DatriumCrowdChat

many customers I speak with start with test dev and VDI, then move to web apps and databases that drive their business

the growing prevalence of Virtual Desktops is a big one, many organizations are rapidly expanding virtualize End User desktops and cant wait around for the traditional storage allocation processes to do this

Most IT folks I talk to are trying to drive IT transformation generally - trying to get away from legacy infrastructure admin issues so they can focus on improving their customer experience

Every consumer is demanding instant gratification from the businesses they interact with. What is fast today, will be slow tomorrow. So, every business needs to adopt to this new instant economy model. To do that, we need infrastructure to deliver instant outcomes.

Data protection and recovery is one - human error, ransomware can happen. How can you make it a non-event?

Increasing amounts of distributed data in public clouds, on premises, in SaaS clouds, and at the Edge.

Simplicity. Want something that is simple and easy with low to no work. I want to be able to sleep at night and know that my data is protected

and then ransomware! A platform that can instantly recover an entire infrastructure to a known good state is invaluable

here's a recent report that talks about the status of IT transformation https://cdna.datrium.com/wp-content/uploads/2019/0...

Test/dev and the speed of it is another one. How long should a developer wait before the production data can be made available to them for their use. Should be instant.

@BrianBiles you hit the nail on the head

Another use case is improving workload performance. Nothing is fast enough. As Bezos would say, a customer is never going to say I don't want it faster. So, how fast can it go?

my users need more, faster and more often, by simplifying and automating we are actually seeing LESS total demand for systems (users are willing to give up old systems if it is quick and easy to get new ones)

Data is everywhere - (multi) cloud is forcing users to rethink everything

The chief use case driving data infrastructure simplification is 360-degree customer view for multichannel target marketing, personalization, and experience management. That calls for convergence of OLTP, OLAP, DW, and machine learning within common end-to-end infra.

@tushar_agr our CAD team love the performance of not having to wait to render anymore

Lastly how about maintenance and optimizations. How can you shift your workloads to another site reliably, do what you need to do and bring them back. And, only pay for the other site for the duration of the event.

all our IT is in the cloud - so everything from test/dev, web, transactional data, analytics, video editing...everything

Now THAT'S a great use case? How is this translating into business outcomes?

Firstly, there is a lot of risk weaving together a bunch of different products together. Modern tech is supposed to simplify and reduce the number of things we need weave together.

single pane of glass to manage process, automation, auditing and operations. data is everywhere and people want it now, fast and reliable.

we see customers struggle to work across 5 data functions. Here is a blog that explains the problem and how you can solve it https://www.datrium.com/blog/the-future-of-it-auto...

lack of true portability and transparency

the problem is that everything IT shops need to do requires a separate product with it's own recurring costs, training and maintenance requirements. What is needed are consolidated platforms that treat all aspects of data as one framework, @Datrium is very unique here

The worst is DR orchestration - there can be multiple vendors involved, with different software releases, management consoles, etc. - so many deployers find it fails when they test it. It's a whole-datacenter problem.

Significant friction of different services for different data in different locations. Usual choice is an integrated data management & to move the data to one place. This is costly and slow. Time value and context is lost in transmission. A hybrid cloud approach is best.

the biggest friction of multiple data silos is time. takes too long to backup and restore. takes too long to use cloud. takes too long to recover from disaster. while my customer demands outcome now

@sazzala We've learned the risk of "weaving" tech together in multiple domains. Think of how complex analytics tool chains are. Baking weak interfaces into data services processes will kill performance and security.

must not be using the right tools, we moved away from things like SRM years ago for that reason

The big friction going forward will be that your data will become land-locked. The many different products managing your data are going to be drag on mobility. We all want to leverage the multi-cloud, but the friction between products is going to an albatross.

many complain about "API creep" and the increasing complexity of the data pipeline...it's often complicated, confusing and expensive to manage...

many software development teams like Datrium can reduce friction with fast cloning; here's how https://www.datrium.com/blog/with-fast-clone-tech-...

the second biggest friction is innovation. i can no longer access all my data. my primary has some piece of it. my secondary some piece of it. cloud has other pieces. i can't figure out what is where and how to use it.

And, security and data integrity checks across these multiple silos is hard to establish

A chief source of friction when weaving together multiple data services is usually the need to harmonize all the schemas, aggregate and cleanse the data, and converge it all in a data lake for unified DataOps. Addressing this involves investment in data consolidation.

VM level simplicity, with always-on data reduction, and encryption.

keeping the data in flash on the host always helps too

Getting high performance IOPS without simplicity slows down everyone

definitely vm level management, from inside vCenter

our customers are leveraging this for high performance Virtual Desktops and to have a simple and effective Replication/DR solution

IOPS is one thing, but the most important thing is latency that your applications feel. That is the true test.

Note also with DVX, split provisioning automatically takes full advantage of newer server hardware as its performance layer, separated from persistence

Storage efficiency services, such as data dedup, are always essential. Likewise, data protection services--such as backup and recovery--are a keen perennial enterprise requirement.

The ideal services such as orchestration or backup allow you to keep the data where it should b, and offer the same services in every node of a hybrid cloud. It must be implemented as software running in every node in the hybrid cloud.

Locating data needed for reporting or recovery is essential as is Automated Compliance Reporting

And by the way, everyone desires performance in a cost efficient way.

it is interesting to me that for so many storage vendors dedupe is still an after thought, it is the first thing I look for, my previous provider is just now adding it to their platforms after 8 years

lots of focus on data analytics with an underlying infrastructure to support real-time or near real-time insights.

so many IT shops are being asked to take on more responsibilities with less people, @Datrium allows them to let go of repetitive tasks around storage and focus more on the apps and business needs

Common integrated services which can manage the data where it is.

The question here is what does 10X lower latency mean for your end-customers? What do zero backup windows mean your IT team? And what does day 1 DR readiness mean for business. Lastly, what does it mean to use public cloud for DR in less than an hour.

@tushar_agr I will say, that I love being able to restore a VM in the time it takes to do a reboot of the VM, the first time I did it, I thought it did not work, it was just too quick

Query optimization is of course essential for extracting maximum decision support potential from data, so enterprises insist on in-memory, automated data placement, etc.

@VMJedi totally agree on this, we always laugh when our customers do this for the first time and are confused that a restored VM is instantly available, years of experience create the habit of expecting at least some sort of restore time