ob­ject Stor­age Helps Cre­ate Data- driven Busi­ness Mod­els: Joan Wra­betz, HgSt

HGST, Inc. (for­merly Hi­tachi Global Stor­age Tech­nolo­gies), a wholly owned sub­sidiary of Western Dig­i­tal, deals with all-flash ar­rays, ob­ject stor­age, and hy­brid cloud stor­age so­lu­tions. HGST helps en­ter­prises with stor­age sys­tems that sup­ports data ar­chit

Dataquest - - CONTENTS -

—JOAN WRA­BETZ, VP Mar­ket­ing, Dat­a­cen­ter Sys­tems Busi­ness Unit, HGST

Where do stor­age tech­nolo­gies fea­ture in the world of dig­i­tal trans­for­ma­tion? There is no ar­gu­ment that we live in a data-driven world spurred on by dig­i­tal trans­for­ma­tion. The chal­lenges in dig­i­tal trans­for­ma­tion are four­fold: data mon­e­ti­za­tion, data ar­chi­tec­ture strat­egy, best prac­tices in man­ag­ing data, and man­ag­ing data in the cloud. Note that all the chal­lenges of dig­i­tal trans­for­ma­tion have a di­rect im­pli­ca­tion on data. This con­nects the dots be­tween dig­i­tal trans­for­ma­tion and data.

Now let me bring in stor­age tech­nol­ogy and con­nect it to data. The value of data comes in when you can mon­e­tize it ei­ther di­rectly or through the in­sights gleaned from the data. The right kind of stor­age de­ter­mines the avail­abil­ity of data. The data that the or­ga­ni­za­tion al­ready pos­sesses is ei­ther on tape and hence can­not be ac­cessed read­ily or that the data is in si­los or that the data can­not be shared.

What kind of a stor­age so­lu­tion would help with data?

The so­lu­tion to this is to build a large-scale stor­age sys­tem, specif­i­cally a petabyte-scale ob­ject stor­age with cloud-na­tive S3 in­ter­face, that lends it­self to ex­treme dura­bil­ity. Such a sys­tem helps cre­ate data-cen­tric busi­ness mod­els.

Such as stor­age sys­tem is based on ob­ject stor­age. Legacy stor­age is de­signed for trans­ac­tional pur­poses (fre­quent read-writes), while ob­ject stor­age is de­signed for archival; rare writes, very fre­quent reads.

Are or­ga­ni­za­tions able to ap­proach this in the right way?

Man­age­ment is tuned to to­day’s stor­age, they have big in­vest­ments in ex­ist­ing stor­age and man­ag­ing the tech­nol­ogy life­cy­cle of to­day’s stor­age prod­ucts. This has huge im­pli­ca­tions on the data ar­chi­tec­ture strat­egy. What is needed is a change in the sta­tus quo, a new ap­proach.

What is the new ap­proach?

This new ap­proach is pred­i­cated on the idea that data ar­chi­tec­ture strat­egy is dif­fer­ent from stor­age ar­chi­tec­ture strat­egy. The sug­gested data ar­chi­tec­ture strat­egy should be amenable to im­ple­ment­ing cloud scale op­er­a­tions that is run on a sim­ple ar­chi­tec­ture that yields low TCO.

The ben­e­fits of such a data ar­chi­tec­ture in­clude ac­cess to un­struc­tured data with scal­a­bil­ity, com­bine file and ob­ject stor­age, and move away from si­los with ca­pa­bil­i­ties like an­a­lyt­ics and backup.

Where does cloud fea­ture in the con­ver­sa­tion about data and stor­age?

Flex­i­bil­ity with cloud is im­por­tant. A hy­brid cloud is in­her­ently flex­i­ble; pri­vate cloud can be used for sen­si­tive and ac­tive data and the pub­lic cloud can be used for data pro­cess­ing. In many new ap­pli­ca­tions, data will live its en­tire life­span in the cloud. Data is cre­ated, man­aged, an­a­lyzed, acted upon and archived there. Cloud providers keep up with the grow­ing de­mand eco­nom­i­cally and cloud stor­age ser­vices of­fer high avail­abil­ity and con­sis­tent I/O per­for­mance.

How does this get used across ver­ti­cals? Are there speci­fici­ties?

Each ver­ti­cal has its own use cases. Me­dia and en­ter­tain­ment in­dus­try has the need for ar­chiv­ing, CDN, distri­bu­tion, and ren­der­ing work­flows. HGST so­lu­tions en­able pro­duc­tion, broad­cast, and distri­bu­tion com­pa­nies to work more ef­fi­ciently by de­liv­er­ing fast, flex­i­ble, and scal­able stor­age for every stage of the work­flow. It en­ables the M&E mar­ket to cre­ate, lever­age and pre­serve data while im­prov­ing pro­duc­tiv­ity and low­er­ing TCO for 2K, 4K, 6K, HFR and UHD work­flows.

In life sci­ences, we en­able ge­nomic se­quenc­ing, med­i­cal imaging, and bi­o­log­i­cal re­search labs to de­liver fast, durable and scal­able data for every point in the data anal­y­sis work­flow us­ing tech­nolo­gies from NVMe solid state drives, rack scale flash plat­forms to ob­ject stor­age and hy­brid cloud stor­age so­lu­tions.

What would be some straight points of ad­vice for CIOs on data and stor­age strat­egy in the era of dig­i­tal trans­for­ma­tion?

• Fo­cus on the eco­nom­ics of data, not stor­age. • As­sume that stor­age costs will go down, the value of data will go up. • Min­i­mize data in legacy sys­tems. • Build cloud-scale data ar­chi­tec­tures and move to cloud na­tive ap­pli­ca­tions. • Sep­a­rate ap­pli­ca­tions from data. • As­sume that data and ap­pli­ca­tions will move across clouds. • Build na­tive sup­port for an­a­lyt­ics. • Drive TCO down with scale.

—JOAN WRA­BETZ VP Mar­ket­ing, Dat­a­cen­ter Sys­tems Busi­ness Unit, HGST

Newspapers in English

Newspapers from India

© PressReader. All rights reserved.