Case Study 22: Utility Computing with Diskless Provisioning

Moab Workload Manager^®

A.22 Case Study: Service Virtualization

Overview

Solution

Moab Utility Computing Suite can enable the creation of the fully customized per request environment by utilizing resource monitors, provisioning systems, and storage and network managers. Tightly or loosely scheduled automated batch runs on utility computing resources can also be enabled using standing triggers or standing reservations.

The configuration below will enable Moab to schedule a weekly accounting package and enable an analysis service during business hours.

moab.cfg

SCHEDCFG[master] SERVER=ww.ifl.com MODE=NORMAL

ENABLESERVICEPROVISIONING  TRUE

# interface to monitor/manage services
RMCFG[direct]   TYPE=native:warewulf

# constrain scheduler to only allocate same architecture nodes to each job
NODESETPOLICY    ONEOF
NODESETATTRIBUTE ARCH

# run accounting package once a week
SRCFG[acct]  HOSTLIST=node013,node014 
SRCFG[acct]  DAYS=FRI STARTTIME=22:00:00 ENDTIME=24:00:00
SRCFG[acct]  OS=rh73 APPLICATION=/usr/acct/tools/weeklysummary.pl

# activate web analysis service during business hours
SRCFG[web3]  TASKCOUNT=16
SRCFG[web3]  DAYS=MON,TUE,WED,THU,FRI STARTTIME=08:00:00 ENDTIME=17:00:00
SRCFG[web3]  OS=sus91 APPLICATION=/opt/w3/bin/webanalysis

# prevent oversubscription of provisioning resources
NODECFG[GLOBAL] GRES=prov:32 # allow 32 simultaneous provisioning operations
CLASSCFG[prov] TGRES=prov:1@0-300

To submit a batch application request which requires operating system provisioning, use standard batch submission commands.

msub

> msub -l nodes=1,walltime=300,arch=x86,os=suse91 applaunch:data3.txt

moab.1043 submitted

Moab will schedule applications across the cluster, grid, or utility computing resource and will package the application with the requested operating system. As needed, Moab will reprovision resources to provide the bundled OS/application on the best available compute node.

With this model, both batch and rigidly scheduled applications can be inter-mixed.

highlights

resource allocation optimized to minimize provisioning overhead
forced provisioning option to allow added security
per architecture image tracking
configurable policies to prevent oversubscription of provisioning resources
automated failure recovery