Origin failover for live streaming

Patent metadata

Field	Value
Publication number	US-9866459-B1
Application number	US-201615061890-A
Country	US
Kind code	B1
Filing date	Mar 4, 2016
Priority date	Mar 4, 2016
Publication date	Jan 9, 2018
Grant date	Jan 9, 2018

Abstract

Techniques are described for automatically migrating client devices away from unreliable origin stacks in ways that mitigate potentially negative impacts on the user experience. Multiple origin stacks are provisioned in parallel to provide encoded versions of the same broadcast content for streaming to client devices. Each of the origin stacks is monitored to ensure that it is operating correctly. When an origin stack fails or becomes unreliable, each of the client devices involved in a session corresponding to the unreliable stack is provided with the information it needs to establish a new session corresponding to another one of the stacks.

First claim

What is claimed is:

1. A computer-implemented method, comprising:
provisioning a plurality of origin stacks, each origin stack including a set of computing resources configured to receive broadcast content, and encode the broadcast content;
for each origin stack, generating one or more streams of the encoded broadcast content for consumption by a corresponding plurality of client devices;
providing first manifest information to a first client device, the first manifest information being configured to enable the first client device to access a first stream of the encoded broadcast content originating from a first origin stack;
providing the first stream of the encoded broadcast content to the first client device;
monitoring each of the origin stacks to determine whether each of the origin stacks is reliably generating the corresponding one or more streams of the encoded broadcast content;
determining that the first origin stack is not reliably generating the corresponding one or more streams of the encoded broadcast content;
in response to determining that the first origin stack is not reliably generating the corresponding one or more streams of the encoded broadcast content, providing second manifest information to the first client device, the second manifest information being configured to enable the first client device to transition to a second stream of the encoded broadcast content originating from a second origin stack; and
providing the second stream of the encoded broadcast content to the first client device such that presentation of the broadcast content on the first client device is substantially uninterrupted.

2. The method of claim 1 , wherein providing the first stream of the encoded broadcast content to the first client device is done via a first content distribution network in communication with the first origin stack, and wherein providing the second stream of the encoded broadcast content to the first client device is done via the first content distribution network in communication with the second origin stack or via a second content distribution network in communication with the second origin stack.

3. The method of claim 1 , wherein monitoring each of the origin stacks includes, for each origin stack, acquiring one of the corresponding one or more streams of the encoded broadcast content with a monitoring client operating in a same data center with the origin stack.

4. The method of claim 3 , wherein the monitoring client associated with each origin stack includes playback error detection logic configured to generate event messages representing playback errors, and wherein determining that the first origin stack is not reliably generating the corresponding one or more streams of the encoded broadcast content includes receiving a first event message representing a first playback error from a first monitoring client associated with the first origin stack.

5. A system, comprising one or more computing devices configured to:
provision a plurality of origin stacks, each origin stack including a set of computing resources configured to encode broadcast content;
for each origin stack, generate one or more streams of the encoded broadcast content for consumption by a corresponding plurality of client devices;
provide first manifest information to a first client device, the first manifest information being configured to enable the first client device to access a first stream of the encoded broadcast content originating from a first origin stack;
provide the first stream of the encoded broadcast content to the first client device;
monitor each of the origin stacks to determine whether each of the origin stacks is reliably generating the corresponding one or more streams of the encoded broadcast content;
determine that the first origin stack is not reliably generating the corresponding one or more streams of the encoded broadcast content; and
in response to determining that the first origin stack is not reliably generating the corresponding one or more streams of the encoded broadcast content, migrate the plurality of client devices corresponding to the first origin stack to one or more of the other origin stacks, wherein migration of the plurality of client devices includes providing second manifest information to the first client device, the second manifest information being configured to enable the first client device to transition to a second stream of the encoded broadcast content originating from a second origin stack.

6. The system of claim 5 , wherein the one or more computing devices are further configured to provide a first stream of the encoded broadcast content to a first client device via a first content distribution network in communication with the first origin stack, and wherein the one or more computing devices are configured to migrate the plurality of client devices corresponding to the first origin stack to one or more of the other origin stacks, in part, by providing a second stream of the encoded broadcast content to the first client device via the first content distribution network in communication with a second origin stack or via a second content distribution network in communication with the second origin stack.

7. The system of claim 5 , wherein the one or more computing devices are configured to monitor each of the origin stacks by, for each origin stack, acquiring one of the corresponding one or more streams of the encoded broadcast content with a monitoring client operating in a same data center with the origin stack.

8. The system of claim 7 , wherein the monitoring client associated with each origin stack includes playback error detection logic configured to generate event messages representing playback errors, and wherein the one or more computing devices are configured to determine that the first origin stack is not reliably generating the corresponding one or more streams of the encoded broadcast content by receiving a first event message representing a first playback error from a first monitoring client associated with the first origin stack.

9. The system of claim 5 , wherein the one or more computing devices are configured to migrate the plurality of client devices corresponding to the first origin stack to one or more of the other origin stacks, in part, by providing first transition information to the first client device, the first transition information representing a first transition point in the broadcast content for transitioning playback of the broadcast content away from the first stream of the encoded broadcast content to the second stream of the encoded broadcast.

10. The system of claim 9 , wherein the one or more computing devices are further configured to migrate the plurality of client devices corresponding to the first origin stack to one or more of the other origin stacks, in part, by providing second transition information to a second client device, the second transition information representing a second transition point in the broadcast content for transitioning playback of the broadcast content, the second transition point being different from the first transition point.

11. The system of claim 5 , wherein the one or more computing devices are configured to monitor each of the origin stacks by, for each origin stack, repeatedly requesting health status information from at least one of the computing resources associated with the origin stack.

12. The system of claim 5 , wherein the one or more computing devices are configured to migrate the plurality of client devices corresponding to the first origin stack to one or more of the other origin stacks, in part, by notifying the client devices corresponding to the first

Assignees

• Amazon Tech Inc

Inventors

• Joliveau Marc
• Burford Kenneth Thomas
• George Elliot Fanwar
• Koceski Kyle Bradley
• Nann Paul Simon
• Richter Stefan Christian

Classifications

• H04L65/4069

  Electricity · mapped topic

• H04L43/0823 Primary

  Errors, e.g. transmission errors · CPC title

• H04L45/22

  Alternate routing · CPC title

• H04L41/0654

  using network fault recovery (ring fault isolation or reconfiguration in loop networks without recovery actions by a network management system H04L12/437) · CPC title

• H04N21/2404 Primary

  Monitoring of server processing errors or hardware failure (error or fault detection G06F11/07; monitoring in general G06F11/30) · CPC title