ＩＯ開始通知を前記論理ディスク開始処理に送信するとともに、当該論理ディスクが所属する物理ディスクの物理ディスク番号を少なくとも含むＩＯ開始通知を前記物理ディスクビジー時間監視処理に送信し、前記物理ディスクのアクセス後、前記物理ディスク番号を少なくとも含むＩＯ完了通知を前記物理ディスクビジー時間監視処理に送信するとともに、前記論理ディスク番号を少なくとも含むＩＯ完了通知を前記論理ディスク開始処理に送信する論理ディスク管理処理を含み、前記論理ディスク再配置処理は、前記ＩＯカウンタテーブルを参照して排他アクセス時間が最大である論理ディスクを調べるとともに、ビジー時間が最小である物理ディスクを調べ、当該論理ディスクを当該物理ディスクに移動させることを特徴とする。
【００１７】
本発明の第４のプログラムは、第３のプログラムにおいて、前記論理ディスク排他アクセス時間監視処理は、前記物理ディスクと前記論理ディスクの対応関係を記憶する論理ディスク管理テーブルを参照して、前記論理ディスクが所属する物理ディスク調べ、前記物理ディスクビジー時間監視処理は、前記論理ディスク管理テーブルを参照して、前記物理ディスクに属する論理ディスクを調べ、前記論理ディスク再配置処理は、排他アクセス時間が最大である論理ディスクをビジー時間が最小である物理ディスクに移動させた後に、前記論理ディスク管理テーブル内の当該論理ディスクと当該物理ディスクとの対応関係を変更することを特徴とする。
【００１８】
【発明の実施の形態】
次に、本発明について図面を参照して詳細に説明する。
【００１９】
まず、本発明の第１の実施の形態について詳細に説明する。
【００２０】
図１を参照すると、本発明の第１の実施の形態は、コンピュータ１００と、物理ディスク群２００と、論理ディスク管理テーブル３００と、ＩＯカウンタテーブル４００とから構成され、コンピュータ１００は、プログラム制御により動作する。
【００２１】
コンピュータ１００は、論理ディスク管理手段１０１と、物理ディスク管理手段１０２と、論理ディスク排他アクセス時間監視手段１０３と、物理ディスクビジー時間監視手段１０４と、論理ディスク再配置手段１０５とを含む。
【００２２】
物理ディスク群２００は、複数の物理ディスク２１０，２２０，２３０，２４０を含み、当該物理ディスク上の領域を分割し、各領域を論理的なデータ格納域として割り当てた論理ディスク２１１，２１２，２１３、論理ディスク２２１，２２２，２２３、論理ディスク２３１，２３２，２３３、論理ディスク２４１，２４２，２４３からなる。
【００２３】
論理ディスク管理テーブル３００は、物理ディスク上の分割した領域と論理ディスクとの対応関係を記憶する。
【００２４】
ＩＯカウンタテーブル４００は、物理ディスク２１０〜２４０のそれぞれについて、ビジー状態となっている時間の累積値を示すビジー時間と、ビジー状態となった時刻を示すビジー状態開始時間とを記憶するとともに、物理ディスク２１０〜２４０のそれぞれに属する論理ディスク２１１〜２４３のそれぞれについて、処理待ちまたは処理中の状態にあるＩＯ要求の件数を示すＩＯカウンタと、排他アクセス状態となった時刻を示す排他アクセス開始時刻と、排他アクセス状態となっている時間の累積値を示す排他アクセス時間とを記憶する。
【００２５】
論理ディスク管理手段１０１は、論理ディスク番号の指定による論理ディスク２１１〜２４３に対するＩ／Ｏ要求を受け、論理ディスク管理テーブル３００を参照して当該論理ディスク番号に対応する物理ディスク番号を取得するとともに、当該物理ディスク番号を指定して物理ディスク管理手段１０２にＩ／Ｏ要求する。
【００２６】
物理ディスク管理手段１０２は、物理ディスク２１０〜２４０に対するＩ／Ｏ要求を受け、要求された物理ディスク２１０〜２４０へＩ／Ｏを発行する。
【００２７】
論理ディスク排他アクセス時間監視手段１０３は、Ｉ／Ｏ開始時とＩ／Ｏ終了時に呼び出され、論理ディスクの排他アクセス状態を監視し、ＩＯカウンタテーブル４００内の論理ディスクの排他アクセス時間等を記録する。
【００２８】
物理ディスクビジー時間監視手段１０４は、Ｉ／Ｏ開始時とＩ／Ｏ終了時に呼び出され、物理ディスクに対するＩ／Ｏ発行状態を監視し、ＩＯカウンタテーブル４００内の物理ディスクのビジー時間等を記録する。
【００２９】
論理ディスク再配置手段１０５は、再配置の指示を受け、ＩＯカウンタテーブル４００を参照して再配置を行うべき移動対象論理ディスクと移動先の物理ディスクを決定し、再配置の可否を判断した後、論理ディスク上のデータを移動先の物理ディスク上の空き領域に転送し、論理ディスク管理テーブル３００上の物理ディスクと論理ディスクとの対応関係を更新する。
【００３０】
次に、図１〜図６を参照して本実施の形態の全体の動作について詳細に説明する。
【００３１】
図２は、論理ディスク管理手段１０１の動作を説明するフローチャートである。
【００３２】
まず、論理ディスク管理手段１０１は、論理ディスク番号が指定されたＩ／Ｏ要求を受け取ると、論理ディスク管理テーブル３００を参照して、当該論理ディスク番号を対応する物理ディスク番号に変換する（Ｓ２０１）。
【００３３】
次に、論理ディスク管理手段１０１は、論理ディスク排他アクセス時間監視手段１０３に、論理ディスクのＩ／Ｏ処理を開始したことを論理ディスク番号によって通知する（Ｓ２０２）。
【００３４】
次に、論理ディスク管理手段１０１は、物理ディスクビジー時間監視手段１０４に、物理ディスクのＩ／Ｏ処理を開始したことを物理ディスク番号によって通知する（Ｓ２０３）。
【００３５】
次に、論理ディスク管理手段１０１は、物理ディスク管理手段１０２に対して、物理ディスク番号を指定することにより、前記物理ディスクに対するＩ／Ｏ要求を行う（Ｓ２０４）。
【００３６】
さらに、論理ディスク管理手段１０１は、Ｉ／Ｏ要求をした後、物理ディスク管理手段１０２からＩ／Ｏ要求の応答が返却されるまで待ち合わせる（Ｓ２０５）。
【００３７】
次に、論理ディスク管理手段１０１は、Ｉ／Ｏ要求の応答が返却された後、物理ディスクビジー時間監視手段１０４に、物理ディスクのＩ／Ｏが完了したことを物理ディスク番号によって通知する（Ｓ２０６）。
【００３８】
次に、論理ディスク管理手段１０１は、論理ディスク排他アクセス時間監視手段１０３に、論理ディスクのＩ／Ｏが完了したことを論理ディスク番号によって通知する（Ｓ２０７）。
図３は、物理ディスク管理手段１０２の動作を示すフローチャートである。
【００３９】
まず、物理ディスク管理手段１０２は、論理ディスク管理手段１０１から、物理ディスク番号の指定による物理ディスクに対するＩ／Ｏ要求を受け取ると、指定された物理ディスクに対してＩ／Ｏを発行し（Ｓ２１１）、発行したＩ／Ｏが完了するまで待ち合わせる（Ｓ２１２）。
【００４０】
次に、物理ディスク管理手段１０２は、物理ディスクに対するＩ／Ｏが完了した後、論理ディスク管理手段１０１にＩ／Ｏ要求に対する応答を返却する（Ｓ２１３）。
【００４１】
図４は、論理ディスク排他アクセス時間監視手段１０３の動作を示すフローチャートである。
【００４２】
まず、論理ディスク排他アクセス時間監視手段１０３は、論理ディスク管理手段１０１から、Ｉ／Ｏ開始時に論理ディスク番号の通知を受けると、ＩＯカウンタテーブル４００内の当該論理ディスク番号に対応する論理ディスクのＩＯカウンタをインクリメントする（Ｓ２２１）とともに、当該ＩＯカウンタが２以上であるか否かを確認する（Ｓ２２２）。
当該ＩＯカウンタが２より小さい場合は、当該論理ディスクに対する最初のＩ／Ｏ要求であると判断し、さらに当該論理ディスクが所属する物理ディスク内の他の論理ディスクのＩＯカウンタが全てゼロであるか確認する（Ｓ２２３）。
ＩＯカウンタの全てがゼロではない場合は、他の論理ディスクに対する先行のＩ／Ｏ要求を処理しているときに後続のＩ／Ｏ要求が到着したことになり、当該後続のＩ／Ｏ要求の対象となる論理ディスクは、排他制御が物理ディスク単位に行われるため、排他アクセス状態にあると判断し、ＩＯカウンタテーブル４００内の当該論理ディスクの排他アクセス開始時刻に現在時刻を記録する（Ｓ２２４）。
【００４３】
さらに、ＩＯカウンタが１以上である他の論理ディスクが一つだけであるかを確認する（Ｓ２２５）。
【００４４】
一つだけある場合は、当該他の論理ディスクは今回のＩ／Ｏ要求の到着によって排他アクセス状態に移行したと判断し、当該他の論理ディスクの排他アクセス開始時刻に現在時刻を記録する（Ｓ２２６）。
以後、物理ディスク管理手段１０２からのＩ／Ｏ完了通知を待ち合わせ（Ｓ２２７）、論理ディスク番号が指定されたＩ／Ｏ完了通知を受け取ると、ＩＯカウンタテーブル４００を参照して、当該論理ディスクのＩＯカウンタが２以上であるか否かを確認する（Ｓ２２８）。
当該ＩＯカウンタが２より小さい場合は、このＩ／Ｏ完了通知によって、当該論理ディスクに対するＩ／Ｏ要求が全て完了したと判断し、さらに当該論理ディスクが所属する物理ディスク内の他の論理ディスクのＩＯカウンタが全てゼロであるか確認する（Ｓ２２９）。
【００４５】
ＩＯカウンタの全てがゼロではない場合は、このＩ／Ｏ要求の完了によって当該論理ディスクの排他アクセス状態が終了したと判断し、現在時刻からＩＯカウンタテーブル４００内の当該論理ディスクの排他アクセス開始時刻を減算した値即ち排他アクセス状態が継続した時間を当該論理ディスクの排他アクセス時間に加算する（Ｓ２３０）。
【００４６】
さらに、ＩＯカウンタが１以上である他の論理ディスクが一つだけであるかを確認する（Ｓ２３１）。
一つだけある場合は、この論理ディスクのＩ／Ｏ完了によって、当該他の論理ディスクに対するＩ／Ｏ要求は排他アクセス状態ではなくなるため、現在時刻からＩＯカウンタテーブル４００内の当該他の論理ディスクの排他アクセス開始時刻を減算した値を当該他の論理ディスクの排他アクセス時間に加算する（Ｓ２３２）。
最後に、Ｉ／Ｏが完了した論理ディスクのＩＯカウンタをデクリメントする（Ｓ２３３）。
図５は、物理ディスクビジー時間監視手段１０４の動作を示すフローチャートである。
まず、物理ディスクビジー時間監視手段１０４は、論理ディスク管理手段１０１からＩ／Ｏ開始時に物理ディスク番号の通知を受けると、ＩＯカウンタテーブル４００を参照して、当該物理ディスクに属する全ての論理ディスクのＩＯカウンタがゼロであるか確認する（Ｓ２４１）。
ＩＯカウンタが全てゼロである場合は、当該物理ディスクはアイドル状態からビジー状態に移行したと判断し、ＩＯカウンタテーブル４００内の当該物理ディスクのビジー状態開始時刻に現在時刻を記録する（Ｓ２４２）。
以後、Ｉ／Ｏ完了を待ち合わせ（Ｓ２４３）、Ｉ／Ｏ完了通知を受け取ると、ＩＯカウンタテーブル４００を参照して、当該物理ディスクに属する論理ディスクのＩＯカウンタが全てゼロか否かを確認し（Ｓ２４４）、ＩＯカウンタが全てゼロの場合はこのＩ／Ｏ完了によって、対象物理ディスクはビジー状態からアイドル状態へ移行したと判断し、現在時刻からＩＯカウンタテーブル４００内の当該物理ディスクのビジー状態開始時刻を減算した値をビジー状態が継続した時間としてビジー時間に加算する（Ｓ２４５）。
図６は、論理ディスク再配置手段１０５の動作を示すフローチャートである。
まず、論理ディスク再配置手段１０５は、再配置実行の指示を受けて呼び出されると、ＩＯカウンタテーブル４００を参照し、全ての論理ディスクに関する排他アクセス時間を調査して、最も大きい排他アクセス時間を示す論理ディスクを移動対象論理ディスクとして決定する（Ｓ２６１）。
次に、全ての物理ディスクに関するビジー時間を調査して、決定した移動対象論理ディスクを格納するだけの十分な空き領域があり、かつ、最も小さいビジー時間である物理ディスクを移動先物理ディスクとして決定する（Ｓ２６２）。
