Scale-out storage system and storage control method

Patent metadata

Field	Value
Publication number	US-11630734-B2
Application number	US-202117474395-A
Country	US
Kind code	B2
Filing date	Sep 14, 2021
Priority date	Feb 19, 2021
Publication date	Apr 18, 2023
Grant date	Apr 18, 2023

Abstract

A scale-out storage system includes a plurality of computer nodes each of which has a memory and a processor, and a storage apparatus. The computer nodes have one or more redundancy groups each of which is a group for metadata protection. Each of the one or more redundancy groups includes two or more of the computer nodes including a primary node being a primary computer node and a secondary node being a secondary computer node, and a failover is performed from the primary node to the secondary node. The memory of the primary node has stored therein metadata related to the redundancy group and to be accessed for control. The metadata is redundantly stored in the memory of the primary node and the memory of the secondary node.

First claim

What is claimed is:

1. A scale-out storage system comprising:
a plurality of computer nodes each of which has a memory and a processor; and
a storage apparatus having a plurality of permanent storage devices and connected to the plurality of computer nodes, wherein
the plurality of computer nodes have one or more redundancy groups each of which is a group for metadata protection,
each of the one or more redundancy groups includes two or more of the computer nodes including a primary node being a primary computer node and a secondary node being a secondary computer node, and a failover is performed from the primary node to the secondary node,
the memory of the primary node has stored therein metadata related to the redundancy group and to be accessed for control, and the metadata is redundantly stored in different ones of the permanent storage devices in the storage apparatus, and
the metadata is redundantly stored in the memory of the primary node and the memory of the secondary node, wherein
(a) with respect to each of the one or more redundancy groups, in response to an update request to update the metadata in the primary node, following processes are performed to accomplish redundancy of the metadata updated in the storage apparatus and redundancy of the metadata updated among the memories, the primary node transmits, to the storage apparatus, a first write request to
store the metadata updated in the memory of the primary node or a difference between the updated metadata and the metadata before being updated in a first storage region based on one or more of the permanent storage devices including a first permanent storage device of the storage apparatus, the primary node transmits, to the secondary node, a second write request to store the metadata updated in the memory of the primary node or the difference between the updated metadata and the metadata before being updated in the memory of the secondary node, and
the secondary node transmits, to the storage apparatus, a third write request to store the metadata updated in the memory of the secondary node or a difference between the updated metadata and the metadata before being updated in a second storage region based on one or more of the permanent storage devices including a second permanent storage device of the storage apparatus, or

(b) with respect to each of the one or more redundancy groups, in response to an update request to update the metadata in the primary node, the primary node transmits, to the storage apparatus, one or more write requests to write the metadata updated in the memory of the primary node or a difference between the updated metadata and the metadata before being updated, whereby the updated metadata is redundantly stored in different ones of the permanent storage devices in the storage apparatus, and
at a time of a failover, the secondary node reads the updated metadata from the storage apparatus into the memory of the secondary node.

2. The scale-out storage system according to claim 1 , wherein
each of the one or more redundancy groups is a group to which two or more control programs in the two or more computer nodes belong,
the metadata is data to be accessed in control performed by the processor of any of the computer nodes executing any of the two or more control programs, and
when a failure has occurred in any of the plurality of computer nodes, in each of one or more of the redundancy groups in which the failed computer node is the primary node, a failover is performed from the control program in the primary node to the control program in the secondary node on a basis of the metadata related to the redundancy group and stored in the memory of the secondary node.

3. The scale-out storage system according to claim 1 , wherein
with respect to each of the one or more redundancy groups in (a),
the secondary node stores data accompanying the second write request in the memory of the secondary node, and, when a completion response for the third write request has been received from the storage apparatus, returns a completion response for the second write request to the primary node, and
when the primary node has received a completion response for the first write request from the storage apparatus, and has received the completion response for the second write request from the secondary node, the primary node issues a completion response for the update request.

4. The scale-out storage system according to claim 1 , wherein
with respect to each of the one or more redundancy groups in (b),
the primary node issues a completion response for the update request when the primary node has received a completion response for each of the one or more write requests from the storage apparatus.

5. The scale-out storage system according to claim 1 , wherein
with respect to each of the one or more redundancy groups,
in response to an update request to update the metadata in the primary node, following processes are performed to accomplish redundancy of the metadata updated in the storage apparatus and redundancy of the metadata updated among the memories,
the primary node transmits, to the storage apparatus, one or more write requests to write the metadata updated in the memory of the primary node or a difference between the updated metadata and the metadata before being updated, whereby the updated metadata is redundantly stored in different ones of the permanent storage devices in the storage apparatus, and
the primary node causes the updated metadata to be reflected in the memory of the secondary node in synchronism with the updating of the metadata in the memory of the primary node.

6. The scale-out storage system according to claim 5 ,
wherein with respect to each of the one or more redundancy groups,
the primary node issues a completion response for the update request when the primary node has received a completion response for each of the one or more write requests from the storage apparatus, and has caused the updated metadata to be reflected in the memory of the secondary node in synchronism with the updating of the metadata in the memory of the primary node.

7. The scale-out storage system according to claim 1 , wherein
with respect to each of the one or more redundancy groups,
in response to an update request to update the metadata in the primary node, following processes are performed to accomplish redundancy of the metadata updated in the storage apparatus,
the primary node transmits, to the storage apparatus, one write request to write the metadata updated in the memory of the primary node or a difference between the updated metadata and the metadata before being updated, and in response to the write request, the storage apparatus produces redundancy of the updated metadata or the difference, whereby the updated meta data is redundantly stored in different ones of the permanent storage devices in the storage apparatus.

8. A scale-out storage system comprising:
a plurality of computer nodes each of which has a memory and a processor; and
a storage apparatus having a plurality of permanent storage devices and connected to the plurality of computer nodes, wherein
the plurality of computer nodes have one or more redundancy groups each of which is a group for metadata protection,
each of the one or more redundancy groups includes two or more of the computer nodes including a primary node being a primary computer node and a secondary node being a secondary computer node, and a failover is performed from the primary node to the secondary node,
the memory of the primary node has stored therein metadata related to the redundancy group and to be accessed for control, and the metadata is redundantly stored in different ones of the permanent storage devi

Assignees

• Hitachi Ltd

Inventors

• Ebara Hiroto
• Yamamoto Akira
• Ohira Yoshinori
• Yamamoto Takahiro
• Agetsuma Masakuni
• Hirai Tatsuya

Classifications

• G06F11/0772

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

• G06F11/1435 Primary

  using file system or storage system metadata · CPC title

• G06F11/2076 Primary

  Synchronous techniques · CPC title

• G06F11/1658

  Data re-synchronization of a redundant component, or initial sync of replacement, additional or spare unit · CPC title

• G06F11/2023

  Failover techniques · CPC title