Electrical computers and digital processing systems: memory – Storage accessing and control – Specific memory composition
Reexamination Certificate
1999-06-29
2003-10-07
Thai, Tuan V. (Department: 2186)
Electrical computers and digital processing systems: memory
Storage accessing and control
Specific memory composition
C711S100000, C711S154000, C711S170000
Reexamination Certificate
active
06631442
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to data storage systems, and more particularly, to an interface mechanism which allows host computing devices to interoperate with the data storage system.
BACKGROUND OF THE INVENTION
The widespread use of computerized data processing systems has created vast amounts of data that must be stored. To serve this need, data storage system providers have developed data storage systems that use tape or disk arrays to store such data. In a typical data storage system using disk or tape array technology, there can be many individual physical storage devices, such as hard disk drives or tape drives. Circuitry and/or software in either the data storage system or in the host computing devices (hosts) which interface to the data storage system can manipulate the way in which data is stored in the various individual storage devices. Various arrangements of the storage devices and of the data maintained in each device are possible. Each arrangement generally allows the storage devices to appear as a single contiguous data repository or as groups of repositories to an application executing on a host that requires storage or access to stored data.
By way of example, a data storage technology such as Redundant Arrays of Inexpensive Disks (RAID) allows a data storage system to present conglomerations of many individual hard disk drives to host computing devices as one or more disk storage areas. In general, the data storage system handles the RAID operations in such a way that they are transparent to the host computing devices which interface to the data storage system. RAID systems also provide various levels of fault tolerance, error correction and error recovery. This allows the hosts to treat a conglomeration of disks within the RAID data storage system without regard to the underlying physical separation or layout of data within or “across” each individual disk drive. For more details concerning RAID technology and systems which implement this technology, the reader is referred to Patterson et al., “A Case for Redundant Arrays of Inexpensive Disks (RAID),” ACM SIGMOND Conference, Jun. 1-3, 1988, the teaching of which are hereby incorporated by reference in their entirety.
The conglomerations of disks presented as a single repository for data to a host computing device from a data storage system are called volumes. A prior art volume identifies and serves as a host interface to the raw physical storage associated with one or more storage devices within the data storage system. Typically, in the prior art, a person known as a systems administrator configures one or more volumes during an initial or periodic configuration process performed on the data storage system. Between configurations, the total amount of available data storage within a prior art volume does not change. For instance, an administrator may configure a data storage system containing four physically separate hard disk drives “A”, “B”, “C” and “D” to present only two volumes of data storage, V
1
and V
2
, to a host computing device that interfaces to the data storage system. The first volume V
1
may be a conglomeration of physical storage space (using RAID or another data-over-multiple-disk technology) located on disks “A” and “B”, while the second volume V
2
may be a conglomeration of storage space existing within hard disk drives “C” and “D”.
A single volume need not be comprised of space from more than one storage device. For example, a volume might be comprised of only a portion of storage space from a single hard disk. In this case, the volume may be equivalent, for example, to a partition, slice or other apportionment of that hard disk.
In any event, prior art data storage systems provide volumes that are essentially interface mechanisms to a host computing device. Generally, in the prior art, a single host directly interfaces (i.e., via a SCSI bus or other connection mechanism) to the data storage system and “serves” the data from the data storage system onto a network for access by other hosts, which do not directly interface to the data storage system. The volumes appear to the server host as individual contiguous repositories for data. Within the data storage system, however, the volumes may actually span portions of one or more storage devices (e.g. disk drives, tape drives, and so forth). Providing volumes insulates server software applications and server host computing devices from the details of complicated data and disk storage mechanisms such as mirroring, error correction, striping and so on. From here on in this description, a host or host computing device generally refers to a host (i.e., a server) that directly interfaces with a data storage system.
Generally, when a host computing device starts-up or “boots”, an operating system that controls the host polls any attached data storage systems via device drivers to determine what volumes are available for access to the host computing device within the data storage system. By way of example, a host executing a version of the Solaris operating system (a variant of Unix), which is manufactured by Sun Microsystems of Mountain View, Calif., (Solaris being a registered trademark of Sun Microsystems Inc.) can use an operating system call, such as “report LUNS” for a SCSI-3 device, to poll the SCSI port coupled to the data storage system in order to determine volume availability. In response to the host poll, the data storage system provides a list of available volumes in the given target device. Host-specific access information stored in the volumes may be reported back to the host as well.
A host filesystem in the host computing device can then access a particular volume by “mounting” a volume identifier associated with the volume (obtained from the initial poll) that is provided from the data storage system to a “mount point” within the host filesystem. A filesystem mount point is essentially a stub directory within the host filesystem that serves as an entry point into the volume once the host has mounted the volume. After the host has mounted the volume, software applications on the host can access data in the volume by referencing the mount point directory and any subdirectories that are now available within the mounted volume of data storage. For example, in a computing device controlled by the Unix operating system, a mount point may correspond to a directory within the Unix filesystem, such as “/home”. A system command such as
mount /home /dev/dsk/c
0
t
0
d
0
s
0
can be used to mount the volume identified by “/dev/dsk/c
0
t
0
d
0
s
0
” to the mount point “/home” in the Unix filesystem. After the volume is mounted, users or applications executing on the computing device can reference files and data within the /home directory and any subdirectories within /home that actually exist within the data storage associated with the volume. That is, all data accessed within the /home directory is physically maintained on storage space within storage devices (e.g., hard disk drives) that are associated with the mounted volume “/dev/dsk/c
0
t
0
d
0
s
0
” in the data storage system.
Most operating systems rely on volumes provided by a data storage system for certain operations. For example, many host operating systems expect storage space within volumes to be organized into sequential blocks of data ranging from Block
0
to Block N. Many operating systems frequently store certain host-specific volume directory information, partition information, track/sector layout and size information, and so forth on a predetermined designated portion of storage space within a volume. Block
0
(i.e. the first block) of a volume is frequently used for storing such host-specific data. As another example, the Microsoft Windows NT operating system, manufactured by Microsoft Corporation of Redmond, Wash. performs periodic checks to make sure that a mounted volume is always accessible within the data storage system. Windows and Windows NT are registered trademarks of Microsoft Corporation.
Generally, once an operating syste
Chapin & Huang , L.L.C.
Chaping, Esq. Barry W.
Thai Tuan V.
LandOfFree
Methods and apparatus for interfacing to a data storage system does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Methods and apparatus for interfacing to a data storage system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Methods and apparatus for interfacing to a data storage system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3114310