Processing system with dynamically allocatable buffer memory

• US 6,081,883 A
• Filed: 12/05/1997
• Issued: 06/27/2000
• Est. Priority Date: 12/05/1997
• Status: Expired due to Term

• Current Assignee: Network Appliance Incorporated (NetApp, Inc.)
• Original Assignee: Auspex Systems, Inc.
• Inventors: Repakula, Murali Sundaramoorthy, Popelka, Paul, Row, Edward John, Craft, Peter Kingsley, Del Fante, Paul Brian, Bodas, Amod Prabhakar, Higgen, David Allan, McCutcheon, Leslie Thomas, Sterk, Donald Wayne, Philbrick, Clive Mathew, Jones, Daniel Murray, Walter, Richard Allen, Tripathy, Tarun Kumar, Narayanaswamy, Lakshman
• Primary Examiner(s): Pan, Daniel H.
• Assistant Examiner(s): Nguyen, Dzung C.
• Application Number:
  US08/985,996
• Time in Patent Office: 935 Days
• Field of Search: 395/800.01, 395/800.03, 395/800.23, 395/800.28, 395/800.29, 395/800.3, 395/800.32, 395/800.36, 712/1, 712/3, 712/23, 712/28, 712/29, 712/30, 712/32, 712/36, 711/111, 711/117, 711/119, 711/120, 711/121, 711/131
• US Class Current: 712/28
• CPC Class Codes: G06F 12/0813 with a network or matrix configuration
• CPC Class Codes: G06F 16/10 File systems; File servers

Abstract

A scalable computer system has an interconnect bus providing communication links among a host processor and one or more function-specific processors, including a network processor (NP) and a file storage processor (FSP). The host processor provides a single interface to network administrators for maintaining the system. A bi-endian conversion system is provided to minimize a need for translating between big and little endian data types generated by diverse processors. The NP shares a single memory image with other processors and has a buffer memory for buffering requests from the network interfaces. The buffer memory has one or more segments which are dynamically allocatable to different processors. The FSP has a metadata cache for maintaining information on data being cached in the NP buffer memory. The FSP also has a write cache for buffering file write operations directed at disks. Upon receiving requests for data from the NP, the FSP checks the metadata cache to see if a copy of the requested data has been cached in the NP buffer and, if the copy exists in the NP buffer, causing the NP with the data to respond to the request. The resulting scalable computer provides higher data availability, faster access to shared data, and reduced administrative costs via data consolidation.

Claims

1. A scalable computer system, comprising:

an interconnect bus;

a host processor coupled to the interconnect bus; and

one or more function-specific processors including network processor, a file processor and a storage processor, the one or more function-specific processors coupled to the interconnect bus for communicating with other processors, each function-specific processor having a buffer memory coupled thereto, the buffer memory having one or more segments which serve as function-specific caches to cache function specific data including network processing data, file processing data and storage processing data, the segments being dynamically allocatable to different processors.

2. The computer system of claim 1, wherein one of the function-specific processor includes:

a network processor coupled to the interconnect bus;

one or more network interfaces coupled to the network processor; and

a read cache coupled to the network processor for buffering requests from the network interfaces.

3. The computer system of claim 1, wherein one of said function-specific processor includes:

a file processor coupled to the interconnect bus;

a storage processor coupled to the file processor; and

one or more data storage devices coupled to the storage processor.

4. The computer system of claim 3, further comprising a metadata cache coupled to the file processor.

5. The computer system of claim 3, further comprising a write cache coupled to the storage processor.

6. The computer system of claim 1, wherein one of said function-specific processor includes:

a file storage processor (FSP) coupled to the interconnect bus;

one or more data storage devices coupled to the FSP, each data storage device having files and associated metadata;

a metadata cache coupled to the FSP for buffering the metadata; and

a write cache coupled to the FSP for buffering file write operations from the FSP.

7. The computer system of claim 6, further comprising first and second FSP buses.

8. The computer system of claim 7, further comprising a bridge coupling the interconnect bus to the first FSP bus.

9. The computer system of claim 8, further comprising a write buffer coupled to first and second FSP buses.

10. The computer system of claim 9, wherein the write buffer further comprises:

a memory array for buffering writes to the data storage devices; and

an exclusive OR engine coupled to the memory array for performing redundant array of inexpensive disk (RAID) operations.

11. The computer system of claim 1, wherein the interconnect bus is a Scalable Coherent Interface (SCI) bus.

12. The computer system of claim 1, further comprising one or more network processors and one or more file storage processors, each of the network processors capable of communicating directly with each of the file storage processors.

13. The computer system of claim 12, further comprising:

a data storage device coupled to each file storage processor; and

a bridge coupled to the interconnect bus and to the data storage device for forwarding requests from the network processors to the respective data storage device.

14. The computer system of claim 1, wherein the host computer supports a Unix operating system.

15. The computer system of claim 1, wherein the host computer supports a Windows-NT operating system.

16. The computer system of claim 1, wherein the computer system has a big endian processor and a little endian processor, further comprising a bi-endian messaging system for interprocessor communication.

