scheduler interfaces in the askalon grid environment

Post on 21-Jan-2016

29 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

DESCRIPTION

Scheduler Interfaces in the ASKALON Grid Environment. Marek Wieczorek, GGF16 Athens, February 14th. Agenda. Scheduling in the ASKALON environment Scheduler interfaces Interesting scenarios Summary. ASKALON Grid Environment. ASKALON. AGWL (Workflow Representation). User Interface. - PowerPoint PPT Presentation

TRANSCRIPT

Scheduler Interfaces in the ASKALON Grid

Environment

Marek Wieczorek, GGF16 Athens, February 14th

Agenda

•Scheduling in the ASKALON environment

•Scheduler interfaces

• Interesting scenarios

•Summary

ASKALON Grid Environment

EnactmentEngine

ResourceManagement

PerformancePredictor

MonitoringService

UserInterface

AGWL(Workflow Representation)

Grid Environment (Globus Toolkit)

ASKALON

Scheduler

Scheduling Model

• Scientific workflow scheduler

• Separate instance for each workflow submission

• Interacts with other ASKALON services

• Dynamic scheduling (rescheduling enabled)

Scheduler Input

• Workflow specification created by the user

• A workflow composed of activity types and dependencies between them

• ASKALON workflow description language (AGWL)

Use

r In

terf

ace

(TEU

TA

)

Sch

edule

rworkflow specificationreceive: AGWL file (XML)protocol: GT4 WS interface

Interface to the Resource Management (GridARM)

• Resource brokerage based on resource constraints (e.g., CPU count, memory)

• Logical resources considered, i.e. software components (executables Grid/Web services)

• Advance reservation support

• Billing/pricing

Sch

edule

r

Reso

urc

eM

anagem

en

t

resource queryreceive: resource listprotocol: GT4 WS interface

service queryreceive: service descriptionprotocol: GT4 WS interface

reservation requestreceive: reservation offersprotocol: GT4 WS interface

reservation confirmationreceive: reservation ticketprotocol: GT4 WS interface

Interface to the Enactment Engine

• Enactment Engine – advanced Job Manager

• Communication protocol based on the internal workflow representation

• Current implementation based on the Java API

Sch

edule

r

Enact

ment

Eng

ine

workflow submissionsend: full workflow graphprotocol: GT4 WS interface

rescheduling requestreceive: executed workflowsend: rescheduled workflowprotocol: GT4 WS interface

Interface to the Performance Predictor

• Auxiliary service

• Important for proper scheduling and for advance reservations

• Model: workflow-oriented activity predictions

Sch

edule

r

Perf

orm

ance

Pre

dic

tor

prediction requestsend: activity type, resource, input parametersreceive: predictionprotocol: Java API/GT4 WS

Interface to the Monitoring Service

• Not implemented yet

• Monitoring of resources and activities

• Control over many aspects of execution (including scheduling)

• Possible execution problems reported

• May trigger workflow rescheduling

Sch

edule

r

Monit

ori

ng

Serv

ice

event notificationreceive: eventprotocol: GT4 WS interface

scheduling eventsend: scheduling/ reschedulingprotocol: GT4 WS interface

Workflow Processing Lifecycle

Scheduler EnactmentEngine

PerformancePredictor

AGWL workflowspecification

ResourceManagement

resource/servicequery

MonitoringService

prediction request

negotiation/reservation

scheduled workflow

rescheduled workflow

rescheduling request

event notification

Scheduling under Uncertainty

• Full-graph scheduling

• ASKALON workflow model (loops, conditional branches, etc.) allow different possible executions

• Runtime assumption made by the scheduler

• Assumption fulfilment controlled by the enactment engine

Scheduler EnactmentEngine

originalworkflow

predicted DAG

+assumptionviolation=

rescheduling

+assumptions=

rescheduled workflow

Advance Reservation

•Negotiations between the scheduler and the Reservation Management– The scheduler acts on behalf of the user,– Resource Management optimizes the

profit of the Grid (resource usage, fair-sharing)

•Two-stage negotiation protocol:– reservation request,– reservation confirmation

Event-driven Dynamic Scheduling

• Different sources of events

• Scheduling and rescheduling performed in a dynamic way.

• Two types of reasons for rescheduling:– execution model

(assumption violation)– unexpected behiavor

(execution problems)

Scheduler

Enactment Engine

ResourceManagement

MonitoringService

assumption violation,execution problem

SLA violation performancebreakdown

Summary

• Importance of the scheduler in the ASKALON environment.

•Multitude of the scheduler interfaces

•Resource negotiation and event-driven scheduling

ASKALON

http://dps.uibk.ac.at/projects/http://dps.uibk.ac.at/projects/askalon/askalon/

top related