FalconStor Building a Managed Service Provider

17:08 · Watch on YouTube ↗

Transcript 2,806 words · about 19 min to read

Auto-generated captions from YouTube, not hand-corrected, so names and technical terms may be imperfect. The video is authoritative.

(bouncy music) (engine starting) >> Hey, Keith Townsend, principal of the CTO advisor. If you didn't follow the first video first off on, if you're watching on YouTube, the link will be here. I got to always remember the left-hand, the link will be here and follow that conversation. I'm joined with Chris Cummings of Falconstor in this sponsored video. >> Thanks for having me Keith. >> Thanks for having me again. Thanks for buying me the cold adult beverage on a very warm, >> Cheers.

Silicon Valley day. You're on the road trip and we're talking about 20 years of history, as long as I've been in IT, especially in a role where I'm responsible for enterprise data centers and enterprise infrastructure. The problem of continuous data protection, the problem of replication of data off prem for disaster recovery has not gone away. >> Yeah. >> Matter of fact, it's getting more complicated. >> That's right. >> because I have my AS 400. I have my IBM powered system.

I have my Itanium systems. I have all of my new VMware infrastructure. I have my cloud native stuff, and I now have all of that plus plus public cloud. If I'm a managed service provider, I'm looking at this and I'm scratching my head. There's opportunity there, but I'm not quite sure how to capture that opportunity. That's what this conversation is about then. >> You bet. >> So tell me, what's the elevator pitch for a managed service provider talking to Falconstor.

>> So we know, the core thing is this, a managed service provider does not get to choose what's on-premise at the customer set. >> No. >> And we know that an individual customer, it's going to be a very diverse environment and now multiply that by however many customers they have. So, no two are alike. How do you deal with that level of diversity and the complexity that's associated with that and still build a scalable business that frankly, for an MSP, where you can make a profit.

And that's where we come into play, which is 20 years of history. Sometimes that's a problem. You know, it was viewed as a problem. In this case, 20 years of history is a benefit. You should see what the certification list looks like between all the backup apps, all the servers, all the storage. And now increasingly the cloud services. That roster is this long. So we guarantee that you're going to find our ability to handle the diversity that an MSP sees every day is unparalleled.

>> So MSPs, again, see a lot of stuff. They have diverse customers and the security requirements are ever burdened. We'll get into, you know, the idea of using immutable infrastructure to protect against ransomware and all of that. But before we get there, let's talk about the practical innards of MSPs challenge. I need to take a solution that's nimble enough to support multiple customers, but extensible enough, from a multitenancy perspective to support multiple customers. Those two things don't always melt together.

>> That's right. >> I need to be able to say, you know what, this controlled domain and this blast area is limited to this customer. And maybe even some boundaries within that customer against other customers. How do you guys support that core capability of MSP? >> Yeah. >> No, that's absolutely key. So sometimes in the past we've seen, you know, replication from a single target on a customer site to a cordoned off target in an MSP. That sounds like a headache to me.

>> Yes, it is a headache. >> So that's where we were 10 years ago. We built a full multi-tenant architecture around this so that all these customers can be run on a common MSP cloud. Now, in this particular case, some of our MSPs are delivering their own cloud. They build their own cloud. It's their data center network. And we have other MSPs who say, I'm not building that myself. I'm going to put that on Azure. I'm going to put that on IBM cloud, AWS, whatever, it's their choice, but you have one console where you're managing all of these individual tenants.

We manage the SLAs against those. So you set up those SLAs and say, I want this kind of RPO. I want this kind of RTO. I want certain of this data to be replicated to a cloud. Certain of it, I may still want to write to tape. Well, we do that too. And then ultimately we provide that information into the billing system so that they can go about the business side of delivering this. >> So let's delve a little bit into this concept of multitenancy and multi SLA.

So, one of the problems a managed service provider has is that they have different varying client requirements. I can have a client with two servers, or I can have a client with a hundred or more for a reason. Within that hundred or more servers, different SLAs. So when I'm building my cloud or I'm provisioning my Azure, AWS, GCP environment, I need to keep that architecture in mind. So you're telling me that Falconstor has the knobs that I can turn to deliver SLAs from, you know, the company that has two servers and a 15 minute RTO RPO, up to the company that has a hundred servers and a variation of RTOs and RPOs.

