Systems, methods, and apparatuses for implementing a streaming platform IO pump and regulator

• US 9,183,090 B2
• Filed: 03/15/2013
• Issued: 11/10/2015
• Est. Priority Date: 10/10/2011
• Status: Active Grant

• Current Assignee: Salesforce.com, Inc.
• Original Assignee: Salesforce.com, Inc.
• Inventors: Spencer, Barry, Egenberger, Jeremy
• Primary Examiner(s): Chan, Wing F
• Assistant Examiner(s): Woo, Andrew
• Application Number:
  US13/840,347
• Publication Number:
  US 20130268807A1
• Time in Patent Office: 970 Days
• Field of Search: 709/223, 709/224, 709/226
• US Class Current: 1/1
• CPC Class Codes: G06F 11/1415 at system level
• CPC Class Codes: G06F 11/1443 Transmit or communication errors
• CPC Class Codes: H04L 43/0882 Utilisation of link capacity
• CPC Class Codes: H04L 65/401 wherein the services involve a main real-time session and one or more additional parallel real-time or time sensitive sessions, e.g. white board sharing or spawning of a subconference
• CPC Class Codes: H04L 65/612 for unicast
• CPC Class Codes: H04L 65/80 Responding to QoS
• CPC Class Codes: H04L 69/16 Implementation or adaptation of Internet protocol [IP], of transmission control protocol [TCP] or of user datagram protocol [UDP]

Abstract

In accordance with disclosed embodiments, there are provided methods, systems, and apparatuses for implementing a streaming platform IO pump and regulator including, for example, means for executing an application at the computing device; requesting data for the application from a second computing device communicably interfaced with the first computing device via a public Internet; calculating an estimated arrival time for the data to be returned from the second computing device; determining the data fails to arrive within the estimated arrival time calculated; initiating a new connection to the second computing device; and re-sending the request for the data for the application to the second computing device via the new connection. Other related embodiments are disclosed.

Claims

1. A method at a first computing device, the method comprising:

executing an application at the first computing device;

requesting data for the application from a second computing device communicably interfaced with the first computing device via a public Internet;

calculating an estimated arrival time for the data to be returned from the second computing device;

determining the data fails to arrive within the estimated arrival time calculated;

initiating a new connection to the second computing device;

re-sending the request for the data for the application to the second computing device via the new connection;

determining the data arrives on the new connection responsive to re-sending the request for the data to the second computing device via the new connection within the estimated arrival time calculated;

consuming the data at the first computing device via the application;

sending an acknowledgement to the second computing device via the new connection;

receiving a second copy of the data via an original connection subsequent to the data arriving initially on the new connection; and

discarding the second copy of the data as waste.

2. The method of claim 1, wherein requesting data from the second computing device communicably interfaced with the first computing device via a public Internet comprises:

initiating a Transmission Control Protocol (TCP) connection exclusively for the application between the first computing device and the second computing device via the public Internet.

3. The method of claim 2, wherein initiating the new connection to the second computing device comprises initiating a second TCP connection exclusively for the application between the first computing device and the second computing device via the public Internet.

4. The method of claim 2, wherein initiating the TCP connection comprises:

selecting the TCP connection from a plurality of multiple simultaneous TCP connections in a connection pool exclusively allocated to the application; and

initiating the selected TCP connection.

5. The method of claim 1, further comprising:

receiving the second copy of the data via the original connection subsequent to re-sending the request for the data to the second computing device via the new connection and in addition to receiving the data via the new connection;

wherein consuming the data at the first computing device via the application comprises consuming either the data received via the new connection or the second copy of the data received via the original connection based on whichever arrives first, wherein both the data received via the new connection or the second copy of the data received via the original connection are identical copies of the same data; and

wherein discarding the second copy of the data as waste comprises discarding the non-consumed copy of the data as waste.

6. The method of claim 1, wherein the application comprises a desktop sharing application requiring a lossless low-latency connection.

7. The method of claim 6, wherein the desktop sharing application exchanges I-Frames and P-Frames with the second computing device as part of a video stream for the desktop sharing application.

8. The method of claim 6:

wherein the first computing device is a presenter computing device for the desktop sharing application; and

wherein the second computing device is one of;

(a) a cloud services platform operating at a host organization remote from the first computing device, wherein the cloud services platform accepts a video stream for the desktop sharing application from the presenter computing device and distributes the video stream to one or more other remote computing devices on behalf of the presenter computing device,

(b) a server remote from the first computing device, wherein the server accepts the video stream for the desktop sharing application from the presenter computing device and distributes the video stream to one or more other remote computing devices on behalf of the presenter computing device, and

(c) a peer computing device operating as a viewer computing device, wherein the viewer computing device accepts the video stream for the desktop sharing application from the presenter computing device and consumes the video stream by presenting the video stream at a display of the viewer computing device.

9. The method of claim 1, further comprising:

monitoring open connections in a connection pool;

determining an original connection over which the data was initially requested remains open and has exceeded an open connection timeout; and

forcibly killing a thread associated with the original connection to terminate the connection and release the terminated connection back into the connection pool.

10. The method of claim 1, wherein calculating the estimated arrival time is based at least in part on the following:

latency for the requested data to be transmitted from the first computing device to the second computing device;

latency for the second computing device to process the request for the data from the first computing device;

latency for the requested data to be transmitted from the second computing device to the first computing device; and

a grace period to accommodate connection variance.

11. The method of claim 10, wherein the estimated arrival time is based further on a packet size corresponding to one or both of the data requested by the first computing device and the data returned from the second computing device.

12. The method of claim 1, wherein calculating the estimated arrival time, comprises:

iteratively adjusting the estimated arrival time for data packets communicated in support of the application between the first computing device and the second computing device based at least in part on observed latency between the first computing device and the second computing device.

