US6256748B1 — Method and device for connecting a data processin…

Claim 1

Independent Claim 252 words

A process for connecting the central processing unit (CPU) of a data processing system to a subsystem comprising data storage resources (MD) placed under the control of two data storage controllers (SP-A, SP-B) operating redundantly, said central processing unit (CPU) being associated with at least one input-output controller (Ctl) communicating with said data storage controllers (SP-A, SP-B) through two separate data transmission paths (B 0 , B 1 ), characterized in that it comprises at least the following steps: 1) an initial step for assigning at least some of said data storage resources (PD 1 -PD 6 ) to a first data storage controller (SP-A), thereby establishing a normal data transmission path (B 0 ) between said assigned some of data storage resources (PD 1 -PD 6 ) and said first data storage controller (SP-A) and a backup data transmission path (B 1 ) between any unassigned data storage resources (PD 1 -PD 6 ) and a second data storage controller (SP-B); 2) a step for directing control through at least one input-output controller, via said normal data transmission path (B 0 ), to at least one of said data storage resources assigned to said data storage controller (SP-A), which is said to be active, wherein said second data storage controller (SP-B) is said to be inactive; and 3) a step for continuously polling, via said backup data transmission path (B 1 ), the state of said backup data storage controller (SP-B) and of said backup data transmission path (B 1 ).

Claim 2

Dependent Claim 252 words

A process for connecting the central processing unit (CPU) of a data processing system to a subsystem comprising data storage resources (MD) placed under the control of two data storage controllers (SP-A, SP-B) operating redundantly, said central processing unit (CPU) being associated with at least one input-output controller (Ctl) communicating with said data storage controllers (SP-A, SP-B) through two separate data transmission paths (B 0 , B 1 ), characterized in that it comprises at least the following steps: 1) an initial step for assigning at least some of said data storage resources (PD 1 -PD 6 ) to a first data storage controller (SP-A), thereby establishing a normal data transmission path (B 0 ) between said assigned some of data storage resources (PD 1 -PD 6 ) and said first data storage controller (SP-A) and a backup data transmission path (B 1 ) between any unassigned data storage resources (PD 1 -PD 6 ) and a second data storage controller (SP-B); 2) a step for directing control through at least one input-output controller, via said normal data transmission path (B 0 ), to at least one of said data storage resources assigned to said data storage controller (SP-A), which is said to be active, wherein said second data storage controller (SP-B) is said to be inactive; and 3) a step for continuously polling, via said backup data transmission path (B 1 ), the state of said backup data storage controller (SP-B) and of said backup data transmission path (B 1 ).

Claim 3

Dependent Claim 102 words

The process according to claim 1 , characterized in that said initial step is comprised of assigning a first partition (PD 1 -PD 3 ) of said data storage resources to a first data storage controller (SP-A) and a second partition (PD 4 -PD 6 ) to a second data storage controller (SP-B), and in the event that a failure state is detected during said continuously polling step, said process comprises an additional step for switching said data storage resources assigned to said backup data storage controller (SP-B) to said active data storage controller (SP-A), and for reassigning them to the latter.

Claim 4

Dependent Claim 106 words

The process according to claim 2 , characterized in that, each input-output controller (Ctl) associated with said central processing unit (CPU) storing a program module (P S ) for continuously polling the state of said backup data storage controller (SP-B) via said backup data transmission path (B 1 ), the process comprises, in the event of a failure state of this supervisory program module (P S ) or of the backup path (B 1 ), an additional step for switching said data storage resources assigned to said backup data storage controller (SP-B) to said active data storage controller (SP-A), and for reassigning them to the latter.

