Transparent redirection and load-balancing in a storage network

• US 7,599,941 B2
• Filed: 07/25/2006
• Issued: 10/06/2009
• Est. Priority Date: 07/25/2005
• Status: Active Grant

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Parascale Incorporated (Hitachi, Ltd.)
• Inventors: Bahar, Cameron, Hopfield, Joseph, Nalam, Naveen, Zafman, David B., Oskouy, Rasoul M.
• Primary Examiner(s): Robinson, Greta L
• Assistant Examiner(s): Wilcox, James J
• Application Number:
  US11/459,905
• Publication Number:
  US 20070022121A1
• Time in Patent Office: 1,169 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06F 16/184 implemented as replicated file system
• CPC Class Codes: G06F 2209/508 Monitor
• CPC Class Codes: G06F 9/5011 the resources being hardware resources other than CPUs, Servers and Terminals
• CPC Class Codes: G06F 9/5016 the resource being the memory
• CPC Class Codes: H04L 61/50 Address allocation
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04L 67/1008 based on parameters of servers, e.g. available memory or workload monitoring of computer activity G06F11/30
• CPC Class Codes: H04L 67/1017 based on a round robin mechanism
• CPC Class Codes: H04L 67/1029 using data related to the state of servers by a load balancer
• CPC Class Codes: H04L 67/1095 Replication or mirroring of data, e.g. scheduling or transport for data synchronisation between network nodes
• CPC Class Codes: H04L 67/1097 for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
• CPC Class Codes: H04L 67/563 Data redirection of data network streams
• CPC Class Codes: Y10S 707/99931 Database or file accessing
• CPC Class Codes: Y10S 707/99952 Coherency, e.g. same view to multiple users
• CPC Class Codes: Y10S 707/99953 Recoverability

Abstract

A storage system includes a plurality of storage servers that store a plurality of files, a monitor module, and a redirector module. The monitor module monitors usage information associated with the plurality of storage servers. The redirector module determines, based on the monitored usage information, a storage server in the plurality of storage servers to service a session from a client. The redirector module then instructs the client to establish the session with the determined storage server.

Claims

1. A computer-implemented method of managing client requests to a file system distributed among a plurality of storage servers each including a processor, the method comprising:

initiating a client session for a client request for accessing one or more files in the file system by connecting to a virtual IP address associated with a master server of the plurality of storage servers, the master server elected by the plurality of storage servers to advertise the virtual IP address;

generating a client to storage server binding decision based on file system to storage server mappings, statistics of a statistics database of the master server, the statistics collected by the plurality of storage servers and grouped into categories including server statistics relating to server load statistics, client statistics relating to client information, and storage statistics relating to file system and input/output (IO) configuration statistics, network parameters associated with the client, and a type of file system being accessed;

redirecting the client session to one of the plurality of storage servers based on the generated binding decision and without modification of client network file protocols used by the client, including achieving the client session redirection for a network file system (NFS) providing NFS access by manufacturing an appropriate symbolic link and forcing the client'"'"'s automounter to establish a connection to the selected storage server, for a common internet file system (CIFS) providing CIFS access by masquerading as a distributed file system (DFS) root server and returning a DFS junction redirect forcing the client to establish a connection to the selected storage server, and for a hypertext transfer protocol providing HTTP access by returning an HTTP redirect forcing the client to establish a connection to the selected storage server.

9. The method of claim 1, further comprising associating storage classes with one or more storage volumes of the plurality of storage servers, composing a set of rules that map file characteristics into a storage class, applying, using the master server, the set of rules to the one or more files to obtain one or more storage classes associated with each file in the one or more files, and matching, using the master server, the obtained one or more storage classes to the one or more storage volumes to determine an appropriate storage volume for each file.

10. The method of claim 9, wherein at least one rule in the set of rules map the file to the storage class, the least one rule in the set of rules maps to replication factors for the file, using the at least one rule by the master server to determine if the file should be replicated, determining on which volumes replicas of the file should be stored, and generating a task to perform replication of the file on the determined volumes.