>> That's right. I mean, the idea is that you aren't just setting those SLAs at a tenant level, right? You're setting them at a subtenant level. And that subtenant is here's a company and this tenant is an application or a division or a group, whatever it may be, that's where you set that. And our system is the one that's showing you, the typology of this is going here and this is going where, so you have one interface for that. And that actually interfaces with some of the monitoring systems that they're using to keep track of all the pieces and parts that are out there on prem.

>> So, talking about pieces and parts that's out there on prem, I have SAP ECC >> Sure. >> and we're in a regulated environment and I'm required to off premise some of the datasets for backup, and that's going to tape. So from an MSP perspective, you're saying, telling me that MSPs can offer that capability and the cool disk to disk-based backup and data protection that we expect today. >> Right. So if in this case that you point out, if it's a big enough environment, that it warrants having a small appliance on premise, put that there.

If the tape operation had been on-premise, now you're moving that tape offload operation back to the MSP cloud. So you're taking it off the secondary copy. You write to tape, or you take that and you're writing to, in this case, say, immutable disk, maybe you don't need that to go to tape. I'm going to write it to immutable disk. Then I've got the kind of locking capability around that data for long-term retention. Because once it typically goes into an MSP cloud, it's not coming out.

>> All right. So let's say that I decide to build my own cloud and I'm a Dell shop. >> Yeah. How does being a Dell shop hinder or help me when it comes to selecting Falconstor? >> You have full license to build your cloud with whatever you want. So, because we are a software-only shop and are heterogeneous, as a consequence, you can choose the Dell servers you want, you can choose the Dell storage you may want. And if over time, you don't like the deal you're getting and you see a different deal that you want from someone else, you can use that too.

So, you have the ability to choose what you want, which is really the value prop of software defined. >> So let's talk about immutable infrastructure. Let's talk about a couple of different use cases for it. One is E-discovery. Another one is what everyone is talking about now, malware, ransomware and version. So I have at the higher level abstraction, I have my backup software. Then I have my disk. Let's say that I'm worried about my backup software being compromised.

Does Falconstor give me the ability to protect against that layer of compromise if the malware was written specifically for, let's say, one of the big ISB software providers? >> Well, so we can certainly take a snapshot of that software at that point in time, so that if that server were to become compromised, we can roll back the software. We can roll back the data. That's something that certainly we had people calling up about that when we saw what happened with solar winds, and this is the....

this topic of conversation around immutability, restorability and then just security in the first place. Where's the encryption taking place? Since we saw at least in the US, the hijacking, the ransomware that happened in Southeast, right? With the power company, my God. That's been the topic of conversation from these customers. And the answer is, yes, you can do that. And increasingly that's key for these MSPs to be able to say to those customers, not just that you have a data headache, or you have a backup headache, you have a security headache.

You can blend those together for this secondary, you know, this copy. >> So let's talk dollars and cents. MSP margins can be thin at time. I need a solution that is predictable in cost. >> Yeah. >> If customer's data footprints go up and down, I need some, you know, some level of comfort in my cost when it comes to data protection >> Right. Or I have to build in that cost model to customers and customers don't really care that much for that type of cost model.

What's the cost model for MSPs? For buyers? >> Yeah. Great question. And I think there's two components to it. The first one is how much data do I actually have to store in the first place? So, the history of Falconstor, when I got to know it 20 years ago, was these are the De-dupe experts. And now De-dupe as a term, as you well know, is commoditized. Everyone's got it. Everyone does the same thing. >> Everyone's got De-dupe.

That is far from the case when you get down into the details. And so we routinely see 20 to one De-dupe for our enterprise-oriented clients. I have an MSP that came on and this is the technical guy, comes on and says, "We're getting a hundred to one". So, why? Because there's low change rates for some of these environments. Of course, it all varies. But if you get that set level of De-dupe, then you've got a way of understanding what kind of capacity you're going to need.