Claim 5

Dependent Claim 242 words

The process for connecting the central processing unit (CPU) of a data processing system to a subsystem comprising data storage resources (MD) placed under the control of two data storage controllers (SP-A, SP-B) operating redundantly, said central processing unit (CPU) being associated with at least one input-output controller (Ctl) communicating with said data storage controllers (SP-A, SP-B) through two separate data transmission paths (B 0 , B 1 ), characterized in that it comprises at least the following steps: 1) an initial step for assigning at least some of said data storage resources (PD 1 -PD 6 ) to a first data storage controller (SP-A); 2) a step comprised of executing input-output commands, transmitted by an input-output controller (Ctl) via said normal data transmission path (B 0 ), in order to access one of said data storage resources assigned to said active data storage controller (SP-A); 3) a step for continuously polling the state of this controller (SP-A) and of said normal data transmission path (B 0 ); and in that when a failure state is detected during this last step, it comprises the following subsequent steps: a/ a step for switching said storage resources assigned to the active data storage controller (SP-A) to the backup data storage controller (SP-B) and for reassigning said switched storage resources to said backup data storage controller (SP-B); and b/ a step for re-executing said input-output commands via said backup data transmission path (B 1 ).

Claim 6

Dependent Claim 242 words

The process for connecting the central processing unit (CPU) of a data processing system to a subsystem comprising data storage resources (MD) placed under the control of two data storage controllers (SP-A, SP-B) operating redundantly, said central processing unit (CPU) being associated with at least one input-output controller (Ctl) communicating with said data storage controllers (SP-A, SP-B) through two separate data transmission paths (B 0 , B 1 ), characterized in that it comprises at least the following steps: 1) an initial step for assigning at least some of said data storage resources (PD 1 -PD 6 ) to a first data storage controller (SP-A); 2) a step comprised of executing input-output commands, transmitted by an input-output controller (Ctl) via said normal data transmission path (B 0 ), in order to access one of said data storage resources assigned to said active data storage controller (SP-A); 3) a step for continuously polling the state of this controller (SP-A) and of said normal data transmission path (B 0 ); and in that when a failure state is detected during this last step, it comprises the following subsequent steps: a/ a step for switching said storage resources assigned to the active data storage controller (SP-A) to the backup data storage controller (SP-B) and for reassigning said switched storage resources to said backup data storage controller (SP-B); and b/ a step for re-executing said input-output commands via said backup data transmission path (B 1 ).

Claim 7

Dependent Claim 44 words

The process according to claim 4 , characterized in that it comprises an additional step, when no input-output command is executed, comprised of the continuous supervision of the state of said active controller (SP-A) and of said normal data transmission path (B 0 ).

Claim 8

Dependent Claim 155 words

The process according to claim 1 , characterized in that each input-output controller (Ctl) associated with said central processing unit (CPU) being equipped with two input-output ports (Po 0 , Po 1 ), connected through data transmission busses (B 0 , B 1 ) to said data storage controllers (SP-A, SP-B), and storing program modules (D 0 , D 1 ) for managing these input-output ports (Po 0 , Po 1 ), the process comprises, in the event of a failure state of one of these port management program modules (D 0 or D 1 ), an additional step for switching said data storage resources assigned to said data storage controller (SP-A or SP-B) associated with said malfunctioning port management program module (D 0 or D 1 ) to said data storage controller (SP-A or SP-B) associated with the other input-output port (D 0 or D 1 ) and for reassigning them to the latter.

Claim 9

Dependent Claim 54 words

The process according to claim 1 , characterized in that it includes an additional step, when a failure state is detected, comprised of transmitting an information signal to said central processing unit (CPU) indicating said switching of data storage resources and the passage from a so-called normal functioning mode into a so-called degraded mode.

Claim 10

Dependent Claim 71 words

The process according to claim 7 , characterized in that it includes an initial step comprised of a test for determining the state of said data storage controllers (SP-A, SP-B) and of the associated paths (B 0 B 1 ), and in accordance with the result of this test, of configuring the operation of said data storage resource subsystem (SP-A, SP-B, MD) in said normal mode or in said degraded mode.

Claim 11

Dependent Claim 27 words

The process according to claim 1 , characterized in that it includes an initial step comprised of applying a given redundant configuration to said storage resources (MD).

