Multi-array disk apparatus

• US 6,038,681 A
• Filed: 09/03/1997
• Issued: 03/14/2000
• Est. Priority Date: 09/04/1996
• Status: Expired due to Term

• Current Assignee: NEC Corporation
• Original Assignee: NEC Corporation
• Inventors: Saegusa, Takuya
• Primary Examiner(s): WRIGHT, NORMAN M
• Application Number:
  US08/922,750
• Time in Patent Office: 923 Days
• Field of Search: 395/182.04, 395/182.03, 395/182.05, 714/6, 714/5, 714/3, 714/25, 714/30, 714/42, 714/47
• US Class Current: 714/6.21
• CPC Class Codes: G06F 11/2092 Techniques of failing over between control units

Abstract

A multi-array disk apparatus includes first and second logical drives, a first array controller, a second array controller, and interface control circuits. Each of the first and second logical drives has two logical structure disks constituting a pair. The first array controller controls the first logical drive. The second array controller controls the second logical drive. When a fault occurs in the other of the first and second array controllers, the interface control circuits bus-connect one of the first and second array controller in which no fault occurs to the logical structure disk connected to the other of the first and second array controller in which the fault occurs.

Claims

1. A multi-array disk apparatus comprising:

array controllers and logical drives, each of said logical drives corresponding to one of said array controllers;

said logical drives comprising a first logical drive and a second logical drive;

said array controllers comprising a first array controller and a second array controller;

each of said logical drives comprising a respective logical disks, including a respective first logical disk and a respective second logical disk;

for each said logical disk of one of said logical drives, a corresponding protocol controller in said corresponding one of said array controllers, said corresponding protocol controller being connected to said logical disk by a respective bus;

all of said protocol controllers having an identical first ID;

all of said logical disks having an identical second ID;

a first interface unit connected between said respective bus of said second logical disk of said first logical drive and said respective bus of said first logical disk of said second logical drive;

a second interface unit connected between said respective bus of said first logical disk of said first logical drive and said respective bus of said second logical disk of said second logical drive;

said first and said second interface units being connected to said first and said second array controllers;

said first array controller, in response to a fault detection relating to said second array controller, activating said first and said second interface units; and

said first and said second interface units, when activated by said first array controller;

connecting together said respective buses,

monitoring messages on said buses, and

changing message ID information for selected ones of said messages relating to said logical disks of said second logical drive.

2. An apparatus according to claim 1, wherein, in the absence of said fault detection relating to said second array controller, said interface units do not connect together said respective buses.

3. An apparatus according to claim 1, wherein said interface control units comprise means for changing said ID information to a predetermined different ID information.

4. An apparatus according to claim 3, wherein said ID information is constituted by bit data, and said ID changing means comprises bit replacement means for replacing a bit representing the ID of said second array controller, and a bit representing the ID of said logical disks of said second logical drive with bits representing said predetermined different ID information.

5. An apparatus according to claim 3, wherein said interface control units each comprise:

first and second driver/receiver units connected to said first and second array control units through a SCSI data bus and a SCSI control bus, and being connected to said ID replacement means;

a SCSI phase monitor unit, connected to a control signal path between said first and second driver/receiver units, for activating said ID changing means on the basis of a monitor result of a SCSI phase; and

a control means for enabling control said first and second driver/receiver means and for controlling a bus connection direction in response to a drive enable request signal from said first array controller.