次に、当該移動先物理ディスク上に当該移動対象論理ディスクを移動した場合に再配置の効果があるかどうか判定する（Ｓ２６３）。
例えば、決定した移動対象論理ディスクの排他アクセス時間が利用者によって予め定められた閾値よりも大きい場合であって、移動先の物理ディスクにおけるアイドル時間（Ｉ／Ｏを一切発行していない時間。実観測時間からビジー時間を減算した値で求める）と、移動対象論理ディスクの排他アクセス時間との比が利用者によって定められた規定値よりも大きなときには再配置の効果があると判定する方法などがある。上記の方法によって再配置の効果がないと判定した場合は、再配置処理は実行せず、再配置の効果があると判定した場合には、決定した移動先物理ディスク上に新たな論理ディスクを構成し（Ｓ２６４）、移動対象論理ディスク上のデータを新たな論理ディスクへ転送する（Ｓ２６５）。
【００４７】
尚、Ｓ２６３で再配置の効果があるかどうかを判定しているが、判定を行わずにＳ２６４からＳ２６５の再配置処理を実行してもよい。
【００４８】
次に、論理ディスク管理テーブル上の移動元の論理ディスク番号を新たな論理ディスクの領域を示すように書き換える（Ｓ２６６）。
【００４９】
最後に、移動元の論理ディスクの領域を空き領域として解放する（Ｓ２６７）。
【００５０】
次に、本発明の第２の実施の形態について図面を参照して説明する。
【００５１】
本発明の第２の実施の形態は、データ再配置用プログラムを備える。データ再配置用プログラムは、コンピュータ１００の動作を制御し、コンピュータ１００に上述の手段、すなわち第１の実施の形態における手段と同一の手段を実行させる。
【００５２】
【発明の効果】
以上説明したように、本発明の論理ディスク再配置手段によれば、Ｉ／Ｏが集中する物理ディスク中の論理ディスクを効率的に再配置することができるという効果がある。
【図面の簡単な説明】
【図１】本発明の第１の実施の形態の構成を示すブロック図である。
【図２】本発明の第１の実施の形態の論理ディスク管理手段の動作を示すフローチャートである。
【図３】本発明の第１の実施の形態の物理ディスク管理手段の動作を示すフローチャートである。
【図４】本発明の第１の実施の形態の論理ディスク排他アクセス時間監視手段の動作を示すフローチャートである。
【図５】本発明の第１の実施の形態の物理ディスクビジー時間監視手段の動作を示すフローチャートである。
【図６】本発明の第１の実施の形態の論理ディスク再配置手段の動作を示すフローチャートである。
【符号の説明】
１００コンピュータ
１０１論理ディスク管理手段
１０２物理ディスク管理手段
１０３論理ディスク排他アクセス時間監視手段
１０４物理ディスクビジー時間監視手段
１０５論理ディスク再配置手段
２００物理ディスク群
２１０物理ディスク
２１１論理ディスク
２１２論理ディスク
２１３論理ディスク
２２０物理ディスク
２２１論理ディスク
２２２論理ディスク
２２３論理ディスク
２３０物理ディスク
２３１論理ディスク
２３２論理ディスク
２３３論理ディスク
２４０物理ディスク
２４１論理ディスク
２４２論理ディスク
２４３論理ディスク
３００論理ディスク管理テーブル
４００ＩＯカウンタテーブル[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a data relocation method, a data relocation method, and a data relocation program, and more particularly, to a data relocation method capable of efficiently relocating a logical disk in a physical disk where I / O is concentrated, The present invention relates to a data relocation method and a data relocation program.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, in a disk array device that divides an area on a plurality of physical disks and forms a plurality of logical disks in which the divided areas are collectively virtualized by a certain unit, when I / O requests are concentrated on a specific physical disk, Subsequent I / O requests to the physical disk are delayed until the preceding I / O request is completed, and the response is degraded. If such performance degradation occurs in a part of the apparatus, there is a problem that the performance degradation of the apparatus as a whole is a bottleneck.
[0003]
An example of a technique for solving such a problem is described in JP-A-2000-215007. In JP-A-2000-215007, a busy value of a physical disk is measured, and when the busy rate of the physical disk exceeds a value specified by a user, a logical disk on the physical disk is selected, and data on the logical disk is selected. The logical disk is transferred to another physical disk and the logical disk is relocated to change the correspondence between the logical disk and the physical disk. This eliminates the situation where I / O is concentrated on some physical disks, thereby improving the performance of the entire disk array device.
[0004]
[Problems to be solved by the invention]
In the above-described conventional technique, there is a problem that extra cost is required because the rearrangement of the logical disk is wastefully performed. The reason is that I / O occurs in the reallocation of the logical disk itself, so even if normal operation is stopped and then reallocation is performed, normal operation cannot be performed during that time. On the other hand, if the relocation processing is executed behind the normal operation, the I / O associated with the rearrangement interferes with the normal operation I / O. Since the rearrangement process is a very expensive process, careless execution of the rearrangement must be suppressed. However, in the conventional relocation method in which the relocation is determined by simply comparing the busy rates of the physical disks, in the physical disks constituting the logical disk, I / O is performed for the same logical disk on the same physical disk. Assuming a situation of concentration, the busy rate of the physical disk is detected to be high. However, even if the logical disk is relocated to another physical disk, the performance of the entire disk array device is not improved just because the physical disk on which I / O is concentrated eventually moves to another. Is a wasteful process.
[0005]
An object of the present invention is to solve the above problems and to provide a logical disk relocation method capable of efficiently relocating a logical disk in a physical disk where I / O is concentrated.
[0006]
[Means for Solving the Problems]
According to a first data relocation method of the present invention, in a computer system in which at least one physical disk including at least one logical disk is connected, a logical disk having a maximum exclusive access time has a minimum busy time. A logical disk relocation means for moving to a physical disk is provided.
[0007]
