Untrue: Primary Storage Needs to Be Local

04/01/2018
90
Embed

The Big Picture: How do you get local storage performance in cloud storage? In this short video learn how Zadara Storage provides an identical stack sitting in collocation connected to the cloud via private fiber and with data only available to your private network in your cloud. It feels to you like it's local storage within your virtual private cloud in the public cloud, and yet, it's dedicated hardware on demand so. it's local. It's fast with dedicated, on-demand and elastic resources just for your environment.

Notable Quotes: 

Noam Shendar: "...we provide enterprise storage as a service, and we, importantly, provide it anywhere."

Noam Shendar: "Think of it as we bring the benefit of the on-premise's environment into the cloud, and also we take the benefits of the cloud, and bring them on-premises."

Mike Matchett:                  Hi, I'm Mike Matchett with Small World, Big Data, and I'm here today with Noam Shendar, who's the GM of Cloud at Zadara Storage, and he's going to tell us a little bit about what they're up to. Just to set the stage ... Hi, Noam. How you doing?

Noam Shendar:                  I'm good. Hi, Mike.

Mike Matchett:                  Set the stage a little bit. Zadara is doing enterprise storage as a service, where you buy storage as a subscription on premise, and I'll let you explain the idea about ... that's what your first part of the business was, and now you're able to do something like subscribe to hard drives, if I understand this right, and SSDs even, in the cloud, as part of that service.

Noam Shendar:                  That's absolutely right. As you said, we provide enterprise storage as a service, and we, importantly, provide it anywhere. By anywhere, we mean literally anywhere. If you want it in a public cloud, then it's in a public cloud. I'll explain how we do that, and if you want it on premises, we deliver it to your premises, wherever you are in the planet, or we do both at the same time. You can actually mix the on-premise's environment and the in cloud environment, for what's commonly known as a hybrid cloud environment.

                                                      The way we do it in the cloud is that we have the actual same hardware and software and the exact identical stack sitting in collocation that is literally next door to the cloud location, connected to the cloud via private fiber and with a networking arrangement that makes the data only available to your private network in your cloud. It feels to you like it's local storage within your virtual private cloud in the public cloud, and yet, it's dedicated hardware on demand, best of both words. It's local. It's fast. It's very high throughput, and at the same time, it has dedicated resources just for you that are on demand and elastic.

Mike Matchett:                  I just want to be clear about that. If I have an existing traditional hybrid storage array and I'm tired, and it's on-prem and going to get a little tired ... it's time to refresh it. I'm little tired of managing it. I can replace that with a subscription on-premise to your storage, and you'll put a footprint of storage there that I just basically buy utilization out of, as if it was a public cloud, but it's on premises. Then, I can get that same service from you in a public cloud, the same exact storage, right? Am I getting that right?

Noam Shendar:                  That's exactly right. Think of it as we bring the benefit of the on-premise's environment into the cloud, and also we take the benefits of the cloud, and bring them on-premises. Which means that on-premises, we'll send to you at no cost your storage stack, so your storage array for your exclusive use.

                                                      You will use it and you will pay for it based on your actual usage. You won't pay for how much capacity you have, you'll pay for how much capacity you're using. That number can vary up or down based on your actual usage, and you have the ability to even give it back. You can say, "I no longer have a need for this," and we will take it back, scrub it, refurbish it, and then redeploy it to another customer who's interested in this flexible and really powerful model.

Mike Matchett:                  You were telling me a little bit before, you have the ability to put flash in there, and even deliver an all-Flash model if somebody wants it. That's not ... the performance isn't the concern, but it's both block and file depending on what the customer wants. You're talking about personalities changing on demand of the storage array. Tell me a little bit about that.

Noam Shendar:                  Yeah, it's actually block, and file, an object. We sometimes describe ourselves as any, any, any. That first any is anywhere, so that's any public cloud, and/or any private location specified by our customer. The second any is any storage type, so block, and file, and object. The last any, which is any protocol, so we have a really wide array of protocols that we support, from traditional ones like Fibre Channel, and then to modern ones like S3 and iSER, the iSCSI Extensions for RDMA.

                                                      With regard to what you're asking, the middle any, storage type, our architecture's a software defined architecture, but delivered as the full appliance. Again, we try to think of how to give the customers the best of all worlds, not improve one thing, and take a step back in another area.

                                                      In this case we are saying it is software defined storage, which means it's an extremely flexible. You can change the data type from block to file, and back. You can move data non-disruptively among tiers from hard drives to SSD's and back. You can change your cache size up or down non-disruptively. You can change your performance up or down non-disruptively, et cetera, et cetera.

                                                      We call it the, you can't make a mistake, model. Make a decision and then go with it. If you are unhappy with it, or your requirements change, no problem. You can non-disruptively make changes to your configuration to the point where you're changing the personality of the storage. That's the power of this software-defined storage.

                                                      At the same time, we don't make you install the software on hardware, worry about compatibility, worry about firmware updates, all of these things. We do all of these things for you by delivering the hardware and the software together in way that's pre-integrated, proven to work, and we will continue to maintain it moving forward, including refreshing the hardware.

                                                      Again, this is SDS. It doesn't care what the hardware is. We have the power to swap out the hardware from underneath the storage stack without disrupting. We call that immortality. Our storage will keep on working forever, and as we refresh the hardware as needed.

Mike Matchett:                  Got multiple personalities, and it's immortal, and I know you even told me, even on the cloud side, 'cause of the way you do it, the discs belong to each customer. You're not sharing that, so you've got isolation for things like GDPR, and other regulations and utilities, even on the cloud side, which is all great. Where can someone find out more about this?

Noam Shendar:                  Our homepage is zadarastorage.com. You can go, if you want to type less, go to zadara.com. Zadara has all A's. All the vowels are A's. Z-A-D-A-R-A.com. You can go find us there. There's a plethora of information there. We try to make it very real, so we have a lot of use-cases. You can find somebody like you, who's using their storage, and find out why they came to us, what they're doing with us. What choices did they make as far as on-prem, or cloud, or both?

                                                      You can find out about the different features, and you can find out about the many cloud providers with whom we work, who include Amazon web services, Microsoft Azure, Google Cloud Platform, and also regional and vertical cloud providers. We actually have a really, really nice global footprint across five continents, with, if you count our public and private locations, over 200 different locations that we operate at this very moment.

Mike Matchett:                  Sounds like an excellent way to tackle the next generation of hybrid storage in any given enterprise. Thank you Noam, today, for taking the time to talk that through with us. I'm Mike Matchett from Small World, Big Data, and we'll see you later. Thanks.

Noam Shendar:                  See you. Thanks, Mike.

 

Categories:
Channels:
Tags: