Workstation-implemented data storage re-routing for server fault-tolerance on computer networks

• US 5,633,999 A
• Filed: 12/18/1995
• Issued: 05/27/1997
• Est. Priority Date: 11/07/1990
• Status: Expired due to Fees

• Current Assignee: Nonstop Networks Ltd.
• Original Assignee: Nonstop Networks Ltd.
• Inventors: Tims, Fred W., Clowes, Richard F.
• Primary Examiner(s): Canney, Vincent P.
• Application Number:
  US08/573,865
• Time in Patent Office: 526 Days
• Field of Search: 395/182.04, 395/183.16
• US Class Current: 714/6.21
• CPC Class Codes: G06F 11/142 Reconfiguring to eliminate the error group management mechanisms in a peer-to-peer network H04L67/1044
• CPC Class Codes: G06F 11/1482 by means of middleware or OS functionality
• CPC Class Codes: G06F 11/2033 switching over of hardware resources
• CPC Class Codes: G06F 11/2035 without idle spare hardware
• CPC Class Codes: G06F 11/2046 where the redundant components share persistent storage G06F11/2043 takes precedence
• CPC Class Codes: G06F 11/2071 using a plurality of controllers
• CPC Class Codes: G06F 11/2076 Synchronous techniques

Abstract

Server fault tolerance on local area computer networks is provided by novel workstation implemented software which detects failure of a primary file server and reroutes data storage activity routed via the primary file server, to be routed via an invention-provided secondary file server. Shared file management of multiple file servers or intelligent hosts supporting a single or common high availability data storage system, for example a RAID drive, enables server or host fault tolerance to be provided for such a single highly, reliable data-storage system. Cross-mirroring in such a multi-host, high availability data storage system, along with secondary write suppression, enables server fault tolerance to be provided without significant redundant hardware costs and without a single point of failure in the data storage access paths.

Claims

1. A fault-tolerant computer network comprising:

a) a high availability permanent data-storage device maintaining a data image;

b) primary and secondary intelligent network hosts connected to and supporting said storage device;

c) multiple data-processing workstation connected in a local area network with both said hosts to access said data image maintained on said high-availability storage device via each said intelligent host independently; and

d) data-storage routing software loadable at said workstations and providing, for each said workstation, primary route addressing and secondary route addressing of said data image maintained on said high availability storage device via said primary or said secondary host respectively;

wherein said data-storage routing software comprises;

i) monitoring means to monitor data storage activities output from said workstation onto said computer network and to detect failure to access said high-availability storage device via a respective one of said intelligent hosts and

ii) rerouting data access requests for said high-availability storage device via a respective other of said intelligent hosts hosting said dam storage device in response to said access failure detection;

whereby workstation access to said dam image maintained on said high-availability storage device can continue transparently to the user despite loss of functionality of one of said intelligent hosts.

2. A fault-tolerant computer network according to claim 1 wherein said data-storage routing software monitoring means monitors operating system activity and interrupts data-related storage requests for said high-reliability storage device.

3. A fault-tolerant computer network according to claim 1 wherein said primary and secondary hosts can both write to common data storage areas on said high-availability data storage device.

4. A fault-tolerant computer system according to claim 1 wherein said high-availability data-storage device comprises a redundant array of inexpensive drives.

5. A fault-tolerant computer system according to claim 1 comprising a bus connecting each said host to said high-availability data-storage device.

6. A fault-tolerant computer system according to claim 1 comprising a number of workstations of from two to two hundred each connected for networked communication with said primary and secondary hosts under control of a network operating system.

7. A fault-tolerant computer system according to claim 6 wherein use of said high-availability storage device is divided between said intelligent hosts whereby said intelligent hosts share network processing duties relating to access of said high-availability storage device.

8. A fault-tolerant computer system according to claim 7 wherein said routing software comprises a setup protocol to effect division of use of said high-availability storage device by segregating said workstations into a first group routing access to said high availability storage device via said primary host and a second group routing access to said high availability storage device via said secondary host.

