Management of discrete namespaces by nonvolatile memory controller
This disclosure provides techniques for managing memory which match per-data metrics to those of other data or to memory destination. In one embodiment, wear data is tracked for at least one tier of nonvolatile memory (e.g., flash memory) and a measure of data persistence (e.g., age, write frequency, etc.) is generated or tracked for each data item. Memory wear management based on these individually-generated or tracked metrics is enhanced by storing or migrating data in a manner where persistent data is stored in relatively worn memory locations (e.g., relatively more-worn flash memory) while temporary data is stored in memory that is less worn or is less susceptible to wear. Other data placement or migration techniques are also disclosed.
1 . A storage device comprising:
an interface;
one or more dies of flash memory, a first one of the dies having planes; and
logic, embodied as at least one of circuitry or instructions on a physical storage medium, that when executed, control circuitry of the storage device to:
map storage subdivisions to respective sets of erase units, the erase units of one of the sets, corresponding to one of the subdivisions, including erase units in respective ones of the planes;
transmit, to a host, information representing a size and logical address information for the one of the storage subdivisions;
receive, via the interface, one or more commands from the host which are logically directed to the one of the storage subdivisions, and responsively control physical erasure of erase units, of the erase units of the one of the sets corresponding to the one of the subdivisions, in at least two different ones of the respective ones of the planes;
store metadata data for the one of the of the storage subdivisions; and
detect that the metadata meets a criterion, and responsively:
copy data from an erase unit of the one of the storage subdivisions to a different erase unit; and
map the one of the subdivisions such that, after the copying of the data to the different erase unit, the different erase unit is mapped exclusively to the one of the subdivisions.
2 . The storage device of claim 1 , wherein the logic is further structured to cause the storage device to, with circuitry of the storage device:
receive a write command addressed to a given one of the storage subdivisions;
identify that the set of the erase units which is mapped to the given one of the storage subdivisions has insufficient free space available; and
responsively map the given one of the storage subdivisions to an additional erase unit, and write data accompanying the write command into the additional erase unit, the additional erase unit being exclusive to erase units mapped to each of the storage subdivisions other than the given one of the storage subdivisions.
3 . The storage device of claim 1 , wherein:
the one or more commands comprise a command to overwrite a logical address with data, the logical address being associated with a given one of the storage subdivisions; and
the logic is further structured to cause the storage device to, with circuitry of the storage device:
track metadata representing release status of physical memory space comprising at least one erase unit in the set which is respective to the given one of the storage subdivisions;
responsive to receipt of the command to overwrite the logical address, deallocate a physical storage location to which the logical address was previously mapped, the physical storage location being in the physical memory space, such that the storage location is marked as released; and
detect a condition where each physical location, in the physical memory space, which contains data, has been marked as released, wherein the physical erasure of erase units in at least two different ones of the respective ones of the planes is responsive to the detection of the condition.
4 . The storage device of claim 1 , wherein:
the one or more commands comprise a deallocate command;
the logic is further structured to cause the storage device to, with circuitry of the storage device:
track metadata representing release status of physical memory space comprising at least one erase unit in the set which is respective to the given one of the storage subdivisions;
responsive to receipt of the deallocate command, deallocate a physical storage location to which the logical address was previously mapped, the physical storage location being in the physical memory space, such that the storage location is marked as released; and
detect a condition where each physical location, in the physical memory space, which contains data, has been marked as released, wherein the physical erasure of erase units in at least two different ones of the respective ones of the planes is responsive to the detection of the condition.
5 . The storage device of claim 1 , wherein the one or more commands from the host are a single command and, further, wherein the single command specifies one or more addresses corresponding to the size of the storage subdivision, such that the logic is operable to cause circuitry of the storage device to control the physical erasure of erase units in the at least two different ones of the respective ones of the planes in response to the single command.
6 . The storage device of claim 1 , wherein the logic is further structured to cause the storage device to, with circuitry of the storage device:
store additional metadata respective to the one of the storage subdivisions;
receive a query from the host specifying the one of the storage subdivisions; and
transmit, to the host, information dependent on the stored additional metadata which is respective to the one of the storage subdivisions.
7 . The storage device of claim 6 , wherein the information which is transmitted to the host indicates an extent to which the one of the storage subdivisions is full.
8 . The storage device of claim 6 , wherein the information which is transmitted to the host represents a time since data was written to at least one erase unit in the set which is respective to the one of the storage subdivisions.