11. The method of claim 1, further comprising receiving a first input in the master server from a user that associates storage classes with one or more volumes of the plurality of storage servers, receiving a second input in the master server from the user to create a set of rules that include criteria for identifying a file and defining a storage class that satisfies the criteria, each rule including a storage class specification having a replication factor, at least one storage class, and which is modified by at least one volume preference qualifier, receiving in the master server a list of files to be created, applying using the master server, the set of rules to a first file of the list of files until a rule of the set of rules is satisfied to yield one or more storage classes associated with the first file, and matching, using the master server, the obtained one or more storage classes to the one or more storage volumes to determine an appropriate storage volume for the first file.

12. The method of claim 11, further comprising using at least one rule of the set of rules to determine if the first file should be replicated, determining on which of the storage volumes that the replicated first file is to be stored, and generating a task to perform replication of the first file on the determined storage volume.

13. The method of claim 11, further comprising using at least one rule of the set of rules to determine if the first file is to be migrated from a first storage volume to a second storage volume, and generating a task to perform migration of the first file from the first storage volume to the second storage volume.

2. A storage system comprising:

a file system distributed among a plurality of storage servers each including at least one processor and at least one storage medium to store files;

a monitor module associated with each storage server, each monitor module being configured to collect statistics regarding the associated storage server;

wherein one of the storage servers is a master server to receive the collected statistics from each storage server and includes a redirector module configured to redirect a client session from a client to one of the plurality of storage servers based on network parameters associated with the client, the file system being accessed, and the collected statistics so as to efficiently distribute client sessions among the plurality of storage servers, the collected statistics grouped into categories including server statistics relating to server load statistics, client statistics relating to client information, and storage statistics relating to file system and input/output (IO) configuration statistics, and wherein the client session redirection includes for a network file system (NFS) providing NFS access manufacturing an appropriate symbolic link and forcing the client'"'"'s automounter to establish a connection to the selected storage server, for a common internet file system (CIFS) providing CIFS access masquerading as a distributed file system (DFS) root server and returning a DFS junction redirect forcing the client to establish a connection to the selected storage server, and for a hypertext transfer protocol providing HTTP access returning an HTTP redirect forcing the client to establish a connection to the selected storage server.

3. The storage system of claim 2 wherein each monitor module is further configured to communicate the collected statistics to the master server.

4. A system comprising:

a control node comprising a first server having a processor and a first storage to store information associated with files of a file system, the information including namespace, metadata, an object identifier, and location information; and

a plurality of storage nodes each comprising a second server coupled to the control node, each storage node including a processor and a second storage to store the files of the file system in an object storage, wherein the files are stored separately from the metadata, wherein at least one of the storage nodes is a master node to receive a client request for access to a file of the file system and to select one of the plurality of storage nodes to handle the client request based at least in part on statistics of the plurality of storage nodes, the statistics including server statistics relating to server load statistics, client statistics relating to client information, and storage statistics relating to file system and input/output (IO) configuration statistics, and to redirect the client request to the selected storage node, and wherein the selected storage node is to communicate with the control node to determine an object identifier and location information associated with the file, and to access the file from an object storage of the selected storage node if the object identifier and the location information indicates that the object storage of the selected storage node includes the file, and to otherwise request the file from a different storage node based on the object identifier and the location information, to service the request from the client.

5. The system of claim 4, wherein the selection of the one of the storage nodes for handling the client request is further based on network parameters associated with the client.

6. The system of claim 5, wherein the selection of the one of the storage nodes is further based on the file system being accessed.

7. The system of claim 4, wherein the master node is elected by the plurality of storage nodes to advertise a virtual IP address for use by the client.

8. The system of claim 4, further comprising achieving a client session redirection for a network file system (NFS) providing NFS access by manufacturing an appropriate symbolic link and forcing the client'"'"'s automounter to establish a connection to the selected storage server, for a common internet file system (CIFS) providing CIFS access by masquerading as a distributed file system (DFS) root server and returning a DFS junction redirect forcing the client to establish a connection to the selected storage server, and for a hypertext transfer protocol providing HTTP access by returning an HTTP redirect forcing the client to establish a connection to the selected storage server.