Sharing spare capacity of disks with multiple sizes to parallelize RAID rebuild

A method implemented by a storage array comprising at least one compute node comprising at least one processor and non-transitory computer-readable memory, and a plurality of non-volatile drives of different sizes in a fixed arithmetic relationship relative to a baseline drive size, the method comprising:

• organizing the drives into a plurality of clusters based on storage capacity such that all the drives within each cluster have equal storage capacity;
• creating M*G same-size partitions on the drives, where M is a positive integer defined by the fixed arithmetic relationship relative to the baseline drive size, and G is a redundant array of independent disks (RAID) group size such that storage capacity of all drives defines matrices of G+1 rows and M*G columns, and each matrix comprises submatrices of G+1 rows and G columns;
• allocating single diagonal spare partitions over groups of G columns of all the clusters;
• vertically distributing members of RAID groups such that members of a given RAID group reside in a single partition index of a single cluster of the plurality of clusters;
• and
• responsive to a drive failure in a first one of the plurality of clusters, selecting spare partitions from multiple other ones of the plurality of clusters and rebuilding RAID group members of the failed drive on the selected spare partitions.

A non-transitory computer-readable storage medium with instructions that when executed by a storage array comprising at least one compute node comprising at least one processor and non-transitory computer-readable memory, and a plurality of non-volatile drives of different sizes in a fixed arithmetic relationship relative to a baseline drive size, cause the storage array to create distributed spare capacity, the method comprising:

• organizing the drives into a plurality of clusters based on storage capacity such that all the drives within each cluster have equal storage capacity;
• creating M*G same-size partitions on the drives, where M is a positive integer defined by the fixed arithmetic relationship relative to the baseline drive size, and G is a redundant array of independent disks (RAID) group size such that storage capacity of all drives defines matrices of G+1 rows and M*G columns, and each matrix comprises submatrices of G+1 rows and G columns;
• allocating single diagonal spare partitions over groups of G columns of all the clusters;
• vertically distributing members of RAID groups such that members of a given RAID group reside in a single partition index of a single cluster of the plurality of clusters;
• and
• responsive to a drive failure in a first one of the plurality of clusters, selecting spare partitions from multiple other ones of the plurality of clusters and rebuilding RAID group members of the failed drive on the selected spare partitions.