Storage system

Patent metadata

Field	Value
Publication number	US-12045133-B2
Application number	US-202217940312-A
Country	US
Kind code	B2
Filing date	Sep 8, 2022
Priority date	Jan 21, 2022
Publication date	Jul 23, 2024
Grant date	Jul 23, 2024

Abstract

In each node constituting a storage system, there is at least one of a storage area (user area) in which a user data set is stored and a storage area (parity area). For the node having the user area, there is user part difference information including information indicating whether or not to be in presence of difference for each user area of the node. For the node having the parity area, there is parity part difference information including the information indicating whether or not to be in the presence of difference for each parity area of the node. Out of the parity part difference information, the information corresponding to the parity area is the information indicating the presence of difference when there is the information indicating the presence of difference in the storage area of any of the data sets used for generating the parity stored in the parity area.

First claim

What is claimed is:

1. A storage system, comprising:
a plurality of nodes, each having a processor, a memory, and a permanent storage device,
wherein, for each of the plurality of nodes,

as a storage area based on the permanent storage device of the node, there is at least one of a user area being a storage area storing a user data set as a data set and a parity area being a storage area storing a parity as a data set,
wherein the parity stored in the parity area of the node is a data set generated by using the data set from each of the two or more nodes other than the node,
wherein, for the node having the user area, there is user part difference information including information indicating whether or not to be in presence of difference for each user area of the node,
wherein, for each user area, the presence of difference denotes that update of the user data set of the user area of the node occurs while there is a failed node being a node in which the failure occurs,
wherein, for the node having the parity area, there is parity part difference information including the information indicating whether or not to be in the presence of difference for each parity area of the node,
wherein, with respect to the parity part difference information, for each parity area, the information corresponding to the parity area is information indicating the presence of difference when there is the information indicating the presence of difference for the storage area of any of the data sets used for generating the parity stored in the parity area, and
wherein, for both the user part difference information and the parity part difference information, the node retaining the difference information in the memory updates the information indicating the presence of difference out of the difference information to information indicating absence of difference when the data sets of the storage areas in all the failed nodes are restored out of a storage area group to which the storage areas corresponding to the information belong.

2. The storage system according to claim 1 ,
wherein, for each of the plurality of nodes, the node has a plurality of physical chunks, each being configured with the two or more storage areas,
wherein each of the user part difference information and the parity part difference information exists for each physical chunk, and
wherein the update to the information indicating the absence of difference is performed in units of a physical chunk.

3. The storage system according to claim 1 ,
wherein, when the failed node is recovered from the failure,
wherein, for each of the plurality of storage areas of a recovered node being the node recovered from the failure, the information corresponding to the storage area is set as the information indicating the presence of difference when at least one storage area among two or more storage areas other than the storage area out of the storage area group to which the storage area belongs is in the presence of difference, and
wherein, among the plurality of storage areas of the recovered node, for each storage area with the presence of difference, the recovered node restores the data set.

4. The storage system according to claim 1 , wherein each of the plurality of nodes writes the difference information of the node from the memory of the node to the permanent storage device of the node at a specific trigger.

5. The storage system according to claim 4 , wherein, for each of the plurality of nodes, the specific trigger is a time of the update of the difference information of the node or a time of a planned stop of the node.

6. The storage system according to claim 1 , wherein the memory of the node having the parity area retains the parity part difference information including the information for each parity area of the node.

7. The storage system according to claim 1 , wherein the memory of the node having the user area retains the user part difference information including the information for each user area of the node.

8. The storage system according to claim 7 , wherein each of the plurality of nodes:
provides a volume; and
stores all the user data sets obtained from the user data accompanying a write request designating the volume in the permanent storage device of the node.

9. A storage system comprising:
a plurality of nodes, each having a processor, a memory, and a permanent storage device,
wherein, for each of the plurality of nodes,

as a storage area based on the permanent storage device of the node, there is at least one of a user area being a storage area storing a user data set as a data set and a parity area being a storage area storing a parity as a data set,
wherein the parity stored in the parity area of the node is a data set generated by using the data set from each of the two or more nodes other than the node,
wherein, for the node having the user area, there is user part difference information including information indicating whether or not to be in presence of difference for each user area of the node,
wherein, for each user area, the presence of difference denotes that update of the user data set of the user area of the node occurs while there is a failed node being a node in which the failure occurs,
wherein, for the node having the parity area, there is parity part difference information including the information indicating whether or not to be in the presence of difference for each parity area of the node,
wherein, with respect to the parity part difference information, for each parity area, the information corresponding to the parity area is information indicating the presence of difference when there is the information indicating the presence of difference for the storage area of any of the data sets used for generating the parity stored in the parity area,
wherein a data protection policy of a stripe is jDkP (j is an integer of 2 or more and k is an integer of 1 or more),
wherein the stripe is configured with the j user areas and the k parity areas,
wherein any one among the (j+k) or more nodes configured with the j nodes having the j user areas and the k nodes having the k parity areas is the failed node being the node in which the failure occurs,
wherein each of surviving nodes being nodes other than the failed node among the (j+k) or more nodes stores the data set in the storage area of a write destination,
wherein the information corresponding to the storage area of the write destination becomes the information indicating the presence of difference, and
wherein, for each parity area of the recovered node being the node that recovered from the failure, the recovered node,
updates the information corresponding to the parity area to the information indicating the presence of difference when the information corresponding to at least one user area in the stripe including the parity area is the information indicating the presence of difference, and
restores the parity in the parity area based on the user data set in the data other than the recovered node when the information corresponding to the parity area indicates the presence of difference.

10. A storage system comprising:
a plurality of nodes, each having a processor, a memory, and a permanent storage device,
wherein, for each of the plurality of nodes,

as a storage area based on the permanent storage device of the node, there is at least one of a user area being a storage area storing a user data set as a data set and a parity area being a storage area storing a parity as a data set,
wherein the parity stored in the parity area of the node is a data set generated by using the data set from each of the two or more nodes other than the node,
whe

Assignees

• Hitachi Ltd

Inventors

• Yamamoto Takahiro
• Saito Hideo
• Ito Shintaro
• Tajima Sachie
• Ebara Hiroto
• Ono Taisuke

Classifications

• G06F11/0772

  Means for error signaling, e.g. using interrupts, exception flags, dedicated error registers · CPC title

• G06F11/1469

  Backup restoration techniques · CPC title

• G06F11/1076 Primary

  Parity data used in redundant arrays of independent storages, e.g. in RAID systems · CPC title