SYNCHRONIZED RELEASE OF RESOURCES USED IN DEFERENTIAL CLOUD SERVICES

• US 20180241804A1
• Filed: 11/08/2017
• Published: 08/23/2018
• Est. Priority Date: 02/22/2017
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Calo, Seraphin B., Freimuth, Douglas M., Le, Franck V., Touma, Maroun, Verma, Dinesh C.
• Granted Patent: US 10,749,941 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 9/466 Transaction processing
• CPC Class Codes: G06F 9/5022 Mechanisms to release resources
• CPC Class Codes: G06F 9/5083 Techniques for rebalancing the load in a distributed system
• CPC Class Codes: H04L 47/741 Holding a request until resources become available
• CPC Class Codes: H04L 47/748 Negotiation of resources, e.g. modification of a request
• CPC Class Codes: H04L 47/781 Centralised allocation of resources
• CPC Class Codes: H04L 47/803 Application aware
• CPC Class Codes: H04L 47/826 Involving periods of time
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers

Abstract

A method is provided of using a set of servers to provide deferential services that have a pre-negotiated time for notice to release the servers. The method includes defining a virtual checkpoint frame interval that is constrained to a duration of up to half of the pre-negotiated time for notice to release the servers. The method includes collecting packets and transactions occurring during the interval that are processed by a current server. The method includes, responsive to an end of the interval, (i) writing, to a shared state database, a state of processing of the packets and transactions occurring during the interval, and (ii) releasing the packets and transactions occurring during the interval. The method includes copying the packets and transactions occurring during the interval, and the state, from the current server to another server for subsequent processing, responsive to an indication of an instance loss on the current server.

Claims

1. A computer program product for using a set of servers to provide deferential services that have a pre-negotiated time for notice to release the servers, the computer program product comprising a non-transitory computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computer to cause the computer to perform a method comprising:

defining, by a deferential cloud service broker, a virtual checkpoint frame interval that is constrained to a duration of up to half of the pre-negotiated time for notice to release the servers;

collecting, by a flow controller, packets and transactions occurring during the virtual checkpoint frame interval that are processed by a current one of the servers;

responsive to an end of the virtual checkpoint frame interval,

(i) writing, by the flow controller to a shared state database, a state of processing of the packets and transactions occurring during the virtual checkpoint frame interval, and

(ii) releasing, by the flow controller, the packets and transactions occurring during the virtual checkpoint frame interval; and

copying, by the flow controller, the packets and transactions occurring during the virtual checkpoint frame interval, and the state, from the current one of the servers to another one of the servers for subsequent processing, responsive to an indication of an instance loss on the current one of the servers.

2. The computer program product of claim 1, wherein the flow controller releases the packets and transactions responsive to (i) the end of the virtual checkpoint frame interval, and (ii) a release notification from a service scheduler.

3. The computer program product of claim 1, further comprising, defining, by the deferential cloud service broker using an obligation policy, a duration for the instance loss.

4. The computer program product of claim 3, further comprising countering, by a service scheduler to the deferential cloud service broker, with a different virtual checkpoint frame interval tailored to a particular one of the deferential services, in place of the virtual checkpoint frame interval.

5. The computer program product of claim 4, further comprising selectively accepting or rejecting, by the deferential cloud service broker, the different virtual checkpoint frame interval countered by the service scheduler.

6. The computer program product of claim 1, wherein the deferential cloud service broker uses an obligation policy to assign resources in pairs or multiples while constraining resource release to a single resource at a time to allow for a gradual compression of the deferential services.

7. The computer program product of claim 6, wherein the deferential cloud service broker uses a pair-wise instance assignment such that responsive to receiving a pre-empt instance signal for pre-empting a first instance in an instance pair, the method further comprises switching to a second instance in the instance pair without incurring any latency.

8. A system for using a set of servers to provide deferential services that have a pre-negotiated time for notice to release the servers, the system comprising:

a deferential cloud service broker for defining a virtual checkpoint frame interval that is constrained to a duration of up to half of the pre-negotiated time for notice to release the servers;

a flow controller for (a) collecting packets and transactions occurring during the virtual checkpoint frame interval that are processed by a current one of the servers, and (b) responsive to an end of the virtual checkpoint frame interval,

(i) writing, to a shared state database, a state of processing of the packets and transactions occurring during the virtual checkpoint frame interval, and

(ii) releasing the packets and transactions occurring during the virtual checkpoint frame interval,

wherein the flow controller copies the packets and transactions occurring during the virtual checkpoint frame interval, and the state, from the current one of the servers to another one of the servers for subsequent processing, responsive to an indication of an instance loss on the current one of the servers.