According to a second data relocation method of the present invention, in the first data relocation method, a logical disk exclusive access time monitoring means for checking an exclusive access time for each logical disk, and a physical disk for checking a busy time for each physical disk. A disk busy time monitoring unit, wherein the logical disk relocation unit includes an exclusive access time for each logical disk checked by the logical disk exclusive access time monitoring unit, and the physical disk checked by the physical disk busy time monitoring unit. The logical disk having the maximum exclusive access time is moved to the physical disk having the minimum busy time, based on the busy time of each disk.
[0008]
According to a third data relocation method of the present invention, in the second data relocation method, a busy time indicating a cumulative value of a busy state of each of the physical disks and a time at which the physical disk becomes busy And an IO counter indicating the number of IO requests in a waiting or processing state for each of the logical disks belonging to each of the physical disks, and an exclusive access state. An IO counter table for storing an exclusive access start time indicating the time when the logical disk has been accessed and an exclusive access time indicating the cumulative value of the exclusive access state, and the logical disk exclusive access time monitoring means includes at least a logical disk number. When the IO start notification including the IO counter table is received, the logical The IO counter of the logical disk corresponding to the logical disk number is incremented by one, and it is checked whether or not the IO counter is 2 or more. If the IO counter is smaller than 2, the IO counter in the physical disk to which the logical disk belongs is checked. Check whether the IO counters of the other logical disks are all zero, and when all of the IO counters are not zero, record the current time in the exclusive access start time of the logical disk in the IO counter table and The number of other logical disks whose IO counter is 1 or more is checked, and if there is only one, the current time is recorded at the exclusive access start time of the other logical disk, and IO completion including at least the logical disk number is completed. When the notification is received, the logical disk number is referred to by referring to the IO counter table. It is checked whether the IO counter of the corresponding logical disk is 2 or more. If the IO counter is smaller than 2, the IO counters of other logical disks in the physical disk to which the logical disk belongs are all zero. If all of the IO counters are not zero, the value obtained by subtracting the exclusive access start time of the logical disk in the IO counter table from the current time is added to the exclusive access time of the logical disk. The number of other logical disks whose IO counter is 1 or more is checked, and if there is only one, the value obtained by subtracting the exclusive access start time of the other logical disk in the IO counter table from the current time is calculated. It is added to the exclusive access time of another logical disk, and the IO counter of the logical disk is decremented by one. When receiving the IO start notification including at least the physical disk number, the physical disk busy time monitoring unit refers to the IO counter table and checks the IO of all the logical disks belonging to the physical disk corresponding to the physical disk number. Check whether the counter is zero, and if all the IO counters are zero, record the current time at the busy state start time of the physical disk in the IO counter table, and store at least the IO including the physical disk number. When the completion notification is received, it is checked whether or not the IO counters of the logical disks belonging to the physical disk corresponding to the physical disk number are all zero by referring to the IO counter table. If the physical disk is in the IO counter table from the current time, Is added to the busy time of the physical disk, and when an IO request to a logical disk including at least a logical disk number is received, the IO start notification including at least the logical disk number is transmitted to the logical disk. Along with transmitting to the disk start means, an IO start notification including at least the physical disk number of the physical disk to which the logical disk belongs is transmitted to the physical disk busy time monitoring means, and after accessing the physical disk, the physical disk number is Logical disk management means for transmitting an IO completion notification including at least the physical disk busy time monitoring means to the physical disk busy time monitoring means and transmitting an IO completion notification including at least the logical disk number to the logical disk start means; Is the IO counterte With exclusive access time with reference to the table is checked logical disk is a maximum, examine the physical disks busy time is minimal, the logical disks and moving to the physical disk.
