datacontrol

a single one of glass view

I imagine a universally adopted data storage API which most developers are familiar with would be very helpful across clouds

Being able to search data over multiple cloud not just isolated islands.

All of the above :-)

All of the above. Like @Tyna_callahan said, orgs want a single pane of glass for simplified governance.

sorry -- pane, not one

Intelligent data placement according to the compute they need. Based on data associated metadata may be?

@bennettb83 What would be your vote for this universally adopted data storage API?

API's are the future of storage and applications. Orgs will also want an easy to use API that is powerful enough for governance.

data interoperability is so huge and containers kubernetes driving #microservices will pace this trend imo

Control APIs for all applications to be able to leverage data in the cloud

data is the asset and so controlling it is key - access and security management, location control, data search at least for metadata across clouds

"There's an API for that"......

@Tyna_callahan I like the idea of a gateway data access API. No more repeating lines of codes for different SDKs... Same for centralized authentication!

Data access, high speed data movement, data placement, copy services, data protection, many...

@jlecat S3 has been working for us :-)

Currently searching for existing data on #aws and other platforms is like finding a needle in a haystack, forget about unified search across clouds today but we're working on it :)

Single-pane, intelligent data placement, common data protection tools, support for Hybrid (Cloud/On-premises), "transparency" for the IT Pro

too big of a question. If I had to select one I'd say Data Access API. Once accessing data across clouds is normalized, the remaining data services become possible.

workflow based data movement - for disaster recovery/avoidance for certain applications..

None of the above. The most important aspect is having the same architecture when data spans multiple clouds. If the task is development, use the best development cloud, and move the code at the end...

@dvellante I like the "high speed data movement", we are going to need new wavelengths on all the fiber in the ground.... We are still far from having enough in and out throughput betweeen cloud and the real world...

... if the task is global file management, use the same architecture in every cloud (e.g., Avere Global File System)

@dfloyer More in-line with my thoughts. I hate to use the term Cloud File System, but the idea translates. Once that is normalized the pieces fall together.

Is there a standard set of data services that the industry is getting behind? That would be my question to this

to me data services is a fragmented by cloud

@furrier which is why we need to start at the access for format level. Add on services such as meta data search follow.

@Giorgioregni Search across multiple clouds will help bridge the gap between clouds. Having that ability in a single pane of glass would be ideal for the multi-cloud world.

Having a unified, resilient API is key to managing multiple clouds

... The problem with high-speed data movement is that modern application systems have going from milliseconds to microseconds. High speed is only over meters, not kilometers. Cost and elapsed time for data movement reduces value of applications.

yes for sure - the risk of depending on a single vendor for any service is a risk

historically about 15% of customers are concerned about the lock-in issue...much higher w/ cloud

Businesses are worrying about that, but not obsessively. As multi-cloud scenarios become commonplace, cloud service providers are pursuing open microservice & data persistence frameworks to support cross-cloud data reuse.

yes, businesses are starting to worry about cloud data lock-in, nobody likes their data to be taken hostage

Lockin imo is the data layer but opensource creates great opportunities for choice #multicloud is huge oppty

starting -- it's been a concern for awhile. Big issue

S3 compatibility has made movement relatively simple. Cost of data movement/access is a powerful concern.

I've validated from many on @theCUBE #multicloud is real but not yet ready for primetime as #hybridcloud is currently the reality for most CxOs

You want to talk about Lock-in.... I can show you our SFDC bill, increasing every year... and there is nothing we can do about it... just pay!

can anyone say monopolys are good - they've never been a benefit to technology

Using the right cloud for the right job should minimize impact of lock-in. Cannot get rid of some lock-in.

@jameskobielus really important not to get stuck in one place -- cloud or not -- for resiliency sake.

@jlecat SFDC, AWS, $NOW...all up and to the right on the expense line

many confuse lockin with differentiation #multicloud is oppty to break lockin #microservices #containers

@furrier yes, it's hard to move existing data, the concern is to be forced to store new data with the same cloud vendor

@CTOAdvisor yes we need to have independant interfaces across clouds - S3 helps

@GiorgioRegni Have you seen any instances where a cloud provider deliberately prevented a customer from reusing, synchronization, or replicating their data to a competing cloud service?

why move data around if you can move #microservices and #compute to the #data

Yes, organizations are beginning to research and adopt multi-cloud strategies to avoid vendor lock-ins. They want the freedom to manage their data according to business needs and switch between clouds.

@dvellante i'm surprised it's only 15%

i think it's a matter of interoperability. Having the ability to choose is highly important

@ScalityP I thought that Petee Thiel stated that Monopolies are essential to technology creation in zero to one ?

Principle of Multi-cloud: 1/Place applications where the data is generated 2/ Move code to data 3/ Keep cloud architecture the same for application movement e.g., keep same architecture if recovering true private cloud to a public cloud

the dependence on aws was very evident when they had their big outage a while ago - that got businesses to rethink

@furrier that's a big question. Moving compute isn't as easy as it used to be. As cloud IaaS starts to bleed over to PaaS, moving those services/micro service closer to the data becomes non-trivial.

@jlecat Is that a larger issue than just data lock-in? Customers make a large investment in integrating their apps with SaaS providers such as SFDC. It's a deep commitment to run their entire CRM or other business processes in that cloud.

yes, that aws outage was scary, and made a big wave!

@furrier there's a need to virtualize the data and services for multi cloud use. A complex undertaking as the compute consumes proprietary cloud services.

@jameskobielus agreed with you, it is beyond data lock-in, it is really data-model lock-in.... nonetheless, any cloud de-facto monopoly is likely to create very high bills for customers