Virtualization

Implementing automated replication for cost-effective disaster recovery

By Fabian Salamanca and Javier L. Jiménez

Dell™ Advanced Infrastructure Manager (AIM) software and Compellent™ storage together help create a reliable disaster recovery platform that supports multisite replication and optimizes business continuity cost-effectively.

echnology forms a front line of defense The first step in creating a business continuity Dive deeper: Multisite replication when enterprises require continuous plan is to identify critical applications and functions availability for business-critical that must be protected. Next, organizations A successful business continuity applications and data. As a result, IT delineate the recovery time objective (RTO), plan helps ensure availability of T business-critical applications. efforts are intensifying to establish protection from which specifies the maximum allowable time to Download this Dell technical sudden disruption or downtime in mission-critical restore each critical process. Then, they define the white paper, and discover how Dell AIM and applications such as e-mail, Internet presence, recovery point objective (RPO), which targets the storage provide an efficient enterprise resource planning (ERP), and customer maximum acceptable amount of data at risk of disaster recovery platform using relationship management (CRM). Although high- loss after an adverse event occurs. Time to data multisite replication. availability clustering provides local protection, critical (TTD) is the time required for retrieving backup dell.to/s9UZI0 applications also require geographical protection. data and delivering it to the recovery site. And the stakes for preserving business continuity are RTO and RPO are key measures that drive the high: among organizations that experience a major configuration of a disaster recovery implementation, loss of business data, a significant number face critical which also affects its cost. Figure 1 shows problems and only a few are able to overcome them. the relationship among RTO, RPO, and TTD.

72 2011 Issue 04 | dell.com/powersolutions Reprinted from Dell Power Solutions, 2011 Issue 4. Copyright ­© 2011 Dell Inc. All rights reserved. Virtualization

Reduced RTO and RPO can translate Data Restore into an enhanced business continuity Snapshot Disaster restoration operation Snapshot response and a cost-effective disaster

recovery implementation. Time One approach to creating a cost- RPO effective business continuity deployment leverages Dell Advanced Infrastructure Manager (AIM)—a component of the Dell TTD Virtual Integrated System (VIS) portfolio— together with Dell Compellent Storage Center™ RTO storage area network (SAN) arrays. This architecture is designed to provide reliable Figure 1. Mapping RTO, RPO, and TTD in a disaster recovery scenario data replication, OS image integrity, and fast workload migration. The example either a virtual or a physical . Included center assets and allows IT organizations to configuration described in the section persistent identification settings help leverage existing investments for increased “Efficiently implementing disaster recovery” ensure the persona has access to the same return on investment (ROI). deploys for OS image and resources no matter what changes may data access, IP and networks for occur in a managed with AIM. Intelligently managing and replicating long-distance wide area network (WAN)– AIM network-booted personas can enterprise data based replication using Internet SCSI boot on any validated resource, and AIM Using Dell Compellent storage helps (iSCSI) connectivity, and AIM to manage helps ensure that whichever resource is administrators actively and intelligently the provisioning of workload identities and running the persona, that resource has manage enterprise data throughout its life network configuration. the required network and storage access. cycle, enabling organizations to constantly For example, at any given time a network- adapt to changing conditions. Together, Flexibly managing recovery and booted persona can be running on a the Dell Fluid Data architecture, storage workload mobility physical server, such as a Dell PowerEdge™ virtualization, intelligent software, and modular Dell AIM helps IT organizations build a R610 rack server, and after business hours it hardware enhance IT efficiency, simplicity, next-generation data center by providing can run on either a VMware® virtualization– and security for enterprise storage. Built-in rapid workload mobility and recovery based or a Microsoft® Hyper-V™–based intelligence and automation help ensure data management across physical and virtual virtual machine to help reduce power is available when and where it is needed, and machines in heterogeneous environments. consumption and cooling costs. This an open, persistent hardware platform scales IT administrators can deploy AIM to create dynamic capability enables the workload to in line with business needs to help protect a cost-effective IT infrastructure that use low-cost hardware or virtual machines long-term storage procurements. easily changes—physically and virtually—as when the load is expected to be low and Leveraging Remote Instant Replay for circumstances demand. It helps increase IT retarget to high-performance resources Dell Compellent array replication across asset utilization, can recover services quickly when the load is expected to be high. multiple sites helps reduce the cost and if problems occur, and enables services to be All data related to a network-booted complexity of replication while redefining scaled rapidly to support business needs. persona resides on a SAN, leveraging the backup and recovery. Because this approach AIM creates a dynamic and flexible integrated management functionality is designed to keep all locations active, IT cable-once data center infrastructure by associated with SANs. If a network-booted organizations can maximize utilization— coordinating end-to-end network, storage, persona resides within a Dell Compellent enabling even legacy and entry-level systems and computing resources. AIM captures virtual storage array, for example, it can to be leveraged as appropriate. workload images—personas—in central benefit from the Dell Fluid Data™ architecture, storage that include the OS, the optional AIM using Dell Compellent Data Instant Replay™ Efficiently implementing agent software, application software, and software for backup and recovery and Dell disaster recovery storage and networking settings—including Compellent Remote Instant Replay™ software Together, Dell AIM and Dell Compellent either iSCSI or Fibre Channel—and other for long-distance replication. AIM provides storage offer rich complementary features settings for running an application on an open approach that works with many data that form a cost-effective foundation for

