Disaster Recovery And Other Stuff Jay Weaver, Network Manager @ - PowerPoint PPT Presentation

Disaster Recovery And Other Stuff

� Jay Weaver, � Network Manager @ Bank of Utah � jweaver@bankofutah.com � Chris McIntyre � IT Data Center Manager � cmcintyre@bankofutah.com

� Overview of IT DR Plan � What it used to be like � Locations � Data Centers � Network Topology � Equipment

� AKA Business Continuity Plan � Completed Business Impact Analysis � Found what is important � Management gave time frames � We build systems to accomplish time frames Recovery Group 1 is 0 ‐ 12 hrs Recovery Group 2 is 12 ‐ 24 hrs Recovery Group 3 is 24 ‐ 72 hrs

� 14 Retail Banking Locations � 4 Mortgage Only Locations � 1 Corporate Trust Location � 2 Data Centers

� IBM I Power 7 for core processing � 64 GB of Memory � 1.5 TB of SSD � No HDD � QEDD Data Replication � IBM Flex � 2 Chassis' totaling 7 Compute Nodes � Each Node has 2 Proc with 8 Cores & 132 GB RAM � 10 GB switch / possible 40 GB � NetApp Storage � Model FAS3240 � 54 TB of usable space on 76 spindles � Redundant Controllers � Snapshots � Replication

Internet 10 Mbit 300 Mbit Primary DC Backup DC

� Disaster Declared � IBM QEDD replication is stopped � Scripted network routing changes are run � Fiserv subsystem is started on IBM i � Script to restore Windows servers is run � Creates FlexClones of NetApp data volumes � Attaches cloned data volumes to VMWare via NFS � Turns on Virtual Servers

� The Past � Previous years would take 48 hours PLUS using Evault � IT Personnel would work continuously for 48 hours � Systems were slow � Not sure if everything works, odds and ends. � Present � Group 1 Recovery completed in 6 Minutes � Entire environment functional in under 10 Minutes � Testers were unable to tell the difference

� We know that full DR is possible � Other solutions forced a forklift of core network switches � Less cable mess and expertise � Greater Performance � Expandability � Interoperability – x86 and System I

� Resource Consumption