[0009]
According to a fourth data relocation method of the present invention, in the third data relocation method, the data relocation method includes a logical disk management table for storing a correspondence relationship between the physical disk and the logical disk, and the logical disk exclusive access time monitoring means includes: Referring to the logical disk management table to check the physical disk to which the logical disk belongs; the physical disk busy time monitoring unit refers to the logical disk management table to check the logical disk belonging to the physical disk; The logical disk relocation means, after moving the logical disk having the maximum exclusive access time to the physical disk having the minimum busy time, sets the correspondence between the logical disk and the physical disk in the logical disk management table. Is changed.
[0010]
According to a first data relocation method of the present invention, in a computer system in which at least one or more physical disks including at least one or more logical disks are connected, a computer can use a logical disk with a maximum exclusive access time with a busy time. The method further comprises a logical disk relocation step of moving to the smallest physical disk.
[0011]
According to a second data relocation method of the present invention, in the first data relocation method, the computer is configured to monitor a logical disk exclusive access time for checking an exclusive access time for each logical disk; A physical disk busy time monitoring step of checking a busy time of each disk, wherein the logical disk relocation step includes an exclusive access time of each logical disk checked in the logical disk exclusive access time monitoring step, and a physical disk busy time. The logical disk having the maximum exclusive access time is moved to the physical disk having the minimum busy time based on the busy time of each physical disk checked in the time monitoring step.
[0012]
According to a third data relocation method of the present invention, in the second data relocation method, the logical disk exclusive access time monitoring step includes, when an IO start notification including at least a logical disk number is received, For each of them, a busy time indicating an accumulated value of a busy state and a busy state start time indicating a time at which a busy state is stored are stored for each of the logical disks belonging to each of the physical disks. , An IO counter indicating the number of IO requests in a waiting or processing state, an exclusive access start time indicating a time when the exclusive access state is entered, and an exclusive access indicating an accumulated value of the exclusive access state time The logical disk corresponding to the logical disk number in the IO counter table that stores the time Is incremented by one, and it is checked whether the IO counter is 2 or more. If the IO counter is smaller than 2, the IO counter of another logical disk in the physical disk to which the logical disk belongs is checked. Check whether the counters are all zero. If all of the IO counters are not zero, record the current time at the exclusive access start time of the logical disk in the IO counter table, and set the IO counter to 1 or more. Check the number of certain other logical disks, if there is only one, record the current time at the exclusive access start time of the other logical disk, and if an IO completion notification including at least the logical disk number is received, , Referring to the IO counter table, the IO of the logical disk corresponding to the logical disk number The IO counter is 2 or more, and when the IO counter is smaller than 2, it is checked whether all IO counters of other logical disks in the physical disk to which the logical disk belongs are zero. When all of the IO counters are not zero, a value obtained by subtracting the exclusive access start time of the logical disk in the IO counter table from the current time is added to the exclusive access time of the logical disk, and the IO counter is set to 1 or more. The number of other logical disks is checked, and if there is only one, the value obtained by subtracting the exclusive access start time of the other logical disk in the IO counter table from the current time is calculated as the exclusive time of the other logical disk. Add to the access time, decrement the IO counter of the logical disk by 1, and set the physical disk busy time In the monitoring step, when the IO start notification including at least the physical disk number is received, the IO counters of all the logical disks belonging to the physical disk corresponding to the physical disk number are zero with reference to the IO counter table. If the IO counter is all zero, the current time is recorded at the busy start time of the physical disk in the IO counter table, and an IO completion notification including at least the physical disk number is received. In this case, referring to the IO counter table, it is checked whether the IO counters of the logical disks belonging to the physical disk corresponding to the physical disk number are all zero, and if the IO counters are all zero, At the start of the busy state of the physical disk in the IO counter table from the current time Is added to the busy time of the physical disk, and upon receiving an IO request to the logical disk including at least the logical disk number, transmitting the IO start notification including at least the logical disk number to the logical disk start step Sending an IO start notification including at least the physical disk number of the physical disk to which the logical disk belongs to the physical disk busy time monitoring step, and after accessing the physical disk, an IO completion notification including at least the physical disk number To the physical disk busy time monitoring step, and transmitting an IO completion notification including at least the logical disk number to the logical disk start step, wherein the logical disk relocation step comprises: With exclusive access time by referring to the table determine the logical disk is a maximum, examine the physical disks busy time is minimal, the logical disks and moving to the physical disk.
[0013]
In a fourth data relocation method according to the present invention, in the third data relocation method, the logical disk exclusive access time monitoring step refers to a logical disk management table that stores a correspondence between the physical disk and the logical disk. Then, the physical disk to which the logical disk belongs, the physical disk busy time monitoring step refers to the logical disk management table to check the logical disk belonging to the physical disk, the logical disk relocation step, After moving the logical disk with the longest exclusive access time to the physical disk with the shortest busy time, the correspondence between the logical disk and the physical disk in the logical disk management table is changed. .
[0014]
According to a first program of the present invention, in a computer system in which at least one physical disk including at least one logical disk is connected, a logical disk having a maximum exclusive access time is replaced with a physical disk having a minimum busy time. The present invention is characterized by causing a computer to execute a logical disk relocation process to be moved.
[0015]
A second program according to the present invention, in the first program, is a logical disk exclusive access time monitoring process for checking the exclusive access time for each logical disk, and a physical disk busy time monitoring process for checking the busy time for each physical disk. The logical disk relocation processing includes: an exclusive access time for each logical disk checked in the logical disk exclusive access time monitoring processing; and a busy time for each physical disk checked in the physical disk busy time monitoring processing. Thus, the logical disk having the maximum exclusive access time is moved to the physical disk having the minimum busy time.
