US 12,724,687 B2 · Managing rate-controlled and scheduled workloads in distributed computing systems

Abstract

Systems and methods for managing rate-controlled and scheduled workloads in distributed computing systems. One example method provides a distributed two-phase algorithm, which utilizes an external timer service and local high-resolution timers to orchestrate time critical workloads on distributed computing systems. The first phase is a planning phase implemented as a recurrent process by means of repeated registrations to the external timer service. During the planning phase, a portion or all of the workload to be performed is broken down into small batches to be assigned to computing instances for execution at a later time. The second phase is a batch execution phase. Each batch planned during the planning phase includes fine-grained instructions as to when each item within the batch is to be executed. The computing instance receiving the batch uses local high-resolution timers to synchronize the execution of items relative to one another.

Claims20

Dependent claims

A workload distribution system for a distributed computing environment, the system comprising:

a message broker cluster including a planning queue;
a state cluster including a plan state database storing a plan;
a synchronization cluster operating a global timer service;
and
an engine cluster including a plurality of computing nodes, wherein each of the plurality of computing nodes is configured to:
execute a planning process that includes:
responsive to receiving a planning timeout from the synchronization cluster:
generating a planning job based on the plan, and
storing the planning job in the planning queue;
and
responsive to receiving an available planning job from the planning queue:
processing the available planning job to generate a first execution batch, and
storing the first execution batch in the plan;
and
while concurrently executing the planning process, execute a batch execution process that includes:
responsive to receiving an execution timeout from the synchronization cluster:
retrieving from the state cluster, based on the execution timeout, a second execution batch, and
processing the second execution batch to execute a workload.

15 more elements

A method for distributing computing workloads in a distributed computing environment, the method comprising:

executing, on each of a plurality of computing nodes, a planning process that includes:
responsive to receiving a planning timeout from a synchronization cluster operating a global timer service:
generating a planning job based on the plan, and
storing the planning job in a planning queue;
and
responsive to receiving an available planning job from the planning queue:
processing the available planning job to generate a first execution batch, and
storing the first execution batch on a state cluster;
and
while concurrently executing the planning process, executing, on each of the plurality of computing nodes, a batch execution process that includes:
responsive to receiving an execution timeout from the synchronization cluster:
retrieving from the state cluster, based on the execution timeout, a second execution batch, and
processing the second execution batch to execute a workload.

10 more elements

Dependencies are inferred from the claim text.

Managing rate-controlled and scheduled workloads in distributed computing systems

Life of this patent

Managing rate-controlled and scheduled workloads in distributed computing systems

Life of this patent

Abstract

Claims20