Storage system and cluster configuration control method

1. A storage system including N storage nodes that are members of a storage cluster (N being an integer equal to or larger than 3), wherein

a first storage node that is any one of the N storage nodes

determines whether importance of a second storage node is equal to or larger than a predetermined importance and reliability of the second storage node is equal to or larger than a predetermined reliability, the second storage node being a storage node set as an object among storage nodes other than the first storage node,

when the determination result is true, performs reintegration which is processing for causing the second storage node to leave the storage cluster and causing the second storage node to become a member of the storage cluster again,

the importance of the second storage node depends on highness of availability when assuming that the second storage node has left the storage cluster,

the reliability of the second storage node depends on tendency of operation of the second storage node

each of the N storage nodes comprises a processor unit,

the N storage nodes comprise Q cluster control programs (Q being an integer equal to or larger than 2 and equal to or smaller than N),

the Q cluster control programs are respectively arranged in Q storage nodes,

the Q cluster control programs includes

a primary cluster control program; and

one or more secondary cluster control programs, the one or more secondary cluster control programs each being a cluster control program other than the primary control program, and arranged in one or more storage nodes other than a storage node where the primary cluster control program is arranged,

the primary cluster control program, when executed by the processor unit in a storage node where the primary cluster control program is arranged, manages a cluster serving as the storage system,

when the primary cluster control program is stopped, any one of the one or more secondary cluster control programs becomes primary instead of the primary cluster control program,

the importance of the second storage node depends on a remaining node number that is Q when the second storage node is not included in Q storage nodes where the Q cluster control programs are arranged and is a value obtained by subtracting the number of the second storage nodes from Q when the second storage node is included in the Q storage nodes,

the importance of the second storage node depends on whether the remaining node number is equal to or less than a threshold of the remaining node number, and

the threshold of the remaining node number is a value obtained by adding the number of storage nodes having a possibility to simultaneously be targeted for the reintegration to the majority of Q.

2. The storage system according to claim 1 , wherein

the N storage nodes further comprise

P program clusters (P being a natural number) and

each of the P program clusters includes

an active input/output (I/O) control program; and

one or more standby I/O control programs, the one or more standby I/O control programs each being an I/O control program other than the active I/O control program, and arranged in one or more storage nodes other than a storage node where the active I/O control program is arranged,

the active I/O control program, when executed by the processor unit in a storage node where the I/O control program is arranged, writes a dataset for one or more storage nodes,

when the active I/O control program is stopped, any one of the one or more standby I/O control programs becomes active instead of the active I/O control program,

the importance of the second storage node depends on

redundancy of at least one dataset stored in the second storage node, and

redundancy of I/O control program in the program cluster with respect to at least one I/O control program arranged in the second storage node.

3. The storage system according to claim 1 , wherein

the importance of the second storage node depends on redundancy of the dataset with respect to at least one dataset stored in the second storage node.

4. The storage system according to claim 1 , wherein the tendency of operation of the second storage node includes tendency of failure occurrence in the second storage node.

5. The storage system according to claim 4 , wherein the tendency of failure occurrence in the second storage node includes the tendency of failure occurrence in a recent long period of the second storage node and the tendency of failure occurrence in a recent short period of the second storage node.

6. A cluster configuration control method including:

determining whether importance of an object storage node is equal to or larger than a predetermined importance and reliability of the object storage node is equal to or larger than a predetermined reliability, the object storage node being a storage node set as an object among N storage nodes that are members of a storage cluster (N being an integer equal to or larger than 3); and

when the determination result is true, performing reintegration which is processing for causing the object storage node to leave the storage cluster and causing the object storage node to become a member of the storage cluster again, wherein

the importance of the object storage node depends on highness of availability when assuming that the object storage node has left the storage cluster,

the reliability of the object storage node depends on tendency of operation of the object storage node

each of the N storage nodes comprises a processor unit,

the N storage nodes comprise Q cluster control programs (Q being an integer equal to or larger than 2 and equal to or smaller than N),

the Q cluster control programs are respectively arranged in Q storage nodes,

the Q cluster control programs includes

a primary cluster control program; and

one or more secondary cluster control programs, the one or more secondary cluster control programs each being a cluster control program other than the primary control program, and arranged in one or more storage nodes other than a storage node where the primary cluster control program is arranged,

the primary cluster control program, when executed by the processor unit in a storage node where the primary cluster control program is arranged, manages a cluster serving as the storage system,

when the primary cluster control program is stopped, any one of the one or more secondary cluster control programs becomes primary instead of the primary cluster control program,

the importance of the second storage node depends on a remaining node number that is Q when the second storage node is not included in Q storage nodes where the Q cluster control programs are arranged and is a value obtained by subtracting the number of the second storage nodes from Q when the second storage node is included in the Q storage nodes,

the importance of the second storage node depends on whether the remaining node number is equal to or less than a threshold of the remaining node number, and

the threshold of the remaining node number is a value obtained by adding the number of storage nodes having a possibility to simultaneously be targeted for the reintegration to the majority of Q.

7. A non-transitory computer readable medium storing a program for controlling a computer that improves stability and reduces deterioration, when executed by a processor, causes the processor to:

determine whether importance of an object storage node is equal to or larger than a predetermined importance and reliability of the object storage node is equal to or larger than a predetermined reliability, the object storage node being a storage node set as an object among N storage nodes that are members of a storage cluster (N being an integer equal to or larger than 3); and

when the determination result is true, performing reintegration which is processing for causing the object storage node to leave the storage cluster and causing the object storage node to become a member of the storage cluster again, wherein

the importance of the object storage node depends on highness of availability when assuming that the object storage node has left the storage cluster,

the reliability of the object storage node depends on the tendency of operation of the object storage node

each of the N storage nodes comprises a processor unit,

the N storage nodes comprise Q cluster control programs (Q being an integer equal to or larger than 2 and equal to or smaller than N),

the Q cluster control programs are respectively arranged in Q storage nodes,

the Q cluster control programs includes

a primary cluster control program; and

one or more secondary cluster control programs, the one or more secondary cluster control programs each being a cluster control program other than the primary control program, and arranged in one or more storage nodes other than a storage node where the primary cluster control program is arranged,

the primary cluster control program, when executed by the processor unit in a storage node where the primary cluster control program is arranged, manages a cluster serving as the storage system,

when the primary cluster control program is stopped, any one of the one or more secondary cluster control programs becomes primary instead of the primary cluster control program,

the importance of the object storage node depends on a remaining node number that is Q when the object storage node is not included in Q storage nodes where the Q cluster control programs are arranged and is a value obtained by subtracting the number of the object storage nodes from Q when the object storage node is included in the Q storage nodes,

the importance of the object storage node depends on whether the remaining node number is equal to or less than a threshold of the remaining node number, and

the threshold of the remaining node number is a value obtained by adding the number of storage nodes having a possibility to simultaneously be targeted for the reintegration to the majority of Q.