Reprinted from Dell Power Solutions, 2011 Issue 4. Copyright ­© 2011 Dell Inc. All rights reserved. dell.com/powersolutions | 2011 Issue 04 73 Virtualization

business continuity planning. At the Dell a simulated disaster recovery scenario using the direct SAN access option in AIM. Customer Briefing Center in Mexico City, comprising two sites that support production Administrators can implement SAN zoning Mexico, and the Dell Solution Center in operation of workloads, replication, and using the World Wide Name (WWN) of the Round Rock, Texas, Dell engineers configured disaster recovery. The main site ran a host bus adapter (HBA) that is assigned to persona using the Red Hat® Enterprise the persona. ® 5 64-bit OS on a Dell PowerEdge At the remote site, the persona was #!/bin/bash R610 server that was booted from a Dell configured to boot using an iSCSI image; as # Compellent Storage Center SAN. The volume a result, there is no need to use the exact # ********************************* # containing the persona was replicated to the same hardware that is used at the main site, # (C) 2011 Dell disaster recovery site in Dell Compellent and for this scenario there was no need to # Remote Instant Replay. Both sites were use Fibre Channel hardware at the remote # ********************************* connected using a Layer 3 network using site. At the main site, the persona runs on a Dell PowerConnect™ 6224 switches. physical PowerEdge R610 server, whereas SDK=/opt/dell/aim/bin/sdk Following a simulated disaster, the at the remote site a server running ESX 4.1 ACCOUNT=admin PASS=admin replicated data for the volumes was is used to boot the persona on an AIM- SCRIPT=$1 promoted to boot volumes on the disaster managed virtual machine. recovery site; AIM then configured the if [ "$SCRIPT” == "" ]; then persona to boot on a VMware ESX 4.1–based Easily testing a disaster recovery process echo "Usage: $0 [script filename]" virtual machine. The configuration at both To simulate a failure at the main site, either exit 1 fi sites consisted of the following components: the server can be powered off or the persona can be stopped from the Dell AIM echo "======" • ESX virtual machine running Dell AIM console. The first step following the disaster echo "DELL AIM SDK wrapper" release 3.4.1 or failure is to get to the data restoration echo "Running script $SCRIPT" • Dell Compellent Series 40 controller point (see Figure 1). Administrators can echo "======" • A 3.5-inch Serial Attached SCSI (SAS) use the Instant Validation and Recovery echo $SDK account=$ACCOUNT drive enclosure Wizard to recover the Dell Compellent array password=$PASS ifile=$SCRIPT • Dell PowerConnect 6224 switch for volume from the replicated data. LAN and WAN access To tackle the RTO issue, administrators Figure 2. Running a disaster recovery script • Brocade Silkworm 300 Fibre Channel can use AIM to easily implement a disaster switch recovery script. For example, a generic command-line interface (CLI) wrapper [root@aim341 src]# cat startDR.txt Enterprise Manager software in the Dell and AIM shell script help simplify the open Compellent Management Suite can be development and execution of AIM CLI start persona id=PR.1.3 used to implement Remote Instant Replay scripting, and a separate text file including save exit in a disaster recovery environment. The AIM-related commands is the only input [root@aim341 src]# ./aimcli.sh disaster recovery scenario was designed necessary for this wrapper (see Figure 2). startDR.txt to meet replication requirements for a Because several tools available to ======comprehensive Remote Instant Replay administrators can be used to implement DELL AIM SDK wrapper configuration. Note: By default, data is similar functionality, AIM can be integrated Running script startDR.txt ======replicated from the source volume to the with several open platforms. The output in lowest storage tier of the destination volume.1 Figure 3 shows execution of the disaster >> open The persona that represents the recovery script. >> start persona id=PR.1.3 workload must be created at both sites; it Although this script represents a simple >> save runs the Red Hat Enterprise Linux 5 64-bit start persona, additional tasks can be included >> exit OS, and Fibre Channel access is configured to restore the operating environment of a [root@aim341 src]#

