Advertisment

Disaster Recovery and Business Continuity Planning for Enterprises

author-image
PCQ Bureau
New Update

According to our last month's survey of mid to large Indian enterprises on tech challenges and opportu-nities, we found that DR and BCP was on top of the deployment chart for most mid-sized enterprises. This is aclear indication that even mid-sized enterprises are extrem-ley sensitive about protecting their information so that their business doesn't come to a stand still. They possibly under-stand the implications of application downtime for their business.

Advertisment

The terms DR and BCP come together as the former is just a subset of the other.It is just not enough to recover from disas-ter, but also to swiftly make your organiza-tion get back into business. However, onepoint of differentiation between the twoterms is that while the former comprisesof a technology strategy, the latter entailsmore of a business plan.

Why you should considerDR and BCP?

Advertisment

Until recently, there have been just tworeasons for having a DR and BCP strategyin place. One, readiness during naturalcalamities and two, human interventions.A third reason that has been observedmore recently is the constant exposure tosecurity breaches and attacks. In a recent DR survey doneby Symantec, the most important factor driving the need forDR adoption is virtual or physical security breaches. There is a constant fear of losing critical data among enterprises.Computer system failures, virus attacks, changes in tech-nology infrastructure, pressure from customers and com-petitors and compliance factors are some other prominentreasons why most enterprises are looking for DR solutionsand BCP strategies.

In India, as for the rest of the world, virtualization is akey reason that's enforcing a majority of mid-sized enter-prises to reconsider their existingDR plan. This can be clearly attrib-uted to the challenges that a virtualenvironment poses becauseprocesses that apply to physical en-vironments may not necessarily beapplicable to virtual environments.Though imagining a disaster ina completely up and running busi-ness environment in normal day-to-day life is difficult but the factorsstated above can anytime become amajor concern for IT managementheads and CIOs. So, it is of utmostimportance to have a DR solutionand BCP strategy in place.

DR BCP Strategy andDR System Tests

Advertisment

We have been covering on DR and BCP strategies thatenterprises should formulate and what are the best prac-tices to formulate these strategies. A quick re-capsulation for the uninitiated--a DR plan has to revolve around its fourbasic pillars of data, equipment, alternate site locations andpeople. These factors have to amalgamate well with an or-ganization's business continuity plan comprising of somekey metrics of recovery point objective (RPO) and recoverytime objective (RTO).

As per one of the surveys by Symantec India, 22% of In-dian enterprises conduct full scenario DR tests only once ayear or less because of perceived fear of business disruptionand lack of resources to conducttests. Reasons cited include: lack ofstaff availability (56%), disruption toemployees (58%), budgetary issues(44%) and disruption to customers(46%). In addition, 32% admit DR test-ing could affect sales and revenue

Advertisment

Besides, planning and imple-mentation, another very importantaspect of the DR-BCP plan is thetime-to-time testing of the efficiencyof the DR solution in place. What'sthe benefit of deploying the best DRsolution if it does not stand up to ex-pectations at the time of disaster justbecause of lack of regular test drillschecking the efficiency of the solu-tion.

Sanovi Disaster Recovery Management Suite

Considering the fact that DR and BCP planning is on top ofthe minds of a majority of IT decision makers, we reviewedone of the latest DR suites available in the market today--Sanovi Disaster Recovery Management Suite.Sanovi DR management software for system applica-tions and databases that helps in DR infrastructure monitor-ing, testing and automated recovery of business applica-tions to meet service level objectives. Built on astandards-based platform that leverages your existing data-base replication and storage subsystem replication solu-tions, it offers real-time monitoring, testing, reporting,events, policies and automated recovery features needed forend-to-end visibility and confidence in your DR infrastruc-ture. Sanovi DRM overlays on top of existing recovery sys-tems and interfaces with recoverycomponents.

Advertisment

Sanovi DRM File Replication Method

Sanovi DRM uses its replication softwarecalled Panaces File Replicator (PFR)which comes embedded with the DRMsoftware. PFR runs as a service on theOS, so we need to start the Panace FileReplicator Services from the WindowServices.

Advertisment

The replicator allows an administra-tor to distribute content and data across servers and different platforms across dif-ferent locations (Image 1).The process iscalled Normal Copy Execution which is arecurring process and terminates during the time of disaster.

Why Sanovi DRM Suite?

Advertisment

The entire basis of the Sanovi DRM revolves around twomain objectives: RPO (Recovery Point Objective) and RTO(Recovery Time Objective). RPO is an acceptable amount ofdata loss from the last good backup prior to the point of fail-ure while RTO is the time taken for the business to be backonline after a disaster i.e. the time taken to get back to Busi-ness Continuity. Both these parameters are set by the solu-tion provider as per the demand of the client and the criti-cality of application. Thus, it makes a lot of sense for enter-prises to pay just exactly for the desired configuration,policy-based corrective responses and the desired level ofcriticality of objectives than to go for one standardized fixed price DRM suite.