[0016]
A third program according to the present invention, in the second program, wherein the logical disk exclusive access time monitoring process includes, when an IO start notification including at least a logical disk number is received, a busy state of each of the physical disks. While storing a busy time indicating the accumulated value of the time period, and a busy state start time indicating the time at which the state became busy, and waiting for or during processing for each of the logical disks belonging to each of the physical disks. , An IO counter indicating the number of IO requests in the state, an exclusive access start time indicating the time at which the exclusive access state was established, and an exclusive access time indicating the cumulative value of the time during which the exclusive access state was established. IO counter of the logical disk corresponding to the logical disk number in the counter table Increment by 1 and checking whether the IO counter is 2 or more. If the IO counter is smaller than 2, all IO counters of other logical disks in the physical disk to which the logical disk belongs are set to zero. If the IO counters are not all zero, the current time is recorded in the exclusive access start time of the logical disk in the IO counter table, and another logical counter whose IO counter is 1 or more is recorded. The number of disks is checked, and if there is only one, the current time is recorded at the exclusive access start time of the other logical disk, and if an IO completion notification including at least the logical disk number is received, the IO counter Referring to the table, if the IO counter of the logical disk corresponding to the logical disk number is 2 or more It is checked whether the IO counter is smaller than 2. If the IO counter is smaller than 2, it is checked whether the IO counters of the other logical disks in the physical disk to which the logical disk belongs are all zero. If it is not zero, a value obtained by subtracting the exclusive access start time of the logical disk in the IO counter table from the current time is added to the exclusive access time of the logical disk, and another logical counter whose IO counter is 1 or more is added. The number of disks is checked, and if there is only one, the value obtained by subtracting the exclusive access start time of the other logical disk in the IO counter table from the current time is added to the exclusive access time of the other logical disk. The IO counter of the logical disk is decremented by 1, and the physical disk busy time monitoring At least, when the IO start notification including the physical disk number is received, the IO counter table is checked to see whether the IO counters of all the logical disks belonging to the physical disk corresponding to the physical disk number are zero. If the IO counter is all zeros, the current time is recorded at the busy start time of the physical disk in the IO counter table, and if the IO completion notification including at least the physical disk number is received, By referring to the IO counter table, it is checked whether the IO counters of the logical disks belonging to the physical disk corresponding to the physical disk number are all zero, and if the IO counters are all zero, the current time The value obtained by subtracting the busy state start time of the physical disk in the IO counter table from When the IO request to the logical disk including at least the logical disk number is received by adding the IO request to the logical disk busy time, the IO start notification including at least the logical disk number is transmitted to the logical disk start process, and the logical disk Sends an IO start notification including at least the physical disk number of the physical disk to which the physical disk belongs to the physical disk busy time monitoring process, and after accessing the physical disk, transmits an IO completion notification including at least the physical disk number to the physical disk busy time. A logical disk management process for transmitting to the monitoring process and transmitting an IO completion notification including at least the logical disk number to the logical disk start process, wherein the logical disk relocation process is exclusive with reference to the IO counter table Maximum access time Together determine the logical disk that checks the physical disks busy time is minimal, the logical disks and moving to the physical disk.
[0017]
A fourth program according to the present invention is the third program, wherein the logical disk exclusive access time monitoring process refers to the logical disk management table storing the correspondence between the physical disk and the logical disk. The physical disk to which the physical disk belongs belongs, the physical disk busy time monitoring processing refers to the logical disk management table to check the logical disk belonging to the physical disk, and the logical disk relocation processing has a maximum exclusive access time. After moving a certain logical disk to a physical disk having the minimum busy time, the correspondence between the logical disk and the physical disk in the logical disk management table is changed.
[0018]
BEST MODE FOR CARRYING OUT THE INVENTION
Next, the present invention will be described in detail with reference to the drawings.
[0019]
First, a first embodiment of the present invention will be described in detail.
[0020]
Referring to FIG. 1, the first embodiment of the present invention includes a computer 100, a physical disk group 200, a logical disk management table 300, and an IO counter table 400, and the computer 100 is controlled by a program. Operate.
[0021]
The computer 100 includes a logical disk management unit 101, a physical disk management unit 102, a logical disk exclusive access time monitoring unit 103, a physical disk busy time monitoring unit 104, and a logical disk relocation unit 105.
[0022]
The physical disk group 200 includes a plurality of physical disks 210, 220, 230, and 240, divides an area on the physical disk, and assigns each area as a logical data storage area. It comprises logical disks 221, 222, 223, logical disks 231, 232, 233, and logical disks 241, 242, 243.
[0023]
The logical disk management table 300 stores the correspondence between the divided areas on the physical disk and the logical disks.
[0024]
The IO counter table 400 stores, for each of the physical disks 210 to 240, a busy time indicating an accumulated value of a busy state time and a busy state start time indicating a time when the physical disk is busy. For each of the logical disks 211 to 243 belonging to each of the disks 210 to 240, an IO counter indicating the number of IO requests that are in a processing waiting or processing state, an exclusive access start time indicating a time when the exclusive access state is set, and And the exclusive access time indicating the cumulative value of the time of the exclusive access state.
[0025]
The logical disk management means 101 receives an I / O request for the logical disks 211 to 243 by designating the logical disk number, refers to the logical disk management table 300, acquires the physical disk number corresponding to the logical disk number, An I / O request is made to the physical disk management means 102 by designating the physical disk number.
[0026]
The physical disk management unit 102 receives an I / O request for the physical disks 210 to 240, and issues I / O to the requested physical disks 210 to 240.
[0027]
The logical disk exclusive access time monitoring means 103 is called at the start of I / O and at the end of I / O, monitors the exclusive access state of the logical disk, and records the exclusive access time of the logical disk in the IO counter table 400. .
[0028]
The physical disk busy time monitoring means 104 is called at the start of I / O and at the end of I / O, monitors the I / O issue state for the physical disk, and records the busy time of the physical disk in the IO counter table 400. .
[0029]
The logical disk relocation unit 105 receives the relocation instruction, determines the target logical disk to be relocated and the destination physical disk by referring to the IO counter table 400, and determines whether relocation is possible. Then, the data on the logical disk is transferred to a free area on the physical disk of the transfer destination, and the correspondence between the physical disk and the logical disk on the logical disk management table 300 is updated.
[0030]
Next, the overall operation of the present embodiment will be described in detail with reference to FIGS.
[0031]
FIG. 2 is a flowchart illustrating the operation of the logical disk management unit 101.
[0032]
First, upon receiving an I / O request specifying a logical disk number, the logical disk management unit 101 refers to the logical disk management table 300 and converts the logical disk number into a corresponding physical disk number (S201). .
[0033]
Next, the logical disk management means 101 notifies the logical disk exclusive access time monitoring means 103 of the start of the logical disk I / O processing by the logical disk number (S202).
