Method of dynamically balancing workload of a storage system

• US 7,886,299 B2
• Filed: 02/14/2005
• Issued: 02/08/2011
• Est. Priority Date: 06/01/2004
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Hitachi, Ltd.
• Inventors: Kitamura, Manabu
• Primary Examiner(s): Wu; Qing
• Assistant Examiner(s): Truong; Camquy
• Application Number:
  US11/058,130
• Publication Number:
  US 20050267929A1
• Time in Patent Office: 2,185 Days
• Field of Search: 711/114, 709223-226, 718104-105, 718/102
• US Class Current: 718/104
• CPC Class Codes: G06F 2206/1012 Load balancing
• CPC Class Codes: G06F 3/061 Improving I/O performance
• CPC Class Codes: G06F 3/0631 by allocating resources to storage systems
• CPC Class Codes: G06F 3/0643 Management of files
• CPC Class Codes: G06F 3/0647 Migration mechanisms
• CPC Class Codes: G06F 3/067 Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]
• 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 67/101 based on network conditions
• CPC Class Codes: H04L 67/1012 based on compliance of requirements or conditions with available server resources
• CPC Class Codes: H04L 67/1023 based on a hash applied to IP addresses or costs
• CPC Class Codes: H04L 67/1034 Reaction to server failures by a load balancer

Abstract

A storage system comprises a plurality of processors, a plurality of volumes, and interconnecting means to connect the plurality of processors and the plurality of volume to each other. The storage system detects and determines the cause of the performance bottleneck, wherein the cause may be one of plurality of reasons. The storage system applies an appropriate load balancing method according to the determined cause of the performance bottleneck.

Claims

1. A storage system coupled to a plurality of clients via a communication link, the storage system comprising:

a first file server including a first network interface and a second network interface, said first network interface being assigned a first group of Internet Protocol (IP) addresses and said second network interface being assigned a second group of IP addresses;

a second file server including a third network interface assigned a third group of IP addresses;

a plurality of disks providing a plurality of storage area;

a storage unit including a storage area of first type that is assigned to the first network interface, a storage area of second type that is assigned to the second network interface, the storage area of first type including a first group of file systems, the storage area of second type including a second group of file systems, each file system being associated with an IP address; and

a memory comprising a program to balance a workload, the program comprising, code for determining whether or not a disk workload of a first group of disks of the plurality of disks is heavy;

code for balancing the disk workload by distributing work to other disks to reduce the heavy disk workload of the first group of disks if the disk workload is determined to be heavy; and

code for balancing a processor workload of a processor of the first file server by distributing work to other processors if the processor workload is determined to be heavy in order to reduce the heavy processor workload of the processor;

wherein the code for balancing the processor workload includes;

code for detaching a first file system and first IP address which is associated with said first file system if it is determined that the workload of the processor of the first file server is heavy, the first file system being part of the first group of file systems; and

code for attaching said first file system to the second file server, said first IP address being assigned to the third network interface, so that the first file system may be accessed via the second file server but not from the first file server.

2. The storage system of claim 1, wherein the code for balancing the disk workload includes code for determining whether or not an unused logical device exists in a second group of disks in the storage system;

code for migrating a logical device associated with the first group of disks to the second group of disks if the second group of disks is determined to have at least one unused logical device.

3. The storage system of claim 2, wherein the code for balancing the disk workload further includes:

code for determining the least busy group of disks from a plurality of groups of disks in the storage system, wherein the least busy group of disks is the second group of disks.

4. The storage system of claim 3, wherein the first and second groups of disks are first and second redundant arrays of inexpensive disks (RAID), respectively.

5. The storage system of claim 1, wherein the first group of disks is a redundant array of inexpensive disks (RAID).

6. The storage system of claim 1, wherein the storage area of first type includes first and second file systems which are associated with first and second IP addresses, respectively, and the storage area of second type includes third and fourth file systems, wherein the program further comprises:

code for determining whether or not a workload of the first network interface is heavy; and

code for reassigning the first file system from the storage area of first type to the storage area of second type if the workload of the first network interface is determined to be heavy.

7. The storage system of claim 6, wherein the program further comprises:

code for determining whether or not reassigning the first file system from the storage area of first type to the storage area of second type would improve the workload of the first network interface without causing the second network interface to experience bottleneck.

8. The storage system of claim 7, wherein the code for balancing the processor workload is executed if it is determined that the second network interface would experience bottleneck if the reassignment of the first file system is carried out.

