Garbage collection and defragmentation for solid state drives (SSD) and shingled magnetic recording (SMR) drives
Example apparatus and methods provide improved reclamation, garbage collection (GC) and defragmentation (defrag) for data storage devices including solid state drives (SSD) or shingled magnetic recording (SMR) drives. An erasure code (EC) layer that facilitates logically or physically erasing data from the SSD or SMR as a comprehensive GC or defrag is added to the SSD or SMR. Erased data may be selectively recreated from the EC layer as needed. Pre-planned EC write zones may be established to further optimize GC and defrag. Recreated data may be written to selected locations to further optimize SSD and SMR performance. Erasure code data may be distributed to co-operating devices to further improve GC or defrag. Example apparatus and methods may also facilitate writing data to an SMR drive using tape or VTL applications or processes and providing a pseudo virtual tape library on the SMR drive.
1. A non-transitory computer-readable storage medium storing computer executable instructions that when executed by a computer control the computer to perform a method, the method comprising:
accessing a first tape and a second tape;
upon determining that a data reduction level achieved by compressing first data from the first tape before writing the first data from the first tape to a shingled magnetic recording (SMR) device and compressing second data from the second tape before writing the second data from the second tape to the SMR device exceeds a data reduction level achieved by compressing the first data on the SMR device after writing the first data to the SMR device and compressing the second data on the SMR device after writing the second data to the SMR device:
compressing the first data before writing the first data from the first tape to the SMR device and compressing the second data from the second tape before writing the second data from the second tape to the SMR device;
writing the first data from the first tape to a first contiguous set of overlapping tracks on the SMR device;
writing the second data from the second tape to a second different contiguous set of overlapping tracks on the SMR device;
upon determining that a data reduction level achieved by compressing the first data after writing the first data from the first tape to the SMR device and compressing the second data from the second tape after writing the second data from the second tape to the SMR device exceeds a data reduction level achieved by compressing the first data on the SMR device before writing the first data to the SMR device and compressing the second data on the SMR device before writing the second data to the SMR device:
compressing the first data after writing the first data from the first tape to the SMR device and compressing the second data from the second tape after writing the second data from the second tape to the SMR device; and
providing the SMR device as a collective virtual tape library that virtualizes the first tape and the second tape.
2. The non-transitory computer-readable storage medium of claim 1 , the method comprising:
compressing the first data before writing the first data from the first tape to the SMR device and compressing the second data from the second tape before writing the second data from the second tape to the SMR device; and
compressing the first data on the SMR device after writing the first data to the SMR device and compressing the second data on the SMR device after writing the second data to the SMR device.
3. The non-transitory computer-readable storage medium of claim 1 , where compressing the first data or compressing the second data includes compressing the first data or the second data using a single instancing approach, a data deduplication approach, or a Lempel-Ziv (LZ) compression approach.
4. The non-transitory computer-readable storage medium of claim 1 , where writing first data from the first tape to the first contiguous set of overlapping tracks includes writing the first data sequentially.
5. The non-transitory computer-readable storage medium of claim 1 , where writing second data from the second tape to the second different contiguous set of overlapping tracks includes writing the second data sequentially.
6. The non-transitory computer-readable storage medium of claim 1 , where the collective virtual tape library addresses the first data or the second data collectively.
7. A non-transitory computer-readable storage medium storing computer executable instructions that when executed by a computer control the computer to perform a method, the method comprising:
accessing data to be stored by a process that writes to a tape drive or to a virtual tape library, where compressing the data before writing the data to a shingled magnetic recording (SMR) device has a first data reduction level, a first computing resource cost, or a first energy requirement, and where compressing the data on the SMR device after writing the data to the SMR device has a second data reduction level, a second computing resource cost, or a second energy requirement;
upon determining that the first data reduction level is greater than the second data reduction level:
compressing the data before writing the data to the SMR device;
writing the data to a first contiguous set of overlapping tracks on the SMR device;
upon determining that the second data reduction level is greater than the first data reduction level:
compressing the data on the SMR device after writing the data to the SMR device; and
providing the SMR device as a virtual tape library that virtualizes the data as a virtual tape.
8. The non-transitory computer-readable storage medium of claim 7 , where compressing the data before writing the data includes compressing the data using a single instancing approach, a data deduplication approach, or a Lempel-Ziv (LZ) compression approach.
9. The non-transitory computer-readable storage medium of claim 7 , where compressing the data on the SMR device after writing the data includes compressing the data using a single instancing approach, a data deduplication approach, or a Lempel-Ziv (LZ) compression approach.
10. The non-transitory computer-readable storage medium of claim 7 , the method comprising:
computing a first cost based on the first computing resource cost and the first energy requirement;
computing a second cost based on the second computing resource cost and the second energy requirement;
upon determining that the first cost is greater than the second cost:
compressing the data on the SMR device after writing the data to the SMR device;
upon determining that the second cost is greater than the first cost:
compressing the data before writing the data to the SMR device.
11. An apparatus, comprising:
a shingled magnetic recording (SMR) drive;
a memory;
a processor;
an input/output (I/O) interface;
a set of circuits comprising a data access circuit, a data reduction level circuit, a compression circuit, a storage circuit, and a virtual tape library (VTL) circuit; and
an interface that connects the memory, the processor, the I/O interface, the SMR drive, and the set of circuits;
where the data access circuit is configured to access data to be stored by a process that writes to a tape drive or to a virtual tape library;
where the data reduction level circuit is configured to:
compute a first data reduction value associated with compressing the data before writing the data to the SMR drive;
compute a second data reduction value associated with compressing the data after writing the data to the SMR drive;
where the compression circuit is configured to:
upon determining that the first data reduction value is greater than the second data reduction value:
compress the data before writing the data to the SMR drive;
upon determining that the second data reduction value is greater than the first data reduction value:
compress the data after writing the data to the SMR drive;
where the storage circuit is configured to write the data to a first contiguous set of overlapping tracks on the SMR drive; and
where the VTL circuit is configured to provide the SMR drive as a collective VTL that virtualizes the data as a virtual tape.
12. The apparatus of claim 11 , where the data reduction level circuit is configured to compute the first data reduction value based, at least in part on a first data reduction level, a first computing resource cost, or a first energy requirement associated with compressing the data before writing the data to the SMR drive, and
where the data reduction level circuit is configured to compute the second data reduction value based, at least in part, on a second data reduction level, a second computing resource cost, or a second energy requirement associated with compressing the data after writing the data to the SMR drive.
13. The apparatus of claim 11 , where the storage circuit is configured to write the data sequentially to a first contiguous set of overlapping tracks on the SMR drive.