[0034]
Next, the logical disk management unit 101 notifies the physical disk busy time monitoring unit 104 of the start of the physical disk I / O processing by the physical disk number (S203).
[0035]
Next, the logical disk management unit 101 issues an I / O request to the physical disk by designating the physical disk number to the physical disk management unit 102 (S204).
[0036]
Further, after making the I / O request, the logical disk management unit 101 waits until a response to the I / O request is returned from the physical disk management unit 102 (S205).
[0037]
Next, after the response to the I / O request is returned, the logical disk management unit 101 notifies the physical disk busy time monitoring unit 104 of the completion of the I / O of the physical disk by the physical disk number (S206). ).
[0038]
Next, the logical disk management unit 101 notifies the logical disk exclusive access time monitoring unit 103 of the completion of the I / O of the logical disk by the logical disk number (S207).
FIG. 3 is a flowchart showing the operation of the physical disk management unit 102.
[0039]
First, upon receiving an I / O request for a physical disk by designating a physical disk number from the logical disk management means 101, the physical disk management means 102 issues I / O to the specified physical disk (S211). And wait until the issued I / O is completed (S212).
[0040]
Next, after the I / O to the physical disk is completed, the physical disk management unit 102 returns a response to the I / O request to the logical disk management unit 101 (S213).
[0041]
FIG. 4 is a flowchart showing the operation of the logical disk exclusive access time monitoring means 103.
[0042]
First, when the logical disk exclusive access time monitoring unit 103 receives a notification of the logical disk number from the logical disk management unit 101 at the start of I / O, the IO of the logical disk corresponding to the logical disk number in the IO counter table 400 is The counter is incremented (S221), and it is checked whether the IO counter is 2 or more (S222).
If the IO counter is smaller than 2, it is determined that the request is the first I / O request for the logical disk, and whether the IO counters of other logical disks in the physical disk to which the logical disk belongs are all zero. Confirmation (S223).
If all of the IO counters are not zero, it means that a subsequent I / O request arrived while processing a previous I / O request for another logical disk, and Since the exclusive control is performed for each physical disk, the target logical disk is determined to be in the exclusive access state, and the current time is recorded as the exclusive access start time of the logical disk in the IO counter table 400 (S224). .
[0043]
Further, it is confirmed whether there is only one other logical disk whose IO counter is 1 or more (S225).
[0044]
If there is only one, it is determined that the other logical disk has transitioned to the exclusive access state due to the arrival of the current I / O request, and the current time is recorded at the exclusive access start time of the other logical disk (S226). ).
Thereafter, the process waits for an I / O completion notification from the physical disk management unit 102 (S227), and upon receiving the I / O completion notification with the designated logical disk number, refers to the IO counter table 400 and checks the IO of the logical disk. It is confirmed whether the counter is 2 or more (S228).
If the I / O counter is smaller than 2, the I / O completion notification determines that all I / O requests for the logical disk have been completed, and further determines that other I / Os in the physical disk to which the logical disk belongs. It is checked whether the IO counter is all zero (S229).
[0045]
If all of the IO counters are not zero, it is determined that the exclusive access state of the logical disk has been terminated by the completion of the I / O request, and the exclusive access start time of the logical disk in the IO counter table 400 is determined from the current time. Is added to the exclusive access time of the logical disk (S230).
[0046]
Further, it is confirmed whether there is only one other logical disk whose IO counter is 1 or more (S231).
When there is only one, the I / O request for the other logical disk is not in the exclusive access state due to the completion of the I / O of this logical disk, so the I / O request of the other logical disk in the IO counter table 400 starts from the current time. The value obtained by subtracting the exclusive access start time is added to the exclusive access time of the other logical disk (S232).
Finally, the IO counter of the logical disk for which I / O has been completed is decremented (S233).
FIG. 5 is a flowchart showing the operation of the physical disk busy time monitoring means 104.
First, when the physical disk busy time monitoring unit 104 receives a notification of the physical disk number at the time of starting I / O from the logical disk management unit 101, the physical disk busy time monitoring unit 104 refers to the IO counter table 400 to check all the logical disks belonging to the physical disk. It is checked whether the IO counter is zero (S241).
If the IO counters are all zero, it is determined that the physical disk has shifted from the idle state to the busy state, and the current time is recorded at the busy state start time of the physical disk in the IO counter table 400 (S242).
Thereafter, the process waits for I / O completion (S243), and upon receiving the I / O completion notification, refers to the IO counter table 400 and checks whether or not the IO counters of the logical disks belonging to the physical disk are all zero (0). S244), when the IO counter is all zero, the completion of this I / O determines that the target physical disk has shifted from the busy state to the idle state, and starts the busy state of the physical disk in the IO counter table 400 from the current time. The value obtained by subtracting the time is added to the busy time as the time during which the busy state continues (S245).
FIG. 6 is a flowchart showing the operation of the logical disk relocation unit 105.
First, when the logical disk relocation unit 105 is called in response to a relocation execution instruction, the logical disk relocation unit 105 refers to the IO counter table 400, checks the exclusive access times for all the logical disks, and indicates the largest exclusive access time. The logical disk is determined as the logical disk to be moved (S261).
Next, the busy times of all the physical disks are investigated, and the physical disk having the shortest busy time that has enough free space to store the determined logical disk to be moved is determined as the destination physical disk. (S262).
Next, it is determined whether there is an effect of relocation when the target logical disk is moved to the destination physical disk (S263).
For example, when the determined exclusive access time of the migration target logical disk is larger than a threshold value predetermined by the user, the idle time (time during which no I / O is issued at the migration destination physical disk. When the ratio of the exclusive access time of the migration target logical disk to the exclusive access time of the migration target logical disk is larger than the specified value determined by the user, it is determined that there is an effect of relocation when the busy time is subtracted from the observation time. is there. If it is determined by the above method that there is no relocation effect, the relocation processing is not performed.If it is determined that the relocation effect is effective, a new logical disk is placed on the determined destination physical disk. It configures (S264) and transfers the data on the migration target logical disk to a new logical disk (S265).
[0047]
Although it is determined in step S263 whether or not the rearrangement has an effect, the rearrangement processing in steps S264 to S265 may be performed without performing the determination.
[0048]
Next, the migration source logical disk number on the logical disk management table is rewritten to indicate a new logical disk area (S266).
[0049]
Finally, the area of the source logical disk is released as a free area (S267).
[0050]
Next, a second embodiment of the present invention will be described with reference to the drawings.
[0051]
The second embodiment of the present invention includes a data relocation program. The data relocation program controls the operation of the computer 100, and causes the computer 100 to execute the above-described means, that is, the same means as the means in the first embodiment.
[0052]
【The invention's effect】
As described above, according to the logical disk relocation means of the present invention, there is an effect that a logical disk in a physical disk where I / O is concentrated can be efficiently relocated.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of a first exemplary embodiment of the present invention.
FIG. 2 is a flowchart illustrating an operation of a logical disk management unit according to the first embodiment of this invention.
FIG. 3 is a flowchart illustrating an operation of a physical disk management unit according to the first embodiment of this invention.
