Direct data retrieval in a distributed computing system

Details Direct data retrieval in a distributed computing system Direct data retrieval in a distributed computing system

1998-02-10

2001-07-10

Rinehart, Mark H. (Department: 2152)

Electrical computers and digital processing systems: multicomput

Computer-to-computer session/connection establishing

Network resources access controlling

C709S213000, C709S214000, C709S215000, C709S216000, C707S793000, C714S006130, C714S012000

Reexamination Certificate

active

06260069

ABSTRACT:

1. FIELD OF THE INVENTION
Preferred embodiments of the present invention relate to a system for direct data retrieval in a distributed computing system to backup files in the system at an attached storage device.
2. BACKGROUND OF THE RELATED ART
In a distributed computing system, different computers, operating systems, and networks interact as if they were all part of a single system. The file system has a single set of global file names. A particular machine in the system need not know where the file is physically located. Instead, the file may be accessed anywhere in the network using the global file name. Global file names are part of the shared name space which devices within the distributed file system may access. One such distributed file system is the Andrew File System (AFS) available through Transarc, Corporation (Transarc). An AFS server performs file mapping between the directory name of a file and the location, making the file space location independent. With file independence, a user at a workstation linked to the network need only know the global file name, which includes the path name, and not the physical location of the file.
Another distributed system, is the Distributed File System (DFS), available from Transarc and International Business Machines, Corp. (IBM), which is a component of the Distributed Computing Environment (DCE) standard promulgated by the Open Software Foundation (OSF). The DFS and AFS systems allow users to have access to data throughout the network. Any changes made by one user to a file is available to all users. The DFS and AFS systems provide authentication services to provide security and limit access to only authorized users.
The AFS system offered by Transarc includes a backup program called “butc” (Backup Tape Coordinator). Butc is a volume backup system used to dump volume images to tape devices attached to the file server. However, the minimum backup unit for the butc program is a volume as the butc program does not provide support for file-level backup and recovery.
Hierarchical storage management programs, such as the IBM Adstar Distributed Storage Management (ADSM) product, provide backup/archive support and migrate less frequently used files to secondary storage to free space in primary storage. The ADSM server provides hierarchical storage management and backup/archive functions, and manages the storage of files on tape drives, optical disks, and other storage medium. The ADSM backup feature saves copies of files from the primary storage in a client computer to the secondary storage space managed by the ADSM server. Thus, data at a client computer running an ADSM client is protected in the event of data loss due to a hardware or software failure, accidental deletion, and/or logical corruption. With the ADSM program, clients can backup volumes, directories, subdirectories or files. ADSM allows incremental backup of only those files that have been changed. In this way, ADSM avoids the need to do a full dump to backup as only those modified files are backed up. This incremental backup reduces network utilization and traffic. The IBM ADSM product is described in “ADSM Version 2 Presentation Guide,” (IBM Document SG24-4532-00, International Business Machines, copyright 1995), which publication is incorporated herein by reference in its entirety.
IBM has combined its ADSM product with AFS and DFS file servers to provide backup support for these products. An AFS or DFS server would include an ADSM client to backup files to an ADSM server. In such case, the ADSM server backs-up files from the AFS or DFS server to a secondary storage device the ADSM server manages. One problem with using such backup software in a distributed file system is that the client managing backup operations, such as the ADSM client, must read a file to be backed-up. This reading operation consumes network resources. The ADSM client must then consume network resources again by transferring the read file to the ADSM server. Network traffic is further increased if the ADSM client is on a separate machine from the AFS/DFS server. The IBM publications entitled “ADSM AFS/DFS Backup Clients Version 2.1” (IBM Document SH26-4048-00, International Business Machines, copyright 1996) and “ADSM Concepts” (IBM Document SG24-4877-00, International Business Machines, copyright 1997) describes the use of the ADSM software in an AFS/DFS distributed file system. These publications are incorporated herein by reference in their entirety.
SUMMARY OF THE INVENTION
To address the shortcomings in the prior art described above, preferred embodiments of the present invention provide a system for backing up files in a distributed computing system. A backup request is initiated with a backup client program to backup a requested file. A determination is made as to whether the requested file is maintained in a shared name space. The backup client program and a backup server program are capable of accessing files maintained in the shared name space. A file server maintains the files in the shared name space. The backup request is transmitted to the backup server program upon determining that the requested file is maintained in the shared name space. The backup server program transmits a message to the file server to provide the requested file. The file server transmits the requested file to the backup server program. The backup server program stores the requested file in a storage device.
In further embodiments, the backup client program is on a first computer machine, the backup server program is on a second computer machine, and the file server is on a third computer machine. The first computer, second computer, and third computer communicate over a network system.
In yet further embodiments, the backup client program and backup server program are on a first computer machine and the file server program is on a second computer machine. The first computer machine and second computer machine communicate over a network system.
It is an object of preferred embodiments of the present invention to provide a system for backing up files in a shared name space maintained in a file server which is part of a distributed computing environment on a storage device managed by a backup server program, such as a hierarchical storage management program.
It is yet a further object to reduce network traffic throughout the distributed computing environment by having the backup server program handle requests by a client computer to backup files maintained in the file server on the storage device managed by the server program.


REFERENCES:
patent: 5005122 (1991-04-01), Griffin et al.
patent: 5434994 (1995-07-01), Shaheen et al.
patent: 5452448 (1995-09-01), Sakuraba et al.
patent: 5495533 (1996-02-01), Linehan et al.
patent: 5588147 (1996-12-01), Neeman et al.
patent: 5628005 (1997-05-01), Hurvig
patent: 5649196 (1997-07-01), Woodhill et al.
patent: 5673381 (1997-09-01), Huei et al.
patent: 5689701 (1997-11-01), Ault et al.
patent: 5713017 (1998-01-01), Lin et al.
patent: 5813017 (1998-09-01), Morris
patent: 5832522 (1998-11-01), Blickenstaff et al.
patent: 5857102 (1999-01-01), McChesney et al.
patent: 5867650 (1999-02-01), Osterman
patent: 5918228 (1999-06-01), Rich et al.
patent: 6026414 (2000-02-01), Anglin
patent: 6148412 (2000-11-01), Cannon et al.
patent: 0 574 900 A2 (1993-12-01), None
patent: 0 773 503 A2 (1997-05-01), None
patent: 2 288 477 (1995-10-01), None
patent: 8202602 (1995-01-01), None
patent: 8044609 (1995-01-01), None
ACM Computing Surveys, vol. 22, No. 4, Dec. 1990, “Distributed File Systems: Concepts and Examples” by Eliezer Levy and Abraham Silberschatz.
IBM Corporation's International Technical Support Organization “ADSM concepts”, SG24-4877-00, Feb. 1997, Chapters 1 & 6.
IBM Corporation's ADSTAR Distributed Storage Manager “Using the UNIX Backup-Achive Clients”, Version 3, Release 1, SH26-4075-01, Chapters 1 & 3.
IBM Corporation's ADSTAR Distributed Storage Manger for HP-UX “Administrator's Guide”, Version 3, GC35-0320-00 Chapters 4 & 12.
La Polla, Stephanie, “Auspex

Affiliated with

Also associated with

Rating

Direct data retrieval in a distributed computing 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 Direct data retrieval in a distributed computing system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Direct data retrieval in a distributed computing system will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2486637