Electrical computers and digital processing systems: memory – Storage accessing and control – Specific memory composition
Reexamination Certificate
1999-12-22
2002-08-20
Hudspeth, David (Department: 2651)
Electrical computers and digital processing systems: memory
Storage accessing and control
Specific memory composition
C711S170000, C710S036000
Reexamination Certificate
active
06438648
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to data storage systems. More particularly, the present invention relates to a system, structure, and method for managing multiple sets of different host computer operating requirements by a disk drive array controller.
BACKGROUND OF THE INVENTION
Disk drives in all computer systems are susceptible to failures caused by temperature variations, head crashes, motor failure, controller failure, and changing voltage conditions. Modern computer systems require, or at least benefit from, a fault-tolerant data storage system, for protecting data in the data storage system against instances of disk drive failure. One approach to meeting this need is to provide a redundant array of independent disks (RAID) system operated by a disk array controller (controller).
Referring to
FIG. 1
, there is shown a block diagram illustrating aspects of a data storage system
100
configured as a RAID, according to the state of the art. The data storage system
100
includes controller
102
coupled across I/O bus
104
to one or more host computers
118
-
125
. The controller
102
is also coupled across I/O bus
114
to a rack of disk drives
106
-
110
. I/O bus
104
and I/O bus
114
can each respectively be one of several different types of I/O buses, including, for example, a parallel SCSI bus, an FC-AL bus, a Fibre Channel Fabric bus, a USB bus, or an IEEE 1394 bus.
The controller
102
coordinates the disk drives
106
-
110
so they appear to host computers
118
-
125
as a single logical drive, but can take advantage of the multiple physical drives
106
-
110
by storing data redundantly across the disk drives
106
-
110
, thus protecting against data loss from any single disk drive
106
-
110
. A logical unit (“LU”) is the medium used by controller
102
to store and retrieve data respectively to and from the disk drives
106
-
110
. (LUs and are known in the art).
A controller
102
typically maps from eight (8) to one-hundred-and-twenty-eight (128) LUs across each of the disk drives
106
-
110
. (Procedures to perform such mapping are known in the art.) For example, LUs
0
-X and
1
-X are mapped across disk drives
106
-
108
, illustrated respectively as LU
0
-A, LU
1
-A, LU
0
-B, and LU
1
-B, and the like.
Controller
102
processes I/O requests from host computers
118
-
125
. Such I/O requests include, for example, Small Computer System Interface (SCSI) I/O requests, which are known in the art of computer programming. Host computers
118
-
125
typically execute one or more computer programs (individual computer programs are not shown) that have particular requirements with respect to the operational behavior of the LUs
0
-N in response to processing such I/O requests.
For example, consider that host computer
118
is used to execute a video playback computer program (not shown), and that host computer
120
is used to execute a video capture computer program (not shown). The video playback application (host computer
118
) requires a timely response from each respective disk drive
106
-
110
to every read data request (for example, a SCSI read data request) that it sends to controller
102
, such that a playback operation of a video image (not shown) by the video playback application is not interrupted by a slow return of the requested data from any respective disk drive
106
-
110
.
In contrast, the video capture application (host computer
120
) does not require a timely response from each respective disk drive
106
-
110
to every I/O request that it sends to controller
102
. However, the video capture application does require each respective disk drive
106
-
110
to implement error recovery information to ensure that the video data that is written to each respective disk drive
106
-
110
by the video capture application is stored with no errors. Such error recovery information is known in the art of computer programming.
If each disk drive
106
-
110
were to be configured to implement such error recovery information to the benefit of computer's
118
video playback application, a write data request sent by the computer's
118
video capture application may result in insufficient amounts of error recovery information being written to disk drives
106
-
110
. Analogously, if each disk drive
106
-
110
were to be configured to implement such error recovery information to the benefit of computer's
120
video capture application, a read data request sent by the computer's
120
video playback application may incur unacceptable delays reading data from disk drive
106
-
110
. Therefore, it is evident that host computers
118
and
120
have different operating requirements with respect to the operational behavior of the LUs
0
-N in response to processing such I/O requests.
When a data storage system
100
includes a plurality of host computers
118
-
125
that have diverse operating requirements with respect to at least a subset of one another, the host computers
118
-
125
(and controller
102
) are said to be operating in a heterogeneous host computer
118
-
125
operating environment. For a host computer
118
-
125
in such an environment to properly operate, it must be able to configure the operational behavior of the LUs
0
-N to parallel the host computer's
118
-
125
particular requirements.
SCSI standards provide a procedure for a particular host computer
118
-
125
to specify its particular operating requirements with respect to the operation of each LU
0
-N that is mapped across disk drives
106
-
110
. This is typically accomplished through the use of the SCSI mode select command that includes one or more sets of mode parameter pages, wherein each mode parameter page identifies one or more of the particular host computer's
118
-
125
particular operating requirements. The use and format of the mode select command and mode parameter pages are known in the art of computer programming.
Because disk drives
106
-
112
are being operated as a RAID, the particular host computer
118
-
125
will send the mode select command to the controller
102
, whereupon, while processing any I/O requests from the particular computer
118
-
125
, the controller
102
configures LUs
0
-N to operate in the appropriate manner. According to SCSI standards (see SCSI-2 document ANSI X3.331 (1994), section 8.2.8, and SCSI-3 document ANSI X3.301 (1997), section 7.8), in a heterogeneous host computer
118
-
125
operating environment, controller
102
will store and maintain in random access memory (RAM)
103
one set of mode parameter pages per host computer
118
-
125
per LU
0
-N, as illustrated by callout
105
.
In such a heterogeneous computer
118
-
125
operating environment, significant problems are presented by the use of conventional techniques to store mode parameter pages in a controller
102
. One problem is that the amount of RAM
103
that a controller
102
may be required to use to store multiple sets of mode parameter pages may be excessive. This problem is a function of the number of host computers
118
-
125
and the number of LUs
0
-N in a data storage system
100
.
The number of host computers
118
-
125
in data storage system
100
can vary. For example: (a) if I/O bus
104
is a parallel SCSI bus, as many as 15 host computers
118
-
125
can be coupled to controller
102
; (b) if I/O bus
104
is an FC-AL bus, as many as 128 host computers
118
-
125
can be coupled to controller
102
; and, (c) if I/O bus
104
is a Fibre Channel Fabric bus, then over 16 million host computers
118
-
125
can be coupled to the controller
102
. However, recall that a controller
102
typically maps from 8 to 128 LUs
0
-N across disk drives
106
-
110
.
In light of the above, the number of mode parameter pages that a controller
102
may need to store in RAM
103
can widely vary. If controller
102
were coupled to 128 host computers in a heterogeneous computer operating environment that includes 128 LUs
0
-N, controller
102
will be required to store and maintain a staggering
Archibald, Jr. John Edward
McKean Brian Dennis
Ananian R. Michael
Dorsey & Whitney LLP
Hudspeth David
Tzeng Fred F.
LandOfFree
System apparatus and method for managing multiple host... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System apparatus and method for managing multiple host..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System apparatus and method for managing multiple host... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2890091