Claim 12

Dependent Claim 69 words

The process according to claim 9 , characterized in that, said storage resources (MD) being constituted by an array of individual magnetic disks (D 1 -D 30 ), said redundant configuration is comprised of associating with each individual magnetic disk a second individual magnetic disk, called a mirror disk, and of redundantly recording on the latter the data recorded on the individual magnetic disk with which it is associated.

Claim 13

Dependent Claim 98 words

The process according to claim 9 , characterized in that, said data storage resources (MD) being constituted by an array of individual magnetic disks (D 1 -D 30 ), said redundant configuration is comprised of subdividing said array into at least one sub-array (PD 1 -PD 6 ) of a given number of individual magnetic disks, of recording on said given number of individual magnetic disks a given number of so-called useable data blocks (S 0 -S 4 ), and of associating with these data blocks at least one block of redundant, so-called parity data (P 1 ).

Claim 14

Dependent Claim 26 words

The process according to claim 9 , characterized in that it includes an initial step comprised of determining the nature of said configuration of given redundancy.

Claim 15

Independent Claim 409 words

A device for connecting the central processing unit (CPU) of a data processing system to a redundant data storage system, the device comprising at least one input-output controller (Ctl) associated with said central processing unit (CPU), two data storage controllers (SP-A, SP-B) operating redundantly, a first normal (B 0 ) and a second backup (B 0 ) data transmission bus (B 0 , B 1 ) for each input-output controller (Ctl) connected to said data storage controllers (SP-A, SP-B) so as to form said normal (B 0 ) and said backup (B 1 ) data transmission path, and an array of data storage resources (MD) under the control of said storage controllers (SP-A, SP-B), in that a first partition (PD 1 -PD 3 ) of said storage resources is assigned, in a so-called normal operating mode, to a first data storage controller (SP-A), and a second partition is assigned to the other data storage controller (SP-B), and in that each input-output controller (Ctl) comprises means (P S ) for continuously polling and detecting a failure state of the data storage controllers (SP-A, SP-B) and of the data transmission paths (B 0 , B 1 ), and for switching and reassigning said data storage resources from one data storage controller to the other in the event that a failure state is detected; said device being configured to perform the following steps: 1) an initial step for assigning at least some of said data storage resources (P 1 -PD 6 ) to a first data storage controller (SP-A), thereby establishing said normal data transmission path (B 0 ) between said all or some of data storage resources (PD 1 -PD 6 ) and said first data storage controller (SP-A) and said backup data transmission path (B 1 ) between any unassigned said all or some of data storage resources (PD 1 -PD 6 ) and a second data storage controller (SP-B); 2) a step for directing control through at least one input-output controller, via said normal data transmission path (B 0 ), to at least one of said data storage resources assigned to said data storage controller (SP-A), which is said to be active, wherein said second data storage controller (SP-B) is said to be inactive and; 3) a step for continuously polling, via said backup data transmission path (B 1 ), the state of said backup data storage controller (SP-B) and of said backup data transmission path (B 1 ).