As discussed earlier in the article, time to time testing ofa DR and BCP solution is as important as choosing a DR so-lution. Sanovi DRM suite offers ease of performance of op-erations of 'switchback' and 'switchover' that are actuallytest drills to put DR site in active mode and PR site in standbyand vice versa, respectively. This is possible at a click of abutton in the application UI. A similar feature is available formaking DR active incase of an actual disaster.

The dashboard view of the application UI offers a com-plete one shot overview of the entire network, hardwarecomponents and application processes that can get effectedduring a disaster. This enables real time monitoring of yourorganization's complete IT health. Anytime downtime in anyof the components is alerted and maintained real time asevent logs in the application.

How we tested?

We created a typical disaster recovery management envi-ronment, where we have a primary database (PR site) and aphysical standby database (DR site) which becomes activein the event of a disaster. The two sites share a common net-work.

The test bed comprised of 3 machines on our labs net-work: The Machine1 (Sanovi DRM Server Site), Machine 2(Production Site) and Machine 3 (DR Site). Windows Server2003 Enterprise Edition was installed on machine 2 and ma-chine 3. Oracle 9i database was installed on both machines.On machine 2, the Oracle database was in Read/Write modeby default. On the other, it was configured on Standby mode.Machine 1 was a Dell R810 server with 750 GB (5x150GB)HDD of which 2 HDD were configured in RAID 1, which wasa pre-requisite for the installation of Sanovi DRM Server.Red Hat Enterprise Linux 5.5 was installed on this machinefollowed by Sanovi DRM Server installation. Sanovi Agentswere installed on the PR and DR machines which can beused to monitor, troubleshoot and perform actions likeswitchover (test drill to make DR site active), switchback(an action to roll back 'switchover' action making PR site ac-tive again), failover (action to make DR site active in a real-time disaster scenario), etc. We created two sites Delhi (PR)and Harayana (DR) on machine 2 and machine 3 respectively. The creation of sites, detecting network components,OS, etc can be done by logging on to the Web-based SanoviDRM client at:

http://:8080/PanacesGUI/

Next we opened the 'Discovery' tab to set up our net-work, and performed the following tests on the Sanovi DRMsoftware:

Test 1: Monitoring/Troubleshooting Test To test whether the environment has been configured prop-erly, we performed 2 monitoring/troubleshooting tests.First, we stopped the listener services of Oracle (Image2).



The Relationship tab in Monitor window of the SanoviDRM client showed status of DR site to be 'inactive' (Image3). The status refers to the non-working condition of site.

Then, we stopped DR Panaces Server from WindowsServices. The status became 'Unknown' (Image 4). This sta-tus is displayed when any of the Sanovi services are down.

Test 2: Switchback/ Switchover

We tested features of switchback/switchover which are testdrills that an enterprise can perform from time to time tocheck the DRM solution working in place.

To launch switchover, we opened 'Test' tab of the SanoviDRM Client, and selected 'Test List View' and clicked on'Start Test' against 'Switchover' (Image 5). The NormalCopy execution has to be manually stopped before this drill.A workflow got activated (visible in 'Workflows Action' tab)which ended by making DR site active changing its status from standby to read/write mode. During this process,archive logs and redo logs from the DR site get populated to those of the PR site.

'Swiitchback' operation can be similarly launched and itreverts the above action (Image 6).

Test 3: Making DR server active by making PR server un-available

To test the DRM solution, under a real time disaster sce-nario, we disabled the network card of the PR site, in a way,making the production server unavailable. DRM client'Monitor' tab displayed 'Inactive' status. In the event log, theissue with timestamp and severity got registered and the sta-tus of the functional group changed to 'Degraded' (Image 7).

The Normal Copy execution was terminated manuallyand to make DR server active, the 'Initiate Failover' waslaunched in the Manage tab (Image 8).

The process runs a workflow (Image 9) which finallymakes the DR site active and the database gets intoRead/Write mode to accept new records in the database.

Sanovi DRM at what cost?

Sanovi DRM Enterprise edition is available in the followinglicense packages options: Sanovi DRM Recovery Monitorand Manager, Sanovi DRM Drill Manager and Sanovi DRMAdvanced Reporter. Pricing of the enterprise edition is basedon the number of servers, number of database instances andnumber of storage based replication instances. Customiza-tion of Sanovi DRM to meet customer business needs is of-fered on a time & material basis. Sanovi also offers an EasyDR package where customers with less than 3 applicationscan deploy a complete DR solution that offers monitoring ofRPO, automation of failover recovery and includes Sanovi'sreplication software for under Rs 5 lakhs.

Go ahead tomake your enterprise disaster proof and BCPready!

Related Articles:

Implementing DR and BCP

Advertisment