Storage systems and methods for writing data to a persistent storage device
A system and method which allows data to be received into a placement intelligence. After the data is analyzed. the data is written to a persistent storage device. Subsequently. the data may be written. Periodically, self-optimization may occur to improve read speeds or other metrics.
1. A method of writing data to a persistent storage device comprising an array of drives, the method comprising:
receiving from a requester a request to write data;
determining metadata; and
selectively writing the data to a physical location on the persistent storage device based on the metadata, wherein the physical location corresponds to a next free block in a sequence of blocks on a drive in the array of drives,
wherein the metadata comprises an affinity of the data,
wherein the affinity of the data is determined by a weighted combination of a plurality of factors, the factors including a first factor based on characteristics of the requester, a second factor based on a temporal analysis of the data, and a third factor based on a positional analysis of the data, and
wherein weights of the weighted combination given to each factor vary over time.
2. The method of claim 1 , wherein selectively writing the data to the physical location on the persistent storage device based on the metadata comprises writing data with the same affinity to a contiguous region in the persistent storage device.
3. The method of claim 2 , wherein the characteristics of the requester include one or more of a process identifier, and a central processing unit (CPU) identifier.
4. The method of claim 2 , wherein the array of drives comprises an array of hard disk drives, and wherein the contiguous region is on the array of hard disk drives.
5. The method of claim 1 , wherein the metadata comprises a priority value of the data.
6. The method of claim 5 , wherein the array of drives comprises an array of hard disk drives, wherein selectively writing the data to the physical location on the persistent storage device based on the metadata comprises writing data with a high priority value to a fast zone in the persistent storage device, and wherein the fast zone is on the array of hard disk drives.
7. The method of claim 5 , wherein the array of drives comprises an array of solid-state drives, wherein selectively writing the data to the physical location on the persistent storage device based on the metadata comprises writing data with a high priority value to a high tier in the persistent storage device, and wherein the high tier comprises the array of solid-state drives.
8. The method of claim 1 , wherein the array of drives comprises multiple solid-state drives, and wherein selectively writing the data to the physical location on the persistent storage device based on the metadata comprises distributing data with affinity across multiple dies on the multiple solid-state drives.
9. The method of claim 1 , wherein the request to write data is a request to write data to a virtual location and comprises a logical block addressing value, the method further comprising recording a mapping from the virtual location to the physical location, wherein the virtual location and the physical location are not correlated.
10. A storage system comprising:
a persistent storage device comprising an array of drives; and
a storage controller configured to perform steps comprising:
receiving from a requester a request to write data;
determining metadata; and
selectively writing the data to a physical location on the persistent storage device based on the metadata, wherein the physical location corresponds to a next free block in a sequence of blocks on a drive in the array of drives,
wherein the metadata comprises an affinity of the data,
wherein the affinity of the data is determined by a weighted combination of a plurality of factors, the factors including a first factor based on characteristics of the requester, a second factor based on a temporal analysis of the data, and a third factor based on a positional analysis of the data, and
wherein weights of the weighted combination given to each factor vary over time.
11. The storage system of claim 10 , wherein selectively writing the data to the physical location on the persistent storage device based on the metadata comprises writing data with the same affinity to a contiguous region in the persistent storage device.
12. The storage system of claim 11 , wherein the characteristics of the requester include one or more of a process identifier, and a central processing unit (CPU) identifier.
13. The storage system of claim 11 , wherein the array of drives comprises an array of hard disk drives, and wherein the contiguous region is on the array of hard disk drives.
14. The storage system of claim 10 , wherein the metadata comprises a priority value of the data.
15. The storage system of claim 14 , wherein the array of drives comprises an array of hard disk drives, wherein selectively writing the data to the physical location on the persistent storage device based on the metadata comprises writing data with a high priority value to a fast zone in the persistent storage device, and wherein the fast zone is on the array of hard disk drives.
16. The storage system of claim 14 , wherein the array of drives comprises an array of solid-state drives, wherein selectively writing the data to the physical location on the persistent storage device based on the metadata comprises writing data with a high priority value to a high tier in the persistent storage device, and wherein the high tier comprises the array of solid-state drives.
17. The storage system of claim 10 , wherein the array of drives comprises multiple solid-state drives, and wherein selectively writing the data to the physical location on the persistent storage device based on the metadata comprises distributing data with affinity across multiple dies on the multiple solid-state drives.
18. The storage system of claim 10 , wherein the request to write data is a request to write data to a virtual location and comprises a logical block addressing value, the storage controller further configured to record a mapping from the virtual location to the physical location, wherein the virtual location and the physical location are not correlated.