FIG. 4 is a flowchart illustrating an operation of a logical disk exclusive access time monitoring unit according to the first embodiment of this invention.
FIG. 5 is a flowchart illustrating an operation of a physical disk busy time monitoring unit according to the first embodiment of this invention.
FIG. 6 is a flowchart illustrating an operation of a logical disk relocation unit according to the first embodiment of this invention.
[Explanation of symbols]
100 computer
101 Logical disk management means
102 Physical disk management means
103 Logical disk exclusive access time monitoring means
104 Physical disk busy time monitoring means
105 Logical disk relocation means
200 Physical Disk Group
210 Physical Disk
211 Logical disk
212 logical disk
213 Logical disk
220 physical disk
221 Logical disk
222 logical disk
223 logical disk
230 Physical Disk
231 logical disk
232 logical disk
233 logical disk
240 physical disks
241 logical disk
242 logical disk
243 logical disk
300 Logical disk management table
400 IO counter table

Claims

In a computer system in which at least two physical disks including at least one logical disk are connected,
After I / O request acceptance, by the I / O processing to a different logical disk and a logical disk as a target of the I / O, exclusive access each logical disk which accumulates only time the issuance of the I / O must wait A data relocation method comprising a logical disk relocation means for moving a logical disk having a longest time to a physical disk having a minimum busy time.

A logical disk exclusive access time monitoring means for checking the exclusive access time for each logical disk,
Physical disk busy time monitoring means for checking the busy time of each physical disk,
The logical disk relocation unit is configured to perform exclusive control based on the exclusive access time of each logical disk checked by the logical disk exclusive access time monitoring unit and the busy time of each physical disk checked by the physical disk busy time monitoring unit. 2. The data relocation method according to claim 1, wherein the logical disk having the longest access time is moved to the physical disk having the shortest busy time.

For each of the physical disks, a busy time indicating an accumulated value of a busy time and a busy state start time indicating a time at which the physical disk is busy are stored, and the logical belonging to each of the physical disks is stored. For each of the disks, an IO counter indicating the number of IO requests that are waiting or being processed, an exclusive access start time indicating the time when the exclusive access state was established, and a cumulative value of the time during which the exclusive access state was established An IO counter table that stores an exclusive access time indicating
When receiving the IO start notification including at least the logical disk number, the logical disk exclusive access time monitoring means increments the IO counter of the logical disk corresponding to the logical disk number in the IO counter table by one, and Check whether the IO counter is 2 or more, and if the IO counter is smaller than 2, check whether the IO counters of other logical disks in the physical disk to which the logical disk belongs are all zero, If all of the IO counters are not zero, the current time is recorded in the exclusive access start time of the logical disk in the IO counter table, and the number of other logical disks whose IO counter is 1 or more is checked. If there is only one, start exclusive access to the other logical disk The current time is recorded every time, and when an IO completion notification including at least the logical disk number is received, the IO counter of the logical disk corresponding to the logical disk number is 2 or more with reference to the IO counter table. If the IO counter is smaller than 2, it is checked whether the IO counters of other logical disks in the physical disk to which the logical disk belongs are all zero, and all of the IO counters are zero. If not, the value obtained by subtracting the exclusive access start time of the logical disk in the IO counter table from the current time is added to the exclusive access time of the logical disk, and the other logical disk whose IO counter is 1 or more is added. Is checked, and if there is only one, the other logic in the IO counter table is counted from the current time. The value obtained by subtracting the exclusive access start time of the disk is added to the exclusive access time of the other logical disk, it is decremented by 1 to IO counter of the logical disk,
The physical disk busy time monitoring means, upon receiving an IO start notification including at least the physical disk number, refers to the IO counter table and checks the IO of all the logical disks belonging to the physical disk corresponding to the physical disk number. Check whether the counter is zero, and if all the IO counters are zero, record the current time at the busy state start time of the physical disk in the IO counter table, and store at least the IO including the physical disk number. When the completion notification is received, it is checked whether or not the IO counters of the logical disks belonging to the physical disk corresponding to the physical disk number are all zero by referring to the IO counter table. In the case of the physical counter in the IO counter table from the current time, The value obtained by subtracting the busy start time of the click is added to the busy time of the physical disk,
Upon receiving an IO request to a logical disk including at least the logical disk number, the IO request is transmitted to the logical disk starting means including the logical disk number and the physical disk number of the physical disk to which the logical disk belongs Transmitting an IO start notification including at least the physical disk number to the physical disk busy time monitoring unit, and transmitting an IO completion notification including at least the physical disk number to the physical disk busy time monitoring unit after accessing the physical disk. Logical disk management means for transmitting an IO completion notification including at least a disk number to the logical disk start means,
The logical disk relocation unit refers to the IO counter table to check a logical disk having the longest exclusive access time, check a physical disk having the shortest busy time, and move the logical disk to the physical disk. 3. The data rearrangement method according to claim 2, wherein:

A logical disk management table that stores a correspondence between the physical disk and the logical disk;
The logical disk exclusive access time monitoring means refers to the logical disk management table, checks the physical disk to which the logical disk belongs,
The physical disk busy time monitoring unit refers to the logical disk management table to check a logical disk belonging to the physical disk,
The logical disk relocation means, after moving the logical disk having the maximum exclusive access time to the physical disk having the minimum busy time, sets the correspondence between the logical disk and the physical disk in the logical disk management table. 4. The data rearrangement method according to claim 3, wherein (i) is changed.

In a computer system in which at least two physical disks including at least one logical disk are connected,
Computer, after I / O request acceptance, by the I / O processing to a different logical disk and a logical disk as a target of the I / O, each logical disk which accumulates only time the issuance of the I / O must wait A logical disk relocation step of moving a logical disk having the maximum exclusive access time to a physical disk having the minimum busy time.

A logical disk exclusive access time monitoring step in which the computer checks an exclusive access time for each logical disk;
The computer comprises a physical disk busy time monitoring step of examining a busy time for each physical disk,
The logical disk relocation step is performed based on the exclusive access time of each logical disk checked in the logical disk exclusive access time monitoring step and the busy time of each physical disk checked in the physical disk busy time monitoring step. 6. The data relocation method according to claim 5, wherein the logical disk having the longest access time is moved to the physical disk having the shortest busy time.

