Tiered persistent memory allocation
The present disclosure relates to one or more memory management techniques. In embodiments, one or more regions of storage class memory (SCM) of a storage array is provisioned as expanded global memory. The one or more regions can correspond to SCM persistent cache memory regions. The storage array's global memory and expanded global memory can be used to execute one or more storage-related services connected to servicing (e.g., executing) an input/output (IO) operation.
1. A method comprising:
provisioning one or more regions of storage class memory (SCM) of a storage array as expanded global memory, the one or more regions corresponding to SCM persistent cache memory regions;
configuring at least one translation lookaside buffer (TLB) to map at least a portion of virtual memory addresses to corresponding physical global cache memory addresses;
updating a virtual memory searchable data structure with a status of each corresponding physical global cache memory address;
generating the virtual memory searchable data structure configured to present the virtual memory addresses to one or more host devices;
mapping each virtual memory address to a physical memory address of the global cache memory or the expanded global memory;
configuring the virtual memory searchable data structure with direct access to the physical global cache memory or the physical expanded global memory; and
executing one or more storage-related services to service an input/output (JO) operation using one or more of the expanded global memory and global cache memory.
2. The method of claim 1 , further comprising:
configuring the global cache memory and the expanded global memory as the storage array's global memory, the global cache memory corresponding to the storage array's local memory, and the expanded global memory corresponding to at least one storage device's persistent page cache memory.
3. The method of claim 1 further comprising assigning at least one of: the global cache memory or the expanded global memory to service each IO operation received by the storage array.
4. The method of claim 3 further comprising identifying one or more patterns related to one or more sets of IOs received by the storage array.
5. The method of claim 4 further comprising:
generating one or more IO workload models based on the identified patterns; and
assigning the global cache memory or the expanded global memory to service each IO operation based on the one or more IO workload models.
6. The method of claim 1 , wherein the at least one TLB corresponds to the storage array's CPU complex.
7. The method of claim 1 further comprising:
establishing an SCM searchable data structure configured to map at least a portion of the virtual memory address to corresponding physical expanded global memory addresses; and
updating the virtual memory searchable data structure with a status of each related physical expanded global memory address.
8. The method of claim 7 further comprising managing JO operations using each physical global cache memory address's statuses, and each physical expanded global memory address.
9. An apparatus comprising a memory and at least one processor configured to:
provision one or more regions of storage class memory (SCM) of a storage array as expanded global memory, the one or more regions correspond to SCM persistent cache memory regions;
configure at least one translation lookaside buffer (TLB) to map at least a portion of virtual memory addresses to corresponding physical global cache memory addresses;
update a virtual memory searchable data structure with a status of each corresponding physical global cache memory address;
generate the virtual memory searchable data structure configured to present the virtual memory addresses to one or more host devices;
map each virtual memory address to a physical memory address correspond to the global cache memory or the expanded global memory;
configure the virtual memory searchable data structure with direct access to the physical global cache memory or the physical expanded global memory; and
execute one or more storage-related services to service an input/output (JO) operation using one or more of the expanded global memory and global cache memory.
10. The apparatus of claim 9 , further configured to:
configure the global cache memory and the expanded global memory as the storage array's global memory, the global cache memory corresponding to the storage array's local memory, and the expanded global memory correspond to at least one storage device's persistent page cache memory.
11. The apparatus of claim 9 further configured to assign at least one of: the global cache memory or the expanded global memory to service each JO operation received by the storage array.
12. The apparatus of claim 11 further configured to identify one or more patterns related to one or more sets of IOs received by the storage array.
13. The apparatus of claim 12 further configured to:
generate one or more IO workload models based on the identified patterns; and
assign the global cache memory or the expanded global memory to service each IO operation based on the one or more IO workload models.
14. The apparatus of claim 9 , wherein the at least one TLB corresponds to the storage array's CPU complex.
15. The apparatus of claim 9 further configured to:
establish an SCM searchable data structure configured to map at least a portion of the virtual memory address to correspond physical expanded global memory addresses; and
update the virtual memory searchable data structure with a status of each related physical expanded global memory address.
16. The apparatus of claim 15 further configured to manage JO operations us each physical global cache memory address's statuses, and each physical expanded global memory address.