Variable QoS management of multiple data streams
Storage devices and systems are capable of dynamically managing QoS requirements associated with host applications via a management interface. The management interface may the enable storage devices to: (i) decide which data needs to be transferred back to the hosts, (ii) choose to skip portions of the data transferred back to the hosts to improve throughput and maintain low cost, and (iii) operate contention resolutions with host applications. Furthermore, storage devices and systems may achieve a virtual throughput that may be greater than its actual physical throughput. The management interface may also be operated at an application level, which advantageously allows the devices and systems the capabilities of managing contention resolutions of host applications, and managing (changing, observing, fetching, etc.) one or more QoS requirements for each host application.
1. A storage device, comprising:
one or more communication channels suitable for connection with a plurality of host devices;
a plurality of memory devices within a memory array;
a host interface, configured to receive and transmit multiple data streams from and to the plurality of host devices, each data stream having a corresponding service level agreement (SLA) measured by a corresponding quality of service (QoS);
a plurality of controllers configured to receive commands sent from a sending host of the plurality of host devices;
a dynamic management interface, executed by at least one of the plurality of controllers, configured to manage each received data stream with the corresponding QoS and associate each received data stream to the sending host;
a dynamic QoS control module coupled with the dynamic management interface and the host interface;
a contention module configured to prioritize one or more applications having one or more contentions with respect to a plurality of shared resources, wherein the one or more applications are prioritized based on a corresponding QoS requirement;
the contention module coupled with a cache configured to cache data from the data stream, wherein the storage device is configured to manage the cache when there are contentions between the plurality of shared resources;
wherein the one or more applications are added to a shared queue used for scheduling;
wherein upon determining that bandwidth is available, the storage device forwards the received data streams back to the sending host; and
wherein the storage device is configured to:
(i) transfer portions of the data streams where a minimum QoS is 100%;
(ii) skip superfluous portions of the data streams to increase virtual throughput; and
(iii) operate a contention resolution priority queue using a non-volatile memory device.
2. The storage device of claim 1 , wherein the storage device comprises a Solid State Drive (SSD) storage device.
3. The storage device of claim 1 , wherein the plurality of host devices are virtual host devices.
4. The storage device of claim 1 , wherein each data stream corresponds to a partition in the memory array.
5. The storage device of claim 1 , wherein the dynamic management interface is configured to determine one or more minimum QoS requirements for each corresponding QoS of each received data stream.
6. The storage device of claim 5 , wherein the one or more minimum QoS requirements comprises anyone of read bandwidth, read latency, write bandwidth, and write latency requirements.
7. The storage device of claim 1 , wherein the dynamic QoS control module is configured to receive first requests for a first data stream and second requests for a second data stream from the sending host.
8. The storage device of claim 7 , wherein the dynamic QoS control module is configured to acknowledge a total data received from the second requests for the second stream from the sending host.
9. The storage device of claim 7 , wherein the dynamic QoS control module is configured to transfer back to the sending host an actual data for the second stream that is less than the total data received from the second requests for the second stream from the sending host.
10. The storage device of claim 7 , wherein the dynamic QoS control module is operative to schedule the first requests and second requests in a manner that may satisfy a first QoS attribute for the first stream and a second QoS attribute for the second stream.
11. The storage device of claim 6 , wherein each QoS attribute is drawn from at least one or more of the minimum QoS requirements.
12. A storage device, comprising:
one or more communication channels suitable for connection with a plurality of host devices;
a plurality of memory devices within a memory array;
a host interface, configured to receive and transmit multiple data streams from and to the plurality of host devices, each data stream having a corresponding service level agreement (SLA) measured by a corresponding quality of service (QoS);
a plurality of controllers configured to receive commands sent from a sending host of the plurality of host devices;
a dynamic management interface, executed by at least one of the plurality of controllers, configured to manage each received data stream with the corresponding QoS and associate the received data stream to the sending host;
a dynamic QoS control module coupled with the dynamic management interface and the host interface, wherein the dynamic QoS control module is configured to receive first requests for a first data stream and second requests for a second stream from the sending host;
a contention module configured to prioritize one or more data streams of the multiple data streams having one or more contentions with respect to a plurality of shared resources, wherein the one or more data streams are prioritized based on a corresponding QoS requirement;
the contention module coupled with a cache configured to cache data from the data streams, wherein the storage device is configured to manage the cache when there are contentions between the plurality of shared resources;
wherein the one or more data streams are added to a shared queue used for scheduling;
wherein upon determining that bandwidth is available, the storage device forwards the received data streams back to the sending host; and
wherein the storage device is configured to:
(i) transfer portions of the data streams where a minimum QoS is 100%;
(ii) skip superfluous portions of the data streams to increase virtual throughput; and
(iii) operate a contention resolution priority queue using a non-volatile memory device.
13. The storage device of claim 12 , wherein a first data stream has a first level of QoS, wherein a second data stream has a second level of QoS, wherein the first level of QoS is greater than the second level of QoS, and wherein a first priority ranking is assigned to the first data stream based on the first level of QoS being greater than the second level of QoS.
14. The storage device of claim 12 , wherein the dynamic management interface is configured to determine one or more minimum QoS requirements for each corresponding QoS of each received data stream.
15. The storage device of claim 14 , wherein the one or more minimum QoS requirements includes read bandwidth, read latency, write bandwidth, and write latency requirements.
16. The storage device of claim 15 , wherein the dynamic QoS control module is configured to acknowledge a total data received from the second requests for the second stream from the sending host.
17. The storage device of claim 16 , wherein the dynamic QoS control module is configured to transfer back to the sending host an actual data for the second stream that is less than the total data received from the second request for the second stream from the sending host, and wherein the dynamic QoS control module is operative to schedule the first requests and second requests in a manner that may satisfy a first QoS attribute for the first stream and a second QoS attribute for the second stream.
18. A method of managing QoS requirements, comprising:
receiving an identifier for a first data stream and an identifier for a second data stream from a host;
receiving first requests for the first data stream and second requests for the second data stream from the host; and
upon determining that bandwidth is available, forwarding a received data stream back to the host;
wherein the identifier for the first data stream further identifies a total data acknowledged from the first requests for the first data stream from a sending host,
wherein the total data acknowledged from the first requests for the first data stream is greater than an actual data transferred back to the host;
wherein the first and second data streams are prioritized using a contention module based on a corresponding QoS requirement in response to contention with respect to a plurality of shared resources;
caching data from the first and second data streams using a cache coupled with the contention module;
managing the cache when there are contentions between the plurality of shared resources;
wherein the first and second data streams are added to a shared queue used for scheduling;
transferring portions of the first and second data streams where a minimum QoS is 100%;
skipping superfluous portions of the first and second data streams to increase virtual throughput; and
operating a contention resolution priority queue using a non-volatile memory device.