System and method for managing a computing cluster

• US 9,459,926 B2
• Filed: 05/09/2012
• Issued: 10/04/2016
• Est. Priority Date: 06/23/2010
• Status: Active Grant

• Current Assignee: Twilio, Inc.
• Original Assignee: Twilio, Inc.
• Inventors: Shakirzyanov, Bulat, Davidson, Keith A., Cooke, Evan, Lawson, Jeffrey G.
• Primary Examiner(s): Chang, Tom Y
• Application Number:
  US13/467,766
• Publication Number:
  US 20120271926A1
• Time in Patent Office: 1,609 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 2209/505 Clust
• CPC Class Codes: G06F 9/5072 Grid computing
• CPC Class Codes: H04L 67/142 Managing session states for stateless protocols; Signalling session states; State transitions; Keeping-state mechanisms
• CPC Class Codes: H04M 3/22 Arrangements for supervision, monitoring or testing
• CPC Class Codes: H04M 7/00 Arrangements for interconnection between switching centres
• CPC Class Codes: H04Q 3/00 Selecting arrangements H04Q5/00 - H04Q11/00 take precedence

Abstract

A method and system for managing a computing cluster including hosting a plurality of machines in a networked computing cluster, wherein the plurality of machines include service instances running on hosts, where the services have configured machine state; and wherein machine state includes configuration data and software of the machine; integrating the plurality of machines of the cluster with at least one configuration controller component; and the at least one configuration controller component, altering the machine state of at least one service instance.

Claims

1. A method comprising:

at a telephony network platform system that includes a cloud manager, a database, and a monitor module, the cloud manager including a pub-sub messaging system, the cloud manager using the pub-sub messaging system for communication with a remote third-party Internet-based multi-tenant computing platform via the Internet, the multi-tenant computing platform being constructed to provide at least one machine group that includes a plurality of machines for implementing a cloud-based telephony platform that provides one of a plurality of telephony platform services including a voice communication service, a messaging communication service and a video communication service;

at the cloud manager, receiving a first user instruction to create a first telephony platform machine group that includes a plurality of machines of the remote Internet-based multi-tenant computing platform, the first user instruction being received from an outside system via a REST API of the cloud manager, the outside system being an external system of a user of the remote multi-tenant computing platform;

responsive to receipt of the first user instruction, the cloud manager storing a first machine group record in the database in accordance with the first user instruction, the first machine group record comprising a group configuration and a group definition, the group configuration comprising a set of rules, the group definition defining the machines of the first machine group, in a domain specific language, the first machine group record relating to an initial state of the first machine group as defined by at least one of the group definition and the group configuration, the initial state of the first machine group record specifying first telephony platform software for a first telephony platform service, the first telephony platform service being one of the voice communication service, the messaging communication service, and the video communication service;

the cloud manager establishing a cloud manager publication channel with the pub-sub messaging system, the pub-sub messaging system being constructed to provide the first telephony platform software via the Internet to each machine of the first machine group, establishing the cloud manager publication channel comprising;

the cloud manager storing a security key of the first machine group and a security key for each machine of the first machine group, the security key for each machine of the first machine group being supplied at boot time of the respective machine and provided to the cloud manger via the pub-sub messaging system, the security key of each machine being used to securely identify the machine;

at the cloud manager, receiving a second user instruction to start the first telephony platform machine group, the second user instruction being received from the outside system via the REST API of the cloud manager;

at the cloud manager and responsive to receipt of the second user instruction to start the telephony platform machine group, the cloud manager launching the first telephony platform machine group at the remote computing platform by using the pub-sub messaging system, launching the first telephony platform machine group comprising;

for each machine of the first telephony platform machine group;

encrypting the first telephony platform software of the first machine group record with the stored security key of the machine,

providing the encrypted first telephony platform software to the machine via the cloud manager publication channel,