13. The method of claim 1, further comprising:

determining a nominal packet size for the application;

wherein requesting the data from the second computing device comprises generating a request packet having a size proportional to the nominal packet size for the application; and

adjusting the nominal packet size upwards responsive to low latency observed for communications on behalf of the application and downwards responsive to high latency and packet loss observed for the communications on behalf of the application.

14. The method of claim 13, further comprising:

forcing a reduction to bandwidth communicated for the application when the data fails to arrive within the estimated arrival time calculated.

15. Non-transitory computer readable storage media having instructions stored thereon that, when executed by a processor in a first computing device, the instructions cause the first computing device to perform operations comprising:

executing an application at the first computing device;

requesting data for the application from a second computing device communicably interfaced with the first computing device via a public Internet;

calculating an estimated arrival time for the data to be returned from the second computing device;

determining the data fails to arrive within the estimated arrival time calculated;

initiating a new connection to the second computing device; and

re-sending the request for the data for the application to the second computing device via the new connection;

determining the data arrives on the new connection responsive to re-sending the request for the data to the second computing device via the new connection within the estimated arrival time calculated;

consuming the data at the first computing device via the application;

sending an acknowledgement to the second computing device via the new connection;

receiving a second copy of the data via an original connection subsequent to the data arriving initially on the new connection; and

discarding the second copy of the data as waste.

16. The non-transitory computer readable storage media of claim 15, wherein the operations further comprise:

determining the data arrives on the new connection responsive to re-sending the request for the data to the second computing device via the new connection within the estimated arrival time calculated;

consuming the data at the first computing device via the application; and

sending an acknowledgement to the second computing device via the new connection.

17. A computing device comprising:

a processor to execute an application, the application being operable to communicate with a second computing device over a public Internet;

an Input/Output (I/O) pump to allocate multiple simultaneous Transmission Control Protocol (TCP) connections in support of the application; and

the Input/Output (I/O) pump to further request data for the application from the second computing device communicably interfaced with the computing device via a public Internet on a first one of the multiple simultaneous TCP connections;

a regulator module to calculate an estimated arrival time for the data to be returned from the second computing device via the first one of the multiple simultaneous TCP connections;

the Input/Output (I/O) pump to determine the data has failed to arrive within the estimated arrival time calculated;

the Input/Output (I/O) pump to initiate a new TCP connection to the second computing device from among the multiple simultaneous TCP connections; and

the Input/Output (I/O) pump to re-send the request for the data for the application to the second computing device via the new TCP connection;

the Input/Output (I/O) pump to determine the data arrives on the new connection responsive to re-send of the request for the data to the second computing device via the new connection within the estimated arrival time calculated;

the application to consume the data received via the new connection;

the Input/Output (I/O) pump to send an acknowledgement to the second computing device via the new connection;

the Input/Output (I/O) pump to receive a second copy of the data via an original connection subsequent to the data having arrived on the new connection; and

the Input/Output (I/O) pump to discard the second copy of the data as waste.

18. The computing device of claim 17, wherein the I/O pump is to allocate the multiple simultaneous TCP connections in support of the application, wherein a connection pool of the I/O pump is to maintain the multiple simultaneous TCP connections exclusively on behalf of the application.

19. A method at a first computing device, the method comprising:

receiving at the first computing device, a request for data from a second computing device communicatively interfaced via a public Internet;

retrieving or generating the data requested;

buffering the data requested into a queue;

sending the data requested to the second computing device;

receiving a new request for the data from the second computing device via a new connection subsequent to the second computing device failing to receive the data requested from the first computing device at the second computing device via an original connection within an estimated arrival time for the data to be returned from the first computing device as calculated by the second computing device;

forwarding the requested data to the second computing device from the queue via the new connection without re-retrieving or re-generating the requested data;

receiving an acknowledgement for the requested data from the second computing device via the new connection pursuant to the second computing device having determined the requested data arrived via the new connection within the estimated arrival time calculated, wherein an application at the second computing device consumes the requested data arriving via the new connection; and

releasing the requested data from the queue of the first computing device and wherein the second computing device discards a second copy of the data which arrives via the original connection as waste.

20. The method of claim 19:

wherein receiving the request for the data from the second computing device communicatively interfaced via the public Internet comprises receiving the request via a Transmission Control Protocol (TCP) connection between the first computing device and the second computing device via the public Internet; and

wherein receiving the new request for the data from the second computing device via the new connection comprises receiving the new request via a second TCP connection different than a first TCP connection corresponding to the original connection.

21. The method of claim 19:

wherein the second computing device is a presenter computing device for a desktop sharing application executing at the second computing device; and

wherein the first computing device is one of;

(a) a cloud services platform operating at a host organization remote from the second computing device, wherein the cloud services platform accepts a video stream for the desktop sharing application from the presenter computing device and distributes the video stream to one or more other remote computing devices on behalf of the presenter computing device,

(b) a server remote from the second computing device, wherein the server accepts the video stream for the desktop sharing application from the presenter computing device and distributes the video stream to one or more other remote computing devices on behalf of the presenter computing device, and

(c) a peer computing device operating as a viewer computing device, wherein the viewer computing device accepts the video stream for the desktop sharing application from the presenter computing device and consumes the video stream by presenting the video stream at a display of the viewer computing device.

22. The method of claim 19, wherein calculating the estimated arrival time is based at least in part on the following:

latency for the requested data to be transmitted from the first computing device to the second computing device;

latency for the second computing device to process the request for the data from the first computing device;

latency for the requested data to be transmitted from the second computing device to the first computing device; and

a grace period to accommodate connection variance.

23. The method of claim 22, wherein the estimated arrival time is based further on a packet size corresponding to one or both of the data requested by the first computing device and the data returned from the second computing device.