9. A fault-tolerant computer system according to claim 7 wherein said routing software comprises a setup protocol to effect division of use of said high-availability storage device by segregating said workstation activity into a first class of activity related to a first shared data type and into a second class of activity related to a second data type, said first class of data activity being routed via said primary intelligent host and said second class of data activity being routed via said secondary host.

19. A fault-tolerant computer network according to claim 1 wherein said rerouting software converts a command line secondary access route specification, prior to primary failure, to a corresponding primary access route specification.

10. A computer network management method for a local area computer network employing a data storage device maintaining a data image and capable of being supported by multiple intelligent network. hosts, said method comprising:

a) coupling at least a primary and a secondary intelligent host to said storage device in a local area network for access of said data image by multiple workstations via either one of said primary and secondary hosts; and

b) employing worlcstation loaded routing software at each said workstation, said routing software performing the functions of;

i) routing all workstation generated network data storage activity destined for said storage device to be segregated between first and second file categories for storage on said storage device through said primary and said secondary intelligent hosts respectively whereby, said intelligent hosts share data storage activities providing network access to said data storage device;

ii) monitoring means to monitor data storage activities output from said workstation onto said computer network and to detect failure to access said high-availability storage device via a respective one of said intelligent hosts; and

iii) rerouting data access requests for said high-availability storage device via a respective other of said intelligent hosts hosting said data storage device in response to said access failure detection;

whereby workstation access to said high-availability storage device can continue transparently to the user despite loss of functionality of one of said intelligent hosts.

11. A computer network management method according to claim 10 comprising routing all said workstation generated data storage activity via said primary intelligent host, performing maintenance or upgrade service functions on said secondary intelligent host, and rerouting data storage activity to be divided between said intelligent hosts after said service functions are completed.

20. A fault-tolerant computer network according to claim 10 wherein said rerouting software converts a command line secondary access route specification, prior to primary failure, to a corresponding primary access route specification.

12. A computer network management method employing on a computer network, a data storage device capable of being supported by multiple intelligent hosts, said method comprising:

a) coupling at least a primary and a secondary intelligent host to said storage device for access by one or more workstations; and

b) employing workstation loaded routing software said routing software performing the functions of;

i) routing all workstation generated network data storage activity destined for said storage device to be segregated between first and second file categories for storage on said storage device through said primary and said secondary intelligent hosts respectively whereby, said intelligent hosts share data storage activities providing network access to said data storage device;

ii) monitoring means to monitor data storage activities output from said workstation onto said computer network and to detect failure to access said high-availability storage device via a respective one of said intelligent hosts and

iii) rerouting data access requests for said high-availability storage device via a respective other of said intelligent hosts hosting said data storage device in response to said access failure detection;

whereby workstation access to said high-availability storage device can continue transparently to the user despite failure of one of said intelligent hosts.

13. A method according to claim 12 comprising physically connecting both intelligent hosts to a single high-availability, random-access, data-storage device to support access of multiple workstation users, each said workstation being loaded with software responsive to loss of activity of either active intelligent host to redirect data-related communication of said users with said commonly supported high-availability randomaccess data-storage device to be supported by the remaining active intelligent host.

14. A method according to claim 13 wherein said data storage device is a high availability multi-host supportable RAID drive.

15. A computer network workstation being one of a number of such workstations connected via a local area network with a primary file server providing access to a data image supported on a network data storage system. connected to a secondary file server to provide an alternative storage system access path loaded with rerouting software for providing redundant data accessibility across the local area network, said rerouting software performing the steps of:

a) monitoring all data output by said workstation to a primary server-supported network data storage system maintaining a data image;

b) detecting at said workstation failure to access said data storage system via said primary server; and

c) rerouting workstation-generated data storage activity to a current version of said data image via the secondary file server and the alternative storage system access path.

16. A computer network workstation according to claim 15 wherein said detection of access failure is effected by monitoring returns from said data storage system to said workstation.

17. A computer network workstation according to claim 15 wherein said a storage system access request is held in workstation RAM and rerouted in response to a bad return.

18. A computer network according to claim 15 having said rerouting software implemented at all network workstations using a specified shared data set, to effect similar data storage routing at each said workstation, to ensure data integrity.