Storage system and performance management method of storage system

• US 8,949,559 B2
• Filed: 03/25/2011
• Issued: 02/03/2015
• Est. Priority Date: 03/25/2011
• Status: Expired due to Fees

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Tanaka, Toru, Nakajima, Noriko, Amano, Takashi, Kaneda, Yasunori
• Primary Examiner(s): Bragdon, Reginald G.
• Assistant Examiner(s): LILLIS, BRENDAN S
• Application Number:
  US13/123,996
• Publication Number:
  US 20120246425A1
• Time in Patent Office: 1,411 Days
• Field of Search: 711/162
• US Class Current: 711/162
• CPC Class Codes: G06F 3/0611 in relation to response time
• CPC Class Codes: G06F 3/0623 in relation to content
• CPC Class Codes: G06F 3/0647 Migration mechanisms
• CPC Class Codes: G06F 3/065 Replication mechanisms
• CPC Class Codes: G06F 3/0664 at device level, e.g. emulation of a storage device or system
• CPC Class Codes: G06F 3/067 Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]
• CPC Class Codes: G06F 3/0689 Disk arrays, e.g. RAID, JBOD

Abstract

It is an object to use a storage region in an efficient manner and maintain a performance of a storage system. A volume group GR11 includes reference volumes 1411, 1413, and 1415 and source volumes 1412, 1414, 1416, and 1417 that derive from the reference volumes. Each of the volumes stores only the difference data from a volume that is a parent thereof. A volume that is a cause of a problem is moved or copied from a pool of the moment 1410 to other pool 1420. A configuration of a volume group is modified.

Claims

1. A storage system comprising a storage apparatus and a management apparatus, the storage apparatus comprising:

a volume group including a plurality of logical volumes, the logical volumes including one or more first logical volumes to store reference data and a second logical volume to store difference data from the one or more first logical volumes, the second logical volume allocated to a computer; and

a controller,

wherein the controller, on receipt of an I/O request to the second logical volume from the computer,

accesses the difference data in the second logical volume if an access destination of the I/O request is the difference data,

accesses the reference data in the one or more first logical volumes if an access destination of the access request is the reference data, and

measures the number of I/O requests to each of the one or more first logical volumes and the second logical volume;

the management apparatus;

identifies a high load volume according to the number of the I/O requests measured by the controller and

judges whether a cause logical unit is a base logical unit or a difference logical unit, and performs a process of improvement of configuration of loads according to a load state;

a storage pool for providing the logical volumes; and

a memory for storing mapping information between logical addresses of each of the logical volumes and addresses of the storage pool;

wherein the controller accesses each of the logical volumes in the volume group according to the mapping information;

wherein the logical volumes comprising a base volume, one or more reference volumes each of which is a snapshot volume of the base volume, and one or more snapshot volumes of the one or more reference volumes;

wherein the management apparatus identifies a logical volume whose load is higher than a predetermined value as the high load volume; and

a configuration of the volume group being classified into the following;

on a first condition that the base logical unit is the cause logical unit and a single child logical unit is included, the base logical unit is moved to another pool, and a first access path between the base logical unit that has been moved and the single child logical unit is set;

on a second condition that the base logical unit is the cause logical unit, a plurality of child logical units is included, and I/O loads of the plurality of child logical units are biased, the base logical unit is moved to the another pool, and a second access path from one of the plurality of child logical units to the base logical unit that has been moved is set;

on a third condition that the base logical unit is the cause logical unit, the plurality of child logical units is included, and the I/O loads of the plurality of child logical units are not biased, a copy of the base logical unit is created in the another pool, and a third access path is set in such a manner that an I/O load of the base logical unit of a copy source and an I/O load of the base logical unit of a copy destination are about equal to each other;

on a fourth condition that the difference logical unit is the cause logical unit, and the single child logical unit is included, the difference logical unit is moved to the another pool, and a fourth access path from the single child logical unit to the difference logical unit that has been moved and a fifth access path from the moved difference logical unit to a parent logical unit are set;

on a fifth condition that the base logical unit is the cause logical unit, the plurality of child logical units is included, and the I/O loads of the plurality of child logical units are biased, a sixth access path from the child logical unit to the difference logical unit that has been moved and a seventh access path from the moved difference logical unit to the parent logical unit are set; and

on a sixth condition that the difference logical unit is the cause logical unit, the plurality of child logical units is included, and the I/O loads of the plurality of child logical units are not biased, a copy of the difference logical unit is created in the another pool, and an eighth access path is set in such a manner that an I/O load of the difference logical unit of the a copy source and an I/O load of the difference logical unit of the copy destination are about equal to each other.