17. A computer system, comprising:

an interconnect bus;

a host processor coupled to the interconnect bus;

a network processor (NP) coupled to the interconnect bus and sharing a single memory image with other processors, including;

one or more network interfaces coupled to the NP; and

a buffer memory coupled to the NP for buffering requests from the network interfaces, the buffer memory having one or more segments which are dynamically allocatable to different processors; and

a file storage processor (FSP) coupled to the interconnect bus, including;

one or more data storage devices coupled to the FSP, each data storage device having files and associated metadata;

a metadata cache coupled to the FSP for buffering the metadata, the metadata cache maintaining information on data stored on the NP buffer memory; and

a write cache coupled to the FSP for buffering file write operations from the FSP,

where the FSP, upon receiving requests for data from the NP, checks the metadata cache to see if a copy of the requested data has been cached in the NP buffer and, if the copy exists in the NP buffer, causing the NP with the data to respond to the request.

18. The computer system of claim 17, wherein the FSP further comprises first and second FSP buses.

19. The computer system of claim 18, further comprising a bridge coupling the interconnect bus to the first FSP bus.

20. The computer system of claim 18, further comprising a write buffer coupled to first and second FSP buses.

21. The computer system of claim 20, wherein the write buffer further comprises:

a memory array for buffering writes to the data storage devices; and

an exclusive OR engine coupled to the memory array for performing redundant array of inexpensive disk (RAID) operations.

22. The computer of claim 17, wherein the interconnect bus is a Scalable Coherent Interface (SCI) bus.

23. The computer system of claim 17, wherein the host computer supports a Unix operating system.

24. The computer system of claim 17, wherein the host computer supports a Windows-NT operating system.

25. The computer system of claim 17, wherein the computer system has a big endian processor and a little endian processor, further comprising a bi-endian messaging system for interprocessor communication.

26. The computer system of claim 17, wherein the FSPs collectively provide a single memory image with a memory management that allows a transfer of authority to another FSP and with distributed data content management.

27. The computer system of claim 17, wherein the computer system provides a single system image (SSI) to an application or to a client.

28. The computer system of claim 17, wherein the SSI is used in performing system administration.

29. The computer system of claim 17, wherein data is transferred between data storage devices without host intervention.

30. The computer system of claim 17, wherein the computer system continues to operate when the host processor fails.

31. The computer system of claim 17, wherein the computer system continues to operate when one of the network processors fails.

32. The computer system of claim 17, wherein the computer system continues to operate when one of the file storage processor fails.

33. The computer system of claim 17, wherein the network processor and the file storage processor are mounted on a single board with a shared cache coherent memory and shared I/O device.

34. The computer system of claim 32, wherein the FSP controls the NP buffer memory.

35. The computer system of claim 17, wherein the interconnect bus is an expansion bus on a board.

36. The computer system of claim 17, wherein each of the NP has a mount table for FSP routing purposes.

37. The computer system of claim 17, wherein the NP examines the request and directly forward the request to a server based on predetermined criteria.

38. The computer system of claim 17, wherein the host processor accesses the data storage devices on the FSP as being virtually connected to the host processor.

39. The computer system of claim 17, further comprising a user application, wherein the application may directly access the data storage device of the FSP.

41. The method of claim 39, wherein the data request arrives at a first NP and wherein the copy of the requested data exists in the buffer of a second NP, further comprising the step of instructing the second NP to respond to the request.

42. The method of claim 39, further comprising the step of storing data in a bi-endian format in the host processor to minimize conversions between big and little endian data types.

44. The method of claim 42, wherein the data request arrives at a first function-specific processor and wherein the copy of the requested data exists in a second function specific processor, further comprising the step of instructing the second processor to respond to the request.

45. The method of claim 42, further comprising the step of storing data in a bi-endian format in the host processor to minimize conversions between big and little endian data types.

40. A method for operating scalable computer system having an interconnect bus;

a host processor coupled to the interconnect bus;

a network processor (NP) coupled to the interconnect bus and sharing a single memory image with other processors, each NP having one or more network interfaces coupled to the NP and a buffer memory coupled to the NP for buffering requests from the network interfaces, the buffer memory having one or more segments which are dynamically allocatable to different processors; and

a file storage processor (FSP) coupled to the interconnect bus with one or more data storage devices coupled to the FSP, each data storage device having files and associated metadata, a metadata cache coupled to the FSP for buffering the metadata, the metadata cache maintaining information on data stored on the NP buffer memory, and a write cache coupled to the FSP for buffering file write operations from the FSP, comprising;

receiving a data request from the NP;

checking the metadata cache to see if a copy of the requested data is cached in the NP buffer; and

if the copy exists in the NP buffer, transmitting the copy from the NP buffer in response to the request and otherwise accessing the data storage device in accordance with the request.

43. A method for operating a scalable computer system having an interconnect bus, a host processor coupled to the interconnect bus, and one or more function-specific processors coupled to the interconnect bus for communicating with other processors, each function-specific processor having a buffer memory coupled thereto, the buffer memory having one or more segments which serve as function-specific caches and which are dynamically allocatable to different processors, comprising:

receiving a data request;

checking if a copy of the requested data is cached in one of the function-specific processor; and

if the copy exists in one of the function-specific processor, transmitting the copy in response to the request and otherwise accessing the data storage device in accordance with the request.