wherein responsive to the machine decrypting the encrypted first telephony platform software by using the security key, the machine executes the first telephony platform software, the first telephony platform software being constructed to control the machine to provide at least a portion of the first telephony platform service;

at the cloud manager, the cloud manager controlling the monitor module to generate a monitoring loop to check status of the first telephony platform machine group at the remote computing platform, the monitor module comprising a rules engine for stateless application of the set of rules of the group configuration of the first telephony platform machine group;

the monitor module checking actual status of the first telephony platform machine group at the remote computing platform, and comparing the actual status with a desired state of the first telephony platform machine group as indicated by the first machine group record stored in the database, comparing the actual status with the desired state comprising;

determining whether each machine of the first telephony platform machine group is executing telephony platform software specified by the first machine group record; and

at the cloud manager, receiving a third user instruction to change operation of the machines of the first telephony platform machine group, the third user instruction being received from the outside system via the REST API of the cloud manager;

at the cloud manager and responsive to receipt of the third user instruction to change operation of the machines, the cloud manager updating the first machine group record in the database in accordance with the third user instruction, updating the first machine group record comprising;

updating the first machine group record to specify second telephony platform software for a second telephony platform service, the second telephony platform service being one of the plurality of telephony platform services that is different from the first telephony platform service;

responsive to a determination by the monitor module that at least one machine of the first telephony platform machine group is executing the first telephony platform software and the first machine group record specifies the second telephony platform software, the monitor module adjusting the actual status of the first telephony platform machine group at the remote computing platform to match the desired state by using the rules engine to apply the set of rules of the group configuration, adjusting the actual status comprising;

for each machine of the first telephony platform machine group;

encrypting the second telephony platform software of the first machine group record with the stored security key of the machine,

providing the encrypted second telephony platform software to the machine via the cloud manager publication channel to replace the first telephony platform software with the second telephony platform software at the machine, the machine decrypting the encrypted second telephony platform software by using the security key of the machine and the machine replacing execution of the first telephony platform software with execution of the second telephony platform software to provide at least a portion of the second telephony platform service,

wherein responsive to the machine decrypting the encrypted second telephony platform software by using the security key, the machine executes the second telephony platform software.

2. The method of claim 1, further comprising compiling the group definition into an internal representation at the cloud manager.

3. The method of claim 1, wherein the set of rules comprises a condition class and an action class.

4. The method of claim 1, wherein the group configuration and the group definition are stateless.

5. The method of claim 1, further comprising the monitor module adjusting the desired state at the database in accordance with the rules engine in response to a change in the actual status of the telephony platform group at the remote computing cluster.

6. The method of claim 1, wherein the monitoring loop is an infinite loop.

7. The method of claim 1, further comprising:

the monitor module checking the desired state of the telephony platform group as indicated by the group record stored at the database; and

responsive to a determination by the monitor module that the desired state has been updated, the monitor module adjusting the actual status of the telephony platform group at the remote computing cluster to match the updated desired state by using the rules engine to apply the set of rules of the group configuration.

8. The method of claim 7, further comprising:

responsive to an API request to update the desired state, the cloud manager updating the desired state at the database in accordance with the API request.

9. The method of claim 1, wherein the monitor module accesses the second telephony platform software from the database.

10. The method of Claim 9, wherein the set of rules of the first machine group record includes a rule that when applied by the rules engine controls the monitor module to provide the encrypted second telephony platform software to each machine of the first telephony platform machine group responsive to a determination that sustained communication sessions of the first telephony platform service have ended.

11. The method of claim 10, wherein the first telephony platform software and the second telephony platform software are each constructed to control each machine of the first telephony platform machine group to publish a machine state notification of the machine to the monitor module via the Internet responsive to a change in a machine state of the machine, the machine publishing the machine state notification by using a publication channel of the pub-sub messaging system.

12. The method of Claim 11, wherein the first telephony platform software is replaced during operation of the first telephony platform service.