Claim 16

Independent Claim 409 words

A device for connecting the central processing unit (CPU) of a data processing system to a redundant data storage system, the device comprising at least one input-output controller (Ctl) associated with said central processing unit (CPU), two data storage controllers (SP-A, SP-B) operating redundantly, a first normal (B 0 ) and a second backup (B 0 ) data transmission bus (B 0 , B 1 ) for each input-output controller (Ctl) connected to said data storage controllers (SP-A, SP-B) so as to form said normal (B 0 ) and said backup (B 1 ) data transmission path, and an array of data storage resources (MD) under the control of said storage controllers (SP-A, SP-B), in that a first partition (PD 1 -PD 3 ) of said storage resources is assigned, in a so-called normal operating mode, to a first data storage controller (SP-A), and a second partition is assigned to the other data storage controller (SP-B), and in that each input-output controller (Ctl) comprises means (P S ) for continuously polling and detecting a failure state of the data storage controllers (SP-A, SP-B) and of the data transmission paths (B 0 , B 1 ), and for switching and reassigning said data storage resources from one data storage controller to the other in the event that a failure state is detected; said device being configured to perform the following steps: 1) an initial step for assigning at least some of said data storage resources (P 1 -PD 6 ) to a first data storage controller (SP-A), thereby establishing said normal data transmission path (B 0 ) between said all or some of data storage resources (PD 1 -PD 6 ) and said first data storage controller (SP-A) and said backup data transmission path (B 1 ) between any unassigned said all or some of data storage resources (PD 1 -PD 6 ) and a second data storage controller (SP-B); 2) a step for directing control through at least one input-output controller, via said normal data transmission path (B 0 ), to at least one of said data storage resources assigned to said data storage controller (SP-A), which is said to be active, wherein said second data storage controller (SP-B) is said to be inactive and; 3) a step for continuously polling, via said backup data transmission path (B 1 ), the state of said backup data storage controller (SP-B) and of said backup data transmission path (B 1 ).

Claim 17

Dependent Claim 32 words

The device according to claim 13 , characterized in that said means for continuously polling and detecting a failure state comprise a program module (P S ) stored in said input-output controller.

Claim 18

Dependent Claim 212 words

The device according to claim 13 , characterized in that it comprises two input-output controllers (Ctl A -Ctl B ), operating redundantly, associated with said central processing unit (CPU), each comprising a first input-output port (Po A0 , Po B0 ) and a second input-output port (Po A1 , Po B1 ), the first input-output port (Po A0 ) of the first of said input-output controllers (Ctl A ) being connected through a first data transmission bus to said first data storage controller (SP-A) and the second input-output port (Po A1 ) of this controller (Ctl A ) being connected through a second data transmission bus to said second data storage controller (SP-B), so as to form said normal path to said first (SP-A) and second (SP-B) data storage controllers, respectively, and the first input-output port (Po B1 ) of the second of said input-output controllers (Ctl B ) being connected through a first data transmission bus to said first data storage controller (SP-A) and the second input-output port (Po B1 ) of this input-output controller (Ctl B ) being connected through a second data transmission bus to said second data storage controller (SP-B), so as to form said backup path to said first (SP-A) and second (SP-B) data storage controllers, respectively.

Claim 19

Dependent Claim 40 words

The device according to claim 15 , characterized in that it comprises two sets of input-output controllers (Ctl 1 -Ctl 2 , Ctl 3 -Ctl 4 ) associated with said central processing unit (CPU), operating redundantly and disposed in parallel.

Claim 20

Dependent Claim 101 words

The device according to claim 13 , characterized in that said data storage resources (MD) are constituted by an array of individual magnetic disks (D 1 -D 3 ), in that these individual disks are grouped into an array forming a configuration of given redundancy, in that this array is divided into logic units (LUN 0 -LUN 2 ) and in that these logic units are subdivided into first and second sub-arrays, the first sub-array of logic units being assigned to a first storage controller (SP-A) and the second sub-array of logic units being assigned to the second storage controller (SP-B).

Annuity Stage	Window Opens	Payment Deadline	Surcharge Period Until	Estimated Fee	Status
3.5 Years (Year 4)	2004-07-02	2005-01-01	2005-07-03	$2000.00	Window Closed / Lapsed
7.5 Years (Year 8)	2008-07-02	2009-01-01	2009-07-03	$3760.00	Window Closed / Lapsed
11.5 Years (Year 12)	2012-07-02	2013-01-01	2013-07-03	$7700.00	Window Closed / Lapsed

Method and device for connecting a data processing system central unit to a redundancy data storage subsystem

Abstract

Inventors

Bibliographic Details

Claims Analysis

References Cited (Backward Citations)

USPTO Annuity & Maintenance Windows

Assigned Categories & Tags

Assign Category Tag