From EMC2 to Dell EMC .Tale of Enterprise storage in DC

Recent times there has been quite a push for implementing software defined everything solution in the enterprise .Enterprise storage never used to have any kind of threats in such order of magnitude.Now,almost all companies will ;at least have a look at the SDS solution available before deciding on their storage solutions.So, what changed and what results in the transformation of the storage king who acquires each and every known storage start ends up being acquired and finally become Dell EMC.

Story begins here…

Problems with enterprise traditional storage was very well known.First its costly and second its damn costly .But given the options and the Enterprise infrastructure design model which we fondly call as Pet model(Ref :Pet vs Cattle design approach).These design approaches put a very high reliability factor on storage ,because if storage fails everything fails along with your data .And for a web service company data availability is everything .

This approach made the companies flocking towards a reliable storage provider .Even at that time opensource options like FreeNAS etc where available and even much stable and reliable, but the Image which EMC had was hard to break and Enterprise trusted them and in return the support which EMC gave back was top notch.You will only come to know of a drive failure or a controller failure when you get a call from EMC representative asking us to get a gate pass inside Data center.

All is well

This marriage went really well.At one point no one thought there is no going back for companies like EMC and its going to be an evergreen business area.Even the competitor like Netapp made a lot of money by doing the same.

EMC and Netapp released versions after versions of their storage products and EMC World was celebrated every year across the world ;all sunshines and rainbows.

But the issue still remained

1. Costly
2. Support is again costly
3. Takes up a lot of rack space
4. Scale UP architecture.
5. Power consumption is very high

Then out of thin air problems started coming to haunt there Storage titans.

Problems,Problems every where

First: AWS-Cloud

This one was huge ,in the beginning of 2010 companies started seriously looking towards AWS offering and there are many startups which built a successful business model on top of that.Netflix/Drop-box etc was thriving on top of AWS .All of a sudden on-premise DC become legacy and traditional and who like to be traditional .There were talks about moving to cloud every where.Every company has some kind of strategy or road map to move to cloud.I ,still remember communicating with many storage admins in the public forums their response was .”Cloud will not do anything to EMC .Let the application be anywhere and even AWS need storage where they will go other than reliable storage provider EMC “.These storage admins where defiant and too busy to look at the sand eroding under their feet.

Later they understood AWS dont need heavy VMAX machines to host public cloud services.And thing can be done in a very different way.

Second:SSD

Now there has been a saying in the business circles that each technology change will provide opportunity for another player to enter the market.It doesn't really matter even EMC-Netapp together hold more that 60% of the market share.And as it happens there is an endless stream of SSD based storage companies coming into the market .Their sales pitch was simple .”We don’t carry the legacy architecture based in HDD.Our architecture is well suited for SSD and we are pure play SSD vendors .And,by the way SSD give far more IOPS than HDD (if not considering the cost factor).

Now considering buying a VMAX and an SSD startup hardware is almost relatively same price and infact we get better performance from the startup

As it happened startup like Pure storage etc flourished and removed EMC from many storage categories in even gartner ratings. EMC/netapp tried to fend off some threats by acquiring some startup like Solid Fire/Extream IO and there is always innovators dilemma for Big companies when engaged in a sales battle with startup. EMC can never state their Extreamio product performance is far more than VMAX even if it is true ,because that will jeopardize their biggest cash cow(VMAX) which no company wanted. Netapp tried to avert the loss by trying to state the they were an SDS company from day one.

Third :Pets vs Cattle design approach

Soon there has been a realization in the industry that treating the servers as their own precious one and throwing money at it to keep them alive at any cost doesn't make much sense at all.Same time Google published their papers and every one started realizing that treating hardware as an entity that can fail at any time and planing your design based on commodity servers makes more sense for them .This realization and other upcoming tools like VMware vSAN(ironically part of EMC) gave confidence to customers .In order to have a discipline with the cattle we definitely need automation and it is almost impossible to take care of all the logistics of installing upgrading and maintaining of hundreds of server manually and this leads to automation revolution inside DC.

With automation comes approach to treat all resources as software only function to make automation across infrastructure more useful and effective .This one cements the status of SDS in enterprise datacenter and saw multiple SDS solution's like Rehat Ceph,Nexanta and EMC Scaleio coming into the picture .

In the coming days it is inevitable that the on premise DC should be automation friendly and for that commodity and software defined every thing based approach fill all the needs.