Com­mon over­sights re­gard­ing dis­as­ter re­cov­ery & how to avoid them

A dis­as­ter re­cov­ery plan can be con­sid­ered a suc­cess only if, when im­ple­mented at the time of a dis­as­ter, it en­sures that the busi­ness continues to op­er­ate smoothly writes Sean Den­dle.

Business First - - CONTENTS - by Sean Den­dle

Though it may seem a sim­ple enough re­quire­ment, more of­ten than not dis­as­ter re­cov­ery plans are bugged with some com­mon mis­takes or er­rors which can be eas­ily avoided with a lit­tle bit of com­mon sense and fore­sight.

Be­low are some com­mon non­tech­ni­cal mis­takes re­gard­ing dis­as­ter re­cov­ery plans and also the pos­si­ble so­lu­tions to avoid them.

1 Ac­tive sup­port and in­volve­ment from the se­nior man­age­ment

It is just not enough to have the ap­proval of the top man­age­ment, for the dis­as­ter re­cov­ery plan to be a suc­cess dur­ing im­ple­men­ta­tion; the top man­age­ment should be proac­tively in­volved in the ex­e­cu­tion. Hav­ing their pres­ence phys­i­cally at the time of dis­as­ter up­lifts the morale of all the par­tic­i­pants and works to­wards suc­cess.

2 Em­ployee buy in and sup­port

To en­sure that the dis­as­ter re­cov­ery plan is ex­e­cuted ef­fec­tively it needs the sup­port of the em­ploy­ees. The em­ploy­ees must be con­vinced on the ef­fi­cacy of the plan against the dis­as­ter. It is im­por­tant to in­volve all the de­part­ments of the busi­ness in the dis­as­ter re­cov­ery op­er­a­tions so that they can as­sist in the plan­ning process and thus take own­er­ship of the plan to an ex­tent. It is im­por­tant to have the non-tech­ni­cal team too as part of the dis­as­ter re­cov­ery team.

3 Dis­as­ter re­cov­ery is not a one per­son job

More of­ten than not com­pa­nies des­ig­nate a per­son as the sole in charge of the dis­as­ter re­cov­ery process. This should not be the case be­cause dis­as­ter re­cov­ery ac­tiv­i­ties are time crit­i­cal, hence there should be a group of well-trained in­di­vid­u­als to ex­e­cute the process.

4 Gen­er­al­i­sa­tions vs. Spe­cific tasks

In the dis­as­ter re­cov­ery plans, most of­ten there is a large ex­tent of gen­er­al­i­sa­tions. This should not be the case. Clear spe­cific steps, ac­tiv­i­ties, and pro­cesses need to be thor­oughly doc­u­mented (and tested) to en­sure that mis­sion crit­i­cal data and busi­ness pro­cesses run un­ham­pered.

5 Bud­getary con­straints

Dis­as­ter re­cov­ery plans can only be a suc­cess if the nec­es­sary budget is al­lo­cated to it. Not only mon­e­tary budget, but hu­man re­sources too, should be al­lo­cated for the suc­cess­ful im­ple­men­ta­tion of dis­as­ter re­cov­ery. Key people from dif­fer­ent de­part­ments should be al­lo­cated and their ex­per­tise should be utilised both for plan­ning and ex­e­cu­tion. Spe­cific is­sues in dis­as­ter re­cov­ery to bear in mind by or­gan­i­sa­tions con­sid­er­ing cloud so­lu­tions are:

1 Data crit­i­cal­ity

Or­gan­i­sa­tions do not seg­re­gate data on the ba­sis of its im­por­tance and con­sider it as a large whole which needs to be saved. Hence they go in for cloud back­ups for the whole data pro­duced. This may not nec­es­sar­ily be re­quired. You may find only data which is mis­sion (busi­ness) crit­i­cal re­quires backup to an off­site lo­ca­tion in cloud back­ups for in­stant re­cov­ery.

2 RTO and PRO

Com­pa­nies do not pay much im­por­tance to set­ting the re­cov­ery time ob­jec­tive (RTO) and the re­cov­ery point ob­jec­tive (RPO) for cloud backup. The RTO and RPO are highly im­por­tant to en­sure that the busi­ness does not suf­fer in any way in sit­u­a­tions of out­ages. Or­gan­i­sa­tions should have a real­is­tic RPO and RTO set so that they can man­age bet­ter in times of out­ages.

3 Dif­fer­ence be­tween High Avail­abil­ity (HA) and Dis­as­ter Re­cov­ery (DR)

It is vi­tally im­por­tant for com­pa­nies to un­der­stand the sub­tle dif­fer­ences be­tween High Avail­abil­ity and Dis­as­ter Re­cov­ery. High Avail­abil­ity is the abil­ity or the re­silience of the data cen­tre or lo­ca­tion whereas dis­as­ter re­cov­ery is the data avail­abil­ity across dif­fer­ent data cen­tres or servers at var­i­ous lo­ca­tions.

4 In­ter-de­pen­den­cies of servers

Com­pa­nies usu­ally tier servers but do not pay at­ten­tion to the in­ter-de­pen­den­cies while work­ing on the dis­as­ter re­cov­ery plan. Be­cause of this fo­cus on in­di­vid­ual tech­nolo­gies with­out con­sid­er­a­tion to the in­ter-de­pen­den­cies, there is a high prob­a­bil­ity that the cloud backup plan might not be suf­fi­cient.

For a suc­cess­ful dis­as­ter re­cov­ery plan it should be bring in the or­gan­i­sa­tion as a whole and should be thought of as a nec­es­sary ex­pense and not as some­thing that can be dealt with later on. My fi­nal thought is: a dis­as­ter re­cov­ery plan is only as good as the last time it was tested. How fre­quently do you test your DR plan? How long did it take? Are the re­sults of these tests pro­vided to man­age­ment for re­view?

Newspapers in English

Newspapers from Australia

© PressReader. All rights reserved.