And then we turn that on its head and said, "Well, our pricing model is this". It's just a flat rate per tenant. So that tenant could be, you know, a hundred terabytes. It could be a petabyte. It doesn't matter to us, one flat rate. And then the margin profile for the MSP goes up because of the De-dupe. >> I was just about to say, it seems like you guys are losing a little money on the table here, >> Well.

because as an MSP, if I can, you know, get up to a hundred to one De-dupe, I can take advantage of that in the public cloud, or even on premises, if, depending on my cloud model, but public cloud, the less I take up, the less I spend and the more customers I can have. And >> That's right. you guys, I don't know. That might... >> May not be the brightest business decision, but I'll tell you what, here's the thing. If you're going to get that kind of rate or something approaching that rate, because it always is going to be changing.

Then, we are going to see MSPs continue to adopt this for additional and additional and additional requirements. And so it's going to pay, it's going to wind up paying us back in the form of new customers. And so this way, our interests as a business are totally aligned to the success of the MSP. And one thing additional, then we have our marketing people come in and throw back some MDF for them, so they can go build out their marketing programs to go capture more business.

So just don't tell them, just don't tell anyone that I said that. >> I was actually just going to ask you about the relationship part of it and that answers the relationship question. But I want to circle back to kind of the larger market. >> Yeah. >> Big OEMs are waking up to the potential specifically, we're working with partners on this. I've gotten into massive debates on Twitter about multi cloud, blah, blah, blah. The reality is customers need people with feet on the ground, solving these difficult, interweaved problems.

And that's the partners. The partners are the ones doing the hard work, and they have a lot of people in their ear. >> Yes. >> And they're trying to decipher the signal from the noise. Why Falconstor? Like, I can go with of the big, fortune 500 vendors. Why little Falconstor? >> Yeah, no, it's a great question. So we're betting the farm on the success of the MSP. I was just looking at a research note from grand view research that said 50% of IT will be delivered by the MSP.

So, we're all in on the MSP side of things. Why come with us? Because an MSP is the one where they've got the skillsets, they've got the talent and the people to go out and get the best. And so if you want the gargantuan that can serve everybody, of course, hey, there's a lot of safe choices out there. But if you want the thing that can fit the heterogeneity and provide this level of throughput and protection, as well as the economics, then we know that MSPs, a lot of them are small businesses.

This sounds like a good, you know, so we're really trying to build a good small business value proposition. >> So Chris, I'll give the last message to you. What did I miss it in the conversation with MSPs? We talked about how do MSPs make money. We talked about the technical capabilities and we talked about why Falconstor. >> Sure. What'd I miss? >> So, I don't think you missed anything. I think the most important thing goes back to that heterogeneity.

We know, lot of the backup and recovery software providers are saying, we are not only the backup and recovery software, we're the disk-based target and we're the cloud, right? That's something that certainly you can go with. But again, it comes back to the capacity. It comes back to the data reduction and then ultimately, the heterogeneity that you want to project so that you can deliver your clients, basically any combination that they want, right? Now within limits, right? You come to our HCL and you can see what combinations are really there, right?

But that will help to define a range of capabilities for customers, as opposed to just standardizing on one thing. >> Well, Chris, I appreciate you taking the time out, although it's not that tough. This is actually really beautiful around here to come have a couple of adult beverages. You're actually having a, what is that called? >> I am. I'm having a mocktail here >> A mocktail. >> Yeah. >> So not quite a cocktail, but.... >> It's still early here on the west coast, Keith.

I know this is not, you know, I know you're still on your road trip. >> It feels like five o'clock. >> Yeah, it does. >> It's been quite a week. If you want to find out more about Falconstor, you can follow the link below you. If you want to find out more about the CTO advisor road trip, the CTO advisor dot com forward slash road dash trip. If I missed a question or if you think we got it completely wrong, you can DM me at CTO advisor on Twitter.

DMS are open. I'm brave like that. Or you can comment in this section below and we'll have the conversation. I'll fill the problem. Until the next CTO advisor. Talk to you, online.