1 For more information about replication requirements when using Enterprise Manager to replicate volumes between Dell Compellent Figure 3. Outputting disaster recovery script Storage Center arrays, see chapter 9, “Managing replications,” in “Storage Center Enterprise Manager user guide,” Revision B, Dell Inc., execution April 2011, support.dell.com/support/edocs/systems/compl_em/en/em_user.pdf.

74 2011 Issue 04 | dell.com/powersolutions Reprinted from Dell Power Solutions, 2011 Issue 4. Copyright ­© 2011 Dell Inc. All rights reserved. production workload. For example, test and Manager software in the Dell Compellent other components of the Dell VIS portfolio, development environments can be gracefully Management Suite provides an easy, wizard- such as Dell VIS Creator, IT organizations shut down to make room for business- driven approach to replication and disaster can be increasingly responsive to end-user critical applications that require the additional recovery that helps administrators manage needs and help them automate routine compute capacity. In addition, creating, the RTO target. RPO can benefit from processes—enabling IT staff to focus on configuring, and virtually cabling new advanced Dell Compellent storage features innovative projects. networks can be automated before starting such as replication, deduplication, bandwidth hundreds of personas, which can follow simulation, and bandwidth shaping. Authors a specific order to correctly sequence the Together, AIM and Dell Compellent start of dependent workloads. storage help simplify the disaster recovery Fabian Salamanca is a sales engineer in the next- process and optimize RTO and RPO targets. generation computing solutions domain at Dell.

Cost-effectively automating The architecture offers a cost-effective and Javier L. Jiménez is a systems principal engineer disaster recovery efficient foundation for business continuity at Dell focusing on next-generation computing Because RTO is related to the processes in environments that simply cannot allow solutions and the VIS portfolio. required to restore operations after data disruptions in the ongoing availability of has been restored, Dell AIM helps reduce business-critical applications. Learn more this target by automating and validating AIM can also impact data centers well tasks related to resource management, beyond application availability. By providing Dell Advanced Infrastructure Manager: synchronization of virtual and physical a single management interface for the entire dell.com/aim resources, workload assignment, networking infrastructure, AIM helps dramatically reduce Dell Compellent Storage Center: configuration, and storage access. Enterprise IT staff workload. When combined with dell.com/compellent

Reprinted from Dell Power Solutions, 2011 Issue 4. Copyright ­© 2011 Dell Inc. All rights reserved. dell.com/powersolutions | 2011 Issue 04 75