This graphic shows at the top two labels: "MT1, MT3, and MT4 Auto Migrate Policy = Failure-Recovery" and "MT2 Auto Migrate Policy = Exactly Once."

Under those labels are three boxes labeled "Machine 1," Machine 2," and Machine 3."

Machine 1 contains an oval labeled "Node Manager" and a pentagon labeled "Managed Server A." Managed Server A contains a dotted line oval containing the label "MT3 JMSServerB QB1, QB2 Store" and another oval containing the label "MT1 JMSServerA QA1, QA2 Store."

Machine 2 contains an oval labeled "Node Manager" and a pentagon labeled "Managed Server B." Managed Server B contains an oval containing the label "MT3 JMSServerB QB1, QB2 Store" and another oval containing the label "MT2 Path Service Store." Managed Server B is crossed out.

Machine 3 contains an oval labeled "Node Manager" and a pentagon labeled "Managed Server C." Managed Server C contains an oval containing the label "MT4 JMSServerB QC1, QC2 Store" and a dotted line oval containing the label "MT2 Path Service Store."

MT2 Path Service Store on Managed Server B points to MT2 Path Service Store on Managed Server C with an arrow labeled "Auto migrates to any available server so it is always active in cluster."

Each machine is connected to a database, labeled SAN/Dual-ported SCSI or JDBC Database," with double-headed arrows.