9 . The storage device of claim 1 , wherein the size of the one of the storage subdivisions corresponds to a minimum amount of storage capacity to be erased by the storage device.
10 . The storage device of claim 1 , wherein:
the metadata comprises release status information for individual logical block addresses (LBAs) associated with the one of the storage subdivisions; and
the logic is further structured to cause the storage device to, with circuitry of the storage device:
detect when the tracked release status information meets a criterion; and
in response to the tracked release status information meeting the criterion, perform at least one maintenance operation on the one of the storage subdivisions to the exclusion of at least one other of the storage subdivisions.
11 . The storage device of claim 1 , wherein the logic is further structured to cause the storage device to, with circuitry of the storage device:
identify that one or more erase units in the set of erase units which is mapped to a second one of the storage subdivisions has stored data but where all physical storage locations containing such stored data are released; and
responsively control physical erasure of each of the identified one or more erase units.
12 . The storage device of claim 1 , wherein the logic is further structured to cause the storage device to, with circuitry of the storage device:
track additional metadata for erase units of the flash memory die, wherein the additional metadata provides information representing respective erase unit wear;
select the different erase unit from among alternative choices of erase units in dependence on respective erase unit wear associated with the alternative choices of erase units.
13 . The storage device of claim 1 , wherein the logic is further structured to cause the storage device to, with circuitry of the storage device:
identify a data access characteristic which is associated with the data being responsively copied to the different erase unit; and
select the different erase unit from among alternative choices of erase units also in dependence on the data access characteristic.
14 . The storage device of claim 13 , wherein the data access characteristic is dependent on access frequency of the data being responsively copied.
15 . The storage device of claim 1 , wherein the logic is further structured to cause the storage device to, with circuitry of the storage device:
store a number indicating a quantity of the storage subdivisions; and
receive a query from the host and responsively transmit, to the host, information dependent on the stored number.
16 . The storage device of claim 1 , wherein the logic is further structured to cause the storage device to, with circuitry of the storage device:
detect an error condition associated with an erase unit in one of the of the sets of erase units which is mapped to a given one of the subdivisions; and
responsively remap the given one of the subdivisions to a new erase unit, the new erase unit being exclusively associated with the given one of the subdivisions.
17 . The storage device of claim 1 , wherein:
the one or more commands comprise a write command to write a particular logical address with data; and
the logic is further structured to cause the storage device to, with circuitry of the storage device, and responsive to the write command, control the write of data to the erase units in at least two different ones of the respective ones of the planes concurrently using a common physical page address.
18 . The storage device of claim 1 , wherein:
the one or more dies of flash memory include a second one of the dies;
the second one of the dies also has planes;
one or more commands comprise a write command to write a particular logical address with data; and
the logic is further structured to cause the storage device to, with circuitry of the storage device, and responsive to the write command, control the write of data to the erase units of one of the storage subdivisions including to an erase unit in a single plane in each of the first one of the dies and the second one of the dies, using a common physical page offset for the respective erase units.
19 . The storage device of claim 1 , wherein the logic is further structured to cause the storage device to, with circuitry of the storage device:
store additional metadata respective to the one of the storage subdivisions, wherein the metadata indicates an extent to which the one of the storage subdivisions is full; and
receive a write command addressed to the one of the subdivisions and responsively update the metadata.
20 . A method of operating a storage device, the storage device having an interface, one or more dies of flash memory, a first one of the dies having planes, and logic, embodied as at least one of circuitry, or instructions on a physical storage medium that, when executed, controls the actions of circuitry of the storage device, the method comprising, using the logic:
mapping storage subdivisions to respective sets of erase units, in a manner such that the erase units of one of the sets, corresponding to one of the subdivisions, includes erase units in respective ones of the planes;
transmitting, to a host, information representing a size and logical address information for the one of the storage subdivisions;
receiving, via the interface, one or more commands from the host which are logically directed to the one of the storage subdivisions, and responsively controlling physical erasure of erase units, of the erase units of the one of the sets corresponding to the one of the subdivisions, in at least two different ones of the respective ones of the planes;
storing metadata data for the one of the of the storage subdivisions; and
detecting that the metadata meets a criterion, and responsively:
copying data from an erase unit of the one of the storage subdivisions to a different erase unit; and
mapping the one of the subdivisions such that, after the copying of the data to the different erase unit, the different erase unit is mapped exclusively to the one of the subdivisions.