Device and method for supervising use of shared storage by multiple caching servers

• US 20030195940A1
• Filed: 04/04/2002
• Published: 10/16/2003
• Est. Priority Date: 04/04/2002
• Status: Active Grant

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Kumar, Rajendra, Basu, Sujoy
• Granted Patent: US 6,868,439 B2
• Time in Patent Office: Days
• US Class Current: 709/213
• CPC Class Codes: G06F 16/9574 of access to content, e.g. by caching

Abstract

A special-purpose appliance (SPA) works in conjunction with a server farm consisting of multiple caching server appliances (CSAs) to supervise a local storage medium (i.e., a shared cache) that is accessible by all the CSAs for storing at least some of the remote objects such as web pages and their embedded objects and/or streaming media objects that have been and/or will be served by one or more of the CSAs to its respective clients. The SPA preferably also determines when to prefetch remote objects such as web pages and their embedded objects and/or streaming media objects that are not currently stored in the shared cache, but which the SPA has determined are likely to be requested in the future by one or more of the CSAs one behalf of one or more of the CSA'"'"'s respective clients. In that regard, the SPA (and/or PSA) does not merely monitor the file requests from each CSA to the remote servers, but rather monitors and aggregates the individual requests from each client to its respective CSA, for example, by monitoring the access logs of each CSA and using that data to decide what to prefetch into the shared cache from the remote server or servers, what is still of value and needs to be updated, and what is no longer of value and can be replaced. What it prefetches can be based, for example, on links present in an already requested web page, on patterns of recent accesses to web pages and streaming media objects, on user profiles, and on past trends.

Claims

1. A caching system for local storage of objects originating from one or more remote origin servers, comprising:

a cluster of caching server appliances for retrieving objects requested by clients from a storage hierarchy of sources including dedicated local storage, shared local storage, and remote servers, and for retaining cached copies of selected said objects in said dedicated local storage and said shared local storage;

a plurality of dedicated local stores, each associated with a respective said caching server appliance for providing a respective said dedicated local storage to said respective caching server appliance;

a shared local store accessible to all said caching server appliances for providing said shared local storage; and

a supervisor appliance responsive to all object requests from said clients for determining when a particular object stored in said shared local store should be replaced with another object.

2. The system of claim 1, further comprising

a shared prefetching server appliance capable of anticipating at least some future requests from more than one of said clients, wherein the supervisor appliance replaces selected objects in the shared local store based at least in part on said anticipated future requests.

3. The system of claim 2, wherein

the shared prefetching server appliance anticipates at least some of the future requests by monitoring all the past requests from said clients.

4. The system of claim 2, wherein

the shared prefetching server appliance anticipates at least some of the future requests by monitoring predetermined expiration dates of objects stored in said shared local storage.

5. The system of claim 2, wherein

the shared prefetching server appliance anticipates at least some of the future requests by examining the objects stored in said shared local storage for links to other objects.

6. The system of claim 2, wherein

the shared prefetching server appliance monitors the requests received by all of the caching server appliances from their respective clients for objects in the shared storage that are being requested only infrequently

7. The system of claim 2, wherein

the shared prefetching server appliance replaces those objects with other objects that are being requested more frequently.

8. The system of claim 2, wherein

the shared prefetching server appliance is integrated with one of the caching server appliances.

9. The system of claim 2, further comprising

a traffic director appliance.

10. The system of claim 9, wherein

the traffic director appliance routes the client requests to a selected caching server appliance based at least in part on prior requests for that same object.

11. The system of claim 9, wherein

the supervisor appliance monitors said requests based on information received from said traffic director appliance.

12. The system of claim 1, wherein

a single conventional origin server API is provided between each said origin server and said cluster, thereby providing the origin server with a single virtual caching appliance.

13. The system of claim 1, wherein

a single conventional client API is provided between each of said clients and said cluster, thereby providing the client with a single virtual caching appliance.

14. The system of claim 1, wherein

a single conventional network manager API is provided between an external network manager and said cluster, thereby providing the network manager with a single virtual caching appliance.

15. The system of claim 1, wherein

said shared local store is part of a storage area network.

16. The system of claim 1,

wherein said shared local store is network attached storage that is connected to the caching server appliances by a local area network.

17. A method for caching objects originating from one or more remote servers, comprising:

retrieving at least two different objects requested by at least two different clients from remote servers,

caching first copies of each of two different retrieved objects in respective different local stores;

caching second copies of the two different retrieved objects in the same shared store; and

independently determining when each of said first and second copies should be replaced with other cached objects.

18. The method of claim 17, wherein

the second copies are placed in the shared storage when the CSAs periodically do a writeback of the first copies in their local cache.

19. The method of claim 17, further comprising

analyzing past requests from said at least two clients to anticipate at least some future requests from said clients and,

using said anticipated future requests to determine when the cached second copies in the shared store should be replaced.

20. The method of claim 19, wherein

all the past requests from said clients are monitored.

21. The method of claim 17, wherein

predetermined expiration dates of objects stored in said shared local store are monitored.

22. The method of claim 17, wherein

the objects stored in said shared local store are monitored for links to other objects.

23. The method of claim 17, wherein

all the requests from said clients are monitored for objects in the shared store that are being requested only infrequently

24. The method of claim 23, wherein

objects in the shared store that are being requested only infrequently are replaced with other objects that are being requested more frequently.

25. The method of claim 17, further comprising

routing a client'"'"'s request for a particular object to a server associated with a particular said dedicated local store based at least in part on prior requests from another client for that same object.

26. The method of claim 17, wherein

a single common origin server API is used for all said requests to a given said origin server.

27. The method of claim 17, wherein

a single common client API is used for all requests from a given said client.