9. The storage system of claim 1,

wherein the first group of disks is located in one of the file servers;

and

wherein the code for balancing the disk workload includes;

code for determining whether or not an unused logical device exists in a second group of disks in the same file server as the first group of disks; and

code for migrating a logical device associated with the first group of disks to the second group of disks if the second group of disks is determined to have at least one unused logical device.

10. The storage system of claim 9, wherein the code for balancing the disk workload further includes:

code for determining the least busy group of disks from a plurality of groups of disks located in the same file server as the first group of disks, wherein the least busy group of disks is the second group of disks.

11. A storage system coupled to a plurality of clients via a communication link, the storage system comprising:

a first file server including a first network interface and a second network interface, said first network interface being assigned a first group of Internet Protocol (IP) addresses and said second network interface being assigned a second group of IP addresses;

a second file server including a third network interface assigned a third group of IP addresses;

a plurality of disks providing a plurality of storage areas;

a storage unit including a storage area of first type that is assigned to the first network interface, a storage area of second type that is assigned to the second network interface, the storage area of first type including a first group of file systems, the storage area of second type including a second group of file systems, each file system being associated with an IP address, wherein the storage area of first type includes first and second file systems which are associated with first and second IP addresses, respectively, and the storage area of second type includes third and fourth file systems; and

disks is heavy, the first memory being provided within a console that is remotely located from the first and second file servers.

12. The storage system of claim 11, wherein the program further comprises:

code for balancing a processor workload of a processor of the first file server if the processor workload is determined to be heavy.

13. The storage system of claim 11, wherein the memory includes a first memory comprising the code for determining whether or not the disk workload of the first group of a first group of file systems, the storage area of second type including a second group of file systems, each file system being associated with an IP address,

wherein the method comprises:

determining whether or not the storage system is experiencing a performance degradation;

determining a cause of the performance degradation, wherein the determined cause can be a first cause, a second cause, or a third cause; and

applying an appropriate load balancing method according to the determined cause of the performance degradation,

wherein the appropriate load balancing method is a first balancing method if the determined cause is the first cause, a second balancing method of the determined cause is the second cause, and a third balancing method if the determined cause is the third cause; and

wherein the first balancing method involves balancing a disk workload by distributing work from a first group of disks to other disks to reduce a heavy disk workload of the first group of disks of the plurality of disks, and the second balancing method involves balancing a processor workload of a processor of the first file server by distributing work to other processors in order to reduce a heavy processor workload of the processor, and the third balancing method involves balancing a LAN workload;

wherein the second balancing method includes;

detaching a first file system and first IP address which is associated with said first file system if it is determined that the workload of the processor of the first file server is heavy, the first file system being part of the first group of file systems; and

attaching said first file system to the second file server, said first IP address being assigned to the third network interface, so that the first file system may be accessed via the second file server but not from the first file server.

14. The storage system of claim 13, wherein the memory includes a second memory comprising the code for balancing the disk workload if the disk workload is determined to be heavy, the second memory being provided within one of the first and second file servers.

15. A method for balancing workload of a storage system including first and second file servers and a storage unit, which includes a plurality of storage areas provided by a plurality of disks,

wherein the file servers include a first file server including a first network interface and a second network interface, said first network interface being assigned a first group of Internet Protocol (IP) addresses and said second network interface being assigned a second group of IP addresses, wherein the second file server includes a third network interface assigned a third group of IP addresses,

wherein the storage unit includes a storage area of first type that is assigned to the first network interface, a storage area of second type that is assigned to the second network interface, the storage area of first type including

a memory system comprising a program to balance a workload, the program comprising,

code for determining whether or not a disk workload of a first group of disks of the plurality of disks is heavy;

code for balancing the disk workload by distributing work to other disks to reduce the heavy disk workload of the first group of disks if the disk workload is determined to be heavy;

code for determining whether or not a workload of a first network interface is heavy; and

code for reassigning the first file system from the storage area of first type to the storage area of second type if the workload of the first network interface is determined to be heavy.

16. The method of claim 15, wherein the first balancing method includes:

determining whether or not an unused logical device exists in a second group of disks in the storage system;

migrating a logical device associated with the first group of disks to the second group of disks if the second group of disks is determined to have at least one unused logical device.

17. The method of claim 16, wherein the first balancing method includes:

determining the least busiest group of disks from a plurality of groups of disks in the storage system, wherein the least busiest group of disks is the second group of disks.

18. The method of claim 15, wherein the determining steps are performed within one of the file servers or a management device coupled to the file servers, the management device being provided external to the file servers.