Granted Patent B2

US 8,713,127 · App. 12/025,999 · Granted Apr 29, 2014

Inventor: Hitesh Menghnani (Bangalore, IN)

Assignee: Novell, Inc.

Abstract

Claims (21)

1. A machine-implemented method that executes on a storage aggregation server, comprising:

identifying, by the storage aggregation server, a plurality of contributing servers over a network connection, which is an Internet connection, each of the contributing servers have excess storage capacity that is collectively shared over the network connection as storage organized in blocks, the contributing servers identified, supplied, configured, and registered by a user via a configuration and management server interface;

exchanging, by the storage aggregation server, configuration information with each of the contributing servers, each contributing server accesses the excess storage capacity, partitions the excess storage capacity, and communicates real-time events with the other contributing servers based on the configuration information, each contributing server implemented as a separate machine and dispersed over the network from the other contributing servers;

logically creating, by the storage aggregation server, logical Redundant Arrays of Independent Disks (RAID) by aggregating the excess storage capacity over the Internet, the logical RAID is accessible via Input/Output (I/O) operations over the Internet using an Internet Small Computer Systems Interface (iSCSI) protocol and appears to be a physical RAID;

maintaining, by the storage aggregation server, the logical RAIDS in three pools, a first pool for dedicated contributing servers and a second pool for part-time contributing servers, and wherein the user assigns a role to a particular server that identifies that particular server's pool assignment; and

using, by network resources over the network connection and via the iSCSI protocol, the excess storage capacity as if it were another physical RAID when in fact the excess storage capacity is dispersed over the network and managed by the storage aggregation server in cooperation with the contributing servers as the logical RAID, the excess storage capacity shared at the block level and not just the file level.

2. The method of claim 1 , wherein identifying further includes receiving an Internet Protocol (IP) address for each of the contributing servers from the user via the configuration and management interface.

3. The method of claim 1 , wherein exchanging further includes establishing a different contributing amount of storage from each of the contributing servers.

4. The method of claim 3 , wherein exchanging further includes setting aside a reserve amount of storage from each contributed amount of storage on each contributing server for failure recovery operations in the event that one of the contributing servers fails, and wherein the logical RAID is created as logical RAID 5.

5. A machine-implemented method, comprising:

registering a configured size of storage as blocked storage with a distributed storage aggregation server, the distributed storage aggregation server is located over a network and is remote and external from the blocked storage, the network is the Internet and the block storage associated with sharing storage over the Internet at the block level and not just a file level, and the block storage assigned to being shared on a dedicated or part time basis, and the block storage registered by a user via a configuration and management interface by the user assigning a role to the block storage by selecting from: the dedicated basis and the part time basis;

partitioning the blocked storage into predefined sizes of shared partitions in response to interaction with the distributed storage aggregation server;

presenting references to the shared partitions to the distributed storage aggregation server for subsequent use by network resources;

organizing, by the distributed storage aggregation server, the shared partitions with other shared partitions from other contributing servers to form logical Redundant Arrays of Independent Disks (RAID) that the network resources access over the Internet using an Internet Small Computer Systems Interface (iSCSI) protocol performing Input/Output (I/O) operations on; and

using, by the network resources over the network and via the iSCSI protocol, the shared partitions as if it were another physical RAID when in fact the shared partitions are dispersed over the network and managed by the distributed storage aggregation server in cooperation with the other contributing servers.

6. The method of claim 5 , wherein registering further includes establishing an interval for sending heartbeat communications to the distributed storage aggregation server and establishing types of events for which notification is to be sent to the distributed storage aggregation server.

7. The method of claim 5 , wherein registering further includes notifying the distributed storage aggregation server as to a type associated with blocked storage during registration, wherein the type includes one of the following: a dedicated storage type and a temporary or part-time usage storage type.

8. The method of claim 5 , wherein partitioning further includes setting aside a reserve amount of the blocked storage for use by the distributed storage aggregation server for disaster recovery.

9. The method of claim 5 , wherein presenting further includes maintaining a block map that maps the references presented to specific blocks within the partitioned blocked storage, and wherein the block map is used to process the I/O operations received from the network resources via the iSCSI protocol.

10. The method of claim 5 , further comprising, permitting the user via the configuration and management interface to manually remove portions of the blocked storage from use within the logical RAID.

11. The method of claim 5 further comprising, processing a received I/O operation from a particular network resource against the partitioned blocked storage.

Assignments (16)

Continuity (1)