2. A storage system according to claim 1, wherein the management apparatus copies or moves the high load volume to another storage pool.

3. A storage system according to claim 2, wherein

a response performance of another storage pool is higher than that of the storage pool.

4. A storage system according to claim 2, wherein

if a plurality of computers accesses the high load volume, the management apparatus allocates I/O requests from some computers of the computers to a copied volume of the high load volume.

5. A method of managing a performance of a storage system comprising a volume group that includes a plurality of logical volumes, the logical volumes including one or more first logical volumes to store reference data and a second logical volume to store difference data from the one or more first logical volumes, the second logical volume allocated to a computer;

the method comprising the steps of;

receiving an I/O request to the second logical volume from the computer,

accessing the difference data in the second logical volume if an access destination of the I/O request is the difference data,

accessing the reference data in the one or more first logical volumes if an access destination of the access request is the reference data,

measuring the number of I/O requests to each of the one or more first logical volumes and the second logical volume;

identifying a high load volume according to the number of the I/O requests measured by a controller,

judging whether a cause logical unit is a base logical unit or a difference logical unit, and

performing a process of improvement of configuration of loads according to a load state;

a storage pool for providing the logical volumes; and

a memory for storing mapping information between logical addresses of each of the logical volumes and addresses of the storage pool;

wherein the controller accesses each of the logical volumes in the volume group according to the mapping information;

wherein the logical volumes comprising a base volume, one or more reference volumes each of which is a snapshot volume of the base volume, and one or more snapshot volumes of the one or more reference volumes;

wherein the management apparatus identifies a logical volume whose load is higher than a predetermined value as the high load volume; and

configuring of the volume group being classified into the following classifications;

on a first condition that the base logical unit is the cause logical unit and a single child logical unit is included, the base logical unit is moved to another pool, and a first access path between the base logical unit that has been moved and the single child logical unit is set;

on a second condition that the base logical unit is the cause logical unit, a plurality of child logical units is included, and I/O loads of the plurality of child logical units are biased, the base logical unit is moved to the another pool, and a second access path from one of the plurality of child logical units to the base logical unit that has been moved is set;

on a third condition that the base logical unit is the cause logical unit, the plurality of child logical units is included, and the I/O loads of the plurality of child logical units are not biased, a copy of the base logical unit is created in the another pool, and a third access path is set in such a manner that an I/O load of the base logical unit of a copy source and an I/O load of the base logical unit of a copy destination are about equal to each other;

on a fourth condition that the difference logical unit is the cause logical unit, and the single child logical unit is included, the difference logical unit is moved to the another pool, and a fourth access path from the single child logical unit to the difference logical unit that has been moved and a fifth access path from the moved difference logical unit to a parent logical unit are set;

on a fifth condition that the base logical unit is the cause logical unit, the plurality of child logical units is included, and the I/O loads of the plurality of child logical units are biased, a sixth access path from the child logical unit to the difference logical unit that has been moved and a seventh access path from the moved difference logical unit to the parent logical unit are set; and

on a sixth condition that the difference logical unit is the cause logical unit, the plurality of child logical units is included, and the I/O loads of the plurality of child logical units are not biased, a copy of the difference logical unit is created in the another pool, and an eighth access path is set in such a manner that an I/O load of the difference logical unit of the a copy source and an I/O load of the difference logical unit of the copy destination are about equal to each other.

6. A method for managing a performance of a storage system according to claim 5,

the method further comprising a step of copying or moving the high load volume from the storage pool providing the logical volumes to a different storage pool.