Methods and systems that manage fetching of commands by a controller from queues of a host
A method and system for managing command fetches by an Non-Volatile Memory express (NVMe) controller from a plurality of queues in a host maintains a predefined ratio of data throughput, based on the command fetches, between the plurality of queues. Each of the plurality of queues is assigned with a particular priority and weight.
1. A method executed by a controller of a storage device for fetching commands from high-priority queues and low-priority queues of a host, the method comprising:
incrementing a first count in response to each instance of transferring an amount of data equivalent to a logical block of a predetermined size between the high-priority queues and the storage device;
incrementing a second count in response to each instance of transferring the amount of data equivalent to the logical block between the low-priority queues and the storage device;
prioritizing the fetching of the commands from the high-priority queues and low-priority queues based on the first count and the second count; and
fetching the commands from the high-priority queues and the low-priority queues in accordance with the prioritization, wherein
the fetching the commands is disabled when data throughput of a plurality of queues categorized in the high-priority queues and the low-priority queues is not equal to a predefined ratio of data throughput based on the command fetches of the plurality of queues categorized in the high-priority queues and the low-priority queues.
2. The method of claim 1 , further comprising determining whether the second count exceeds the first count.
3. The method of claim 2 , further comprising:
in response to determining the second count exceeds the first count, determining whether the high-priority queues have a command pending; and
in response to determining the high-priority queues have a command pending, assigning priority to fetching a command from the high-priority queues.
4. The method of claim 2 , further comprising:
in response to determining the second count does not exceed the first count, determining whether priority has been assigned to fetching a command from the high-priority queues; and
in response to determining priority has been assigned to fetching a command from the high-priority queues, eliminating the priority assigned to fetching the command from the high-priority queues.
5. The method of claim 2 , further comprising:
in response to determining the second count exceeds the first count, determining whether the high-priority queues have a command pending; and
in response to determining the high-priority queues do not have a command pending, setting each of the first count and the second count to a same value.
6. The method of claim 1 , wherein the controller comprises a Non-Volatile Memory express (NVMe) controller that processes the commands from the host.
7. A storage device comprising:
an arbitrator that categorizes a first group of queues of a host as high-priority queues and a second group of queues of the host as low-priority queues; and
a controller that manages fetching of commands from the host by:
incrementing a first count in response to each instance of transferring an amount of data equivalent to a logical block of a predetermined size between the high-priority queues and the storage device,
incrementing a second count in response to each instance of transferring the amount of data equivalent to the logical block between the low-priority queues and the storage device,
prioritizing the fetching of the commands from the high-priority queues and low-priority queues based on the first count and the second count, and
fetching the commands from the high-priority queues and the low-priority queues in accordance with the prioritization, wherein
the fetching is disabled when data throughput of a plurality of queues categorized in the high-priority queues and the low-priority queues is not equal to a predefined ratio of data throughput based on the command fetches of the plurality of queues categorized in the high-priority queues and the low-priority queues.
8. The storage device of claim 7 , wherein the controller further processes commands fetched from the host by determining whether the second count exceeds the first count.
9. The storage device of claim 8 , wherein the controller further processes commands fetched from the host by:
determining, in response to determining the second count exceeds the first count, whether the high-priority queues have a command pending; and
assigning, in response to determining the high-priority queues have a command pending, priority to fetching a command from the high-priority queues.
10. The storage device of claim 8 , wherein the controller further processes commands fetched from the host by:
determining, in response to determining the second count does not exceed the first count, whether priority has been assigned to fetching a command from the high-priority queues; and
eliminating, in response to determining priority has been assigned to fetching a command from the high-priority queues, the priority assigned to fetching the command from the high-priority queues.
11. The storage device of claim 8 , wherein the controller further processes commands fetched from the host by:
determining, in response to determining the second count exceeds the first count, whether the high-priority queues have a command pending; and
setting, in response to determining the high-priority queues do not have a command pending, each of the first count and the second count to a same value.
12. The storage device of claim 7 , wherein the controller comprises a Non-Volatile Memory express (NVMe) controller.
13. A storage device comprising:
an arbitrator that categorizes a first group of queues of a host as high-priority queues and a second group of queues of the host as low-priority queues; and
a Non-Volatile memory express (NVMe) controller that manages fetching of commands from the host by:
incrementing a first count in response to each instance of transferring an amount of data equivalent to a logical block of a predetermined size between the high-priority queues and the storage device,
incrementing a second count in response to each instance of transferring the amount of data equivalent to the logical block between the low-priority queues and the storage device,
prioritizing the fetching of the commands from the high-priority queues and low-priority queues based on the first count and the second count, and
fetching the commands from the high-priority queues and the low-priority queues in accordance with the prioritization, wherein
the fetching is disabled when data throughput of a plurality of queues categorized in the high-priority queues and the low-priority queues is not equal to a predefined ratio of data throughput based on the command fetches of the plurality of queues categorized in the high-priority queues and the low-priority queues.
14. The storage device of claim 13 , wherein the NVMe controller further processes commands fetched from the host by determining whether the second count exceeds the first count.
15. The storage device of claim 14 , wherein the NVMe controller further processes commands fetched from the host by:
determining, in response to determining the second count exceeds the first count, whether the high-priority queues have a command pending; and
assigning, in response to determining the high-priority queues have a command pending, priority to fetching a command from the high-priority queues.
16. The storage device of claim 14 , wherein the NVMe controller further processes commands fetched from the host by:
determining, in response to determining the second count does not exceed the first count, whether priority has been assigned to fetching a command from the high-priority queues; and
eliminating, in response to determining priority has been assigned to fetching a command from the high-priority queues, the priority assigned to fetching the command from the high-priority queues.
17. The storage device of claim 14 , wherein the NVMe controller further processes commands fetched from the host by:
determining, in response to determining the second count exceeds the first count, whether the high-priority queues have a command pending; and
setting, in response to determining the high-priority queues do not have a command pending, each of the first count and the second count to a same value.