The logical disk exclusive access time monitoring step includes, when an IO start notification including at least a logical disk number is received, a busy time indicating an accumulated value of a busy state for each of the physical disks; And an IO counter indicating the number of IO requests in a waiting or processing state for each of the logical disks belonging to each of the physical disks, and an exclusive counter. The IO counter of the logical disk corresponding to the logical disk number in the IO counter table that stores the exclusive access start time indicating the time when the access state has been entered and the exclusive access time indicating the accumulated value of the time during the exclusive access state Is incremented by one, and the IO counter Is checked whether the IO counter is 2 or more, and when the IO counter is smaller than 2, it is checked whether the IO counters of other logical disks in the physical disk to which the logical disk belongs are all zero, and When all of the counters are not zero, the current time is recorded in the exclusive access start time of the logical disk in the IO counter table, and the number of other logical disks having the IO counter of 1 or more is checked. In some cases, the current time is recorded at the exclusive access start time of the other logical disk, and when an IO completion notification including at least the logical disk number is received, the logical disk is referred to by referring to the IO counter table. Check whether the IO counter of the logical disk corresponding to the number is 2 or more, and check that the IO counter is smaller than 2. Sometimes, it is checked whether or not the IO counters of the other logical disks in the physical disk to which the logical disk belongs are all zero. If all of the IO counters are not zero, the IO counter in the IO counter table is counted from the current time. The value obtained by subtracting the exclusive access start time of the logical disk is added to the exclusive access time of the logical disk, and the number of other logical disks whose IO counter is 1 or more is checked. The value obtained by subtracting the exclusive access start time of the other logical disk in the IO counter table from the time is added to the exclusive access time of the other logical disk, and the IO counter of the logical disk is decremented by 1.
In the physical disk busy time monitoring step, when an IO start notification including at least a physical disk number is received, the IO counter table is referred to, and the IO of all the logical disks belonging to the physical disk corresponding to the physical disk number is referred to. Check whether the counter is zero, and if all the IO counters are zero, record the current time at the busy state start time of the physical disk in the IO counter table, and store at least the IO including the physical disk number. When the completion notification is received, it is checked whether or not the IO counters of the logical disks belonging to the physical disk corresponding to the physical disk number are all zero by referring to the IO counter table. , The physical time in the IO counter table from the current time The value obtained by subtracting the busy start time of the disk is added to the busy time of the physical disk,
Upon receiving an IO request to the logical disk including at least the logical disk number, the IO request including at least the logical disk number is transmitted to the logical disk start step, and the physical disk number of the physical disk to which the logical disk belongs Transmitting an IO start notification including at least the physical disk number to the physical disk busy time monitoring step, and transmitting an IO completion notification including at least the physical disk number to the physical disk busy time monitoring step after accessing the physical disk. A logical disk management step of transmitting an IO completion notification including at least a disk number to the logical disk start step,
The logical disk relocation step refers to the IO counter table to check a logical disk having the longest exclusive access time, check a physical disk having the shortest busy time, and move the logical disk to the physical disk. 7. The data rearrangement method according to claim 6, wherein:

The logical disk exclusive access time monitoring step refers to a logical disk management table that stores a correspondence between the physical disk and the logical disk, and checks a physical disk to which the logical disk belongs,
The physical disk busy time monitoring step refers to the logical disk management table to check a logical disk belonging to the physical disk,
The logical disk relocation step includes, after moving the logical disk having the maximum exclusive access time to the physical disk having the minimum busy time, and then moving the logical disk in the logical disk management table between the logical disk and the physical disk. 8. The data relocation method according to claim 7, wherein the data relocation is changed.

In a computer system in which at least two or more physical disks including at least one logical disk are connected,
After I / O request acceptance, by the I / O processing to a different logical disk and a logical disk as a target of the I / O, exclusive access each logical disk which accumulates only time the issuance of the I / O must wait A program for causing a computer to execute a logical disk relocation process for moving a logical disk having a longest time to a physical disk having a minimum busy time.

A logical disk exclusive access time monitoring process for checking the exclusive access time for each logical disk;
Physical disk busy time monitoring processing for examining the busy time of each physical disk,
The logical disk relocation processing is performed based on the exclusive access time for each logical disk checked in the logical disk exclusive access time monitoring processing and the busy time for each physical disk checked in the physical disk busy time monitoring processing. The program according to claim 9, wherein the logical disk having the longest access time is moved to the physical disk having the shortest busy time.

The logical disk exclusive access time monitoring process includes, when an IO start notification including at least a logical disk number is received, a busy time indicating a cumulative value of a busy state for each of the physical disks; And an IO counter indicating the number of IO requests in a waiting or processing state for each of the logical disks belonging to each of the physical disks, and an exclusive counter. The IO counter of the logical disk corresponding to the logical disk number in the IO counter table that stores the exclusive access start time indicating the time when the access state has been entered and the exclusive access time indicating the accumulated value of the time during the exclusive access state Is incremented by one, and the IO counter is It is confirmed whether or not the above is satisfied, and when the IO counter is smaller than 2, it is confirmed whether or not the IO counters of the other logical disks in the physical disk to which the logical disk belongs are all zero, and When all are not zero, the current time is recorded in the exclusive access start time of the logical disk in the IO counter table, and the number of other logical disks whose IO counter is 1 or more is checked, and only one is present. Records the current time at the exclusive access start time of the other logical disk, and when receiving the IO completion notification including at least the logical disk number, refers to the IO counter table and Check whether the IO counter of the corresponding logical disk is 2 or more, and if the IO counter is smaller than 2, It is checked whether the IO counters of the other logical disks in the physical disk to which the logical disk belongs are all zero, and if all of the IO counters are not zero, the current time is counted from the current time in the IO counter table. The value obtained by subtracting the exclusive access start time of the logical disk is added to the exclusive access time of the logical disk, and the number of other logical disks whose IO counter is 1 or more is checked. A value obtained by subtracting the exclusive access start time of the other logical disk in the IO counter table from the current time is added to the exclusive access time of the other logical disk, and the IO counter of the logical disk is decremented by one,
The physical disk busy time monitoring process, when receiving an IO start notification including at least the physical disk number, refers to the IO counter table and checks the IO of all the logical disks belonging to the physical disk corresponding to the physical disk number. Check whether the counter is zero, and if all the IO counters are zero, record the current time at the busy state start time of the physical disk in the IO counter table, and store at least the IO including the physical disk number. When the completion notification is received, it is checked whether or not the IO counters of the logical disks belonging to the physical disk corresponding to the physical disk number are all zero by referring to the IO counter table. In the case of the physical counter in the IO counter table from the current time, The value obtained by subtracting the busy start time of the click is added to the busy time of the physical disk,
Upon receiving an IO request to the logical disk including at least the logical disk number, the IO start notification including at least the logical disk number is transmitted to the logical disk start process, and the physical disk number of the physical disk to which the logical disk belongs Transmitting an IO start notification including at least the physical disk number to the physical disk busy time monitoring process, and transmitting an IO completion notification including at least the physical disk number to the physical disk busy time monitoring process after accessing the physical disk. A logical disk management process of transmitting an IO completion notification including at least a disk number to the logical disk start process,
The logical disk relocation processing refers to the IO counter table to check the logical disk having the longest exclusive access time, check the physical disk having the shortest busy time, and move the logical disk to the physical disk. The program according to claim 10, wherein:

The logical disk exclusive access time monitoring process refers to a logical disk management table that stores a correspondence relationship between the physical disk and the logical disk, checks a physical disk to which the logical disk belongs,
The physical disk busy time monitoring process refers to the logical disk management table to check a logical disk belonging to the physical disk,
The logical disk relocation processing includes, after moving the logical disk having the longest exclusive access time to the physical disk having the shortest busy time, the correspondence between the logical disk and the physical disk in the logical disk management table. 12. The program according to claim 11, wherein the program is changed.