You are here

Interview: Lessons Learned: A Software Leader's Move to SaaS

In this interview, Tim Gage, head of research and development at Alfa, shares an insider’s view on how Alfa tackled moving to a cloud model, including the technology decisions they’ve made to mitigate the risks and maximize the benefits of the cloud.

Video transcript: 

Jeff Boehm:    Hello, and welcome to our webinar on Lessons Learned: A Software Leader Moves to SAAS.  My name is Jeff Boehm, I am the chief marketing officer at NuoDB, and today we’re going to learn a little bit more about a software leader and their move to the cloud.

NuoDB is on the forefront of a new generation of databases that we call Elastic SQL.  Elastic SQL databases bring together two critical elements as companies are moving to the cloud.  The first is that it is SQL-based.  It is ANSI standard SQL, fully asset-compliant, which means you can trust your operational applications to run on this system.  It gives you the capabilities to have durability and consistency and trust in your data as you move to the cloud.

The second is that it is elastic, which means it can scale out seamlessly as you move to the cloud.  If you need to add capacity, add performance, resiliency, you simply add a node.  As that performance or capacity needs diminish, you simply remove a node, which means simple deployment across the cloud architecture while maintaining SQL and asset compliance.  It really does bring together the best of traditional database systems with cloud architectures.

We work with a number of leading customers around the world today, from major software companies like Dassault Systèmes to software development organizations in leading enterprises like the London Stock Exchange.  Today, I’m pleased to be joined by a colleague from Alfa Systems, Tim Gage, who is the head of research and development at Alfa.  And he will talk a little bit about Alfa’s move to the cloud.  Welcome, Tim, and thank you for joining us.

Tim Gage:     Thank you for having me.

BOEHM:    So let’s start by -- maybe you can talk a little bit about your role at -- maybe introduce Alfa and then talk a little bit about your role at Alfa.

GAGE:     Sure.  So Alfa provides software to the asset finance industry.  The most common example of this is vehicle finance.  So if you want to get a vehicle on finance in the US or in the UK or in many countries, you may find that your contract is actually managed by our system.  So it does everything from pricing the contract up front, pricing the repayments, to sending out the request for payments, receiving the money, and receiving the receipts, and then handling all the in-life changes - chasing you if you don’t pay, and then handing the end of the contracts or handling the returning of the vehicle.

So it’s not just vehicles.  It’s also other kinds of assets, some quite interesting ones - things like oil rigs or trains, but also tractors, farm equipment.  Basically, any kind of situation where you’ve got an asset that’s under finance and there’s a contract associated with that asset.

BOEHM:    OK.  And then your role at Alfa - as I introduced, you are head of software research and development there.  Can you talk a little bit about that role and your tenure at Alfa?

GAGE:     Sure.  So I’ve been at Alfa for nearly 16 years now.  A long time.

BOEHM:    OK.

GAGE:     We have a joint CTO role at the company, which I share half of, and I also lead the R&D team.  We have a team of six people who are working on moving the product forward and bringing new technology into the system.

BOEHM:    Great.  So my understanding is that most of Alfa’s deployments today are on-premises, but there’s clearly a move in the marketplace towards cloud deployments.  Can you talk a little bit about how you’re seeing that and what you’re seeing as companies move towards the cloud?

GAGE:     Sure.  There’s a few things driving that move.  I think that the first one is primarily cost.  There’s -- you can -- by moving towards the cloud-based deployment, you can replace a large up-front investment in software and in hardware with regular payments, something that you can control and something you know up front how much it’s going to be.  The second thing is improved service, right?  I think you can actually -- we think we can offer a much better service through deploying our product in the cloud.  We know how to manage our software, and I think we can do it really well if we do it ourselves.

And the other thing is that there’s -- the cloud offers you all sorts of technical advantages, things like the ability to dynamically scale your software or your hardware.  If you know that you’ve got a spiky demand, if you’ve got certain times of the day when you’ve got to do a lot of work, then you can dynamically add hardware to your system and make use of more hardware to make the system run faster and more efficient-- make -- you know, you don’t have to spend so much money on having hardware available all the time to deal with that spiky load.

BOEHM:    Ah, that makes sense.  So certainly in the marketplace at large, there’s this trend not only around cloud but specifically around Software as a Service.  Is that -- how do you view that difference at Alfa?  Do you view yourselves as a Software as a Service, or cloud, or what do you see as that difference?

GAGE:     Right.  So we have a slightly different take on that.

BOEHM:    OK.

GAGE:     So what we -- the way -- a bit of background, I guess, is that we have a slightly different model to many software companies.  Typically, a software firm either ships a product that is kind of shrink-wrapped, and they sell the same -- exactly the same software hundreds, thousands, millions of times over, or they have a single deployment of their product, right?  So maybe some of the financial firms or Facebook or Google, they have one instance of that platform.

BOEHM:    Right.

GAGE:     Typically, companies tend to fall into one or another of those categories.  We actually sit in between.  We have a small number of deployments, in the tens, which are very highly customized to our clients.  Because the core product is the same, but we’ve got interfaces and various other aspects and changes to the deployment model that suits that particular client.  So the consequence of that is we don’t call it -- we don’t call it Software as a Service.  We call it more of a hosted model.  So we are hosting a software on behalf of our clients.

BOEHM:    Got it.  OK.  So as you move from the on-premises deployment to this hosted model and this hosted deployment, how do you see your architecture changing?  Or what impact does that have on your product and your product and your product architecture?

GAGE:     So mostly we want to keep the architecture the same.

BOEHM:    Sure.

GAGE:     We want to be able to deploy the same software in on-premises environments as we can -- as we do in the cloud.  That simplifies the way in which we develop the software.  We don’t want to have to -- basically to give up the ability to deploy on-premises in order to deploy in the cloud.  So we’re looking for something that works in both situations.

BOEHM:    Got it.  OK.  So as you move to these cloud-based deployment models, what sort of technological or architectural changes have you had to think about as you move to this new deployment model?

GAGE:     Right.  So the core product actually transports extremely well to the cloud, right?  We have a -- it’s a Java application; it runs in an application server, it runs on Linux, it’s deployed primarily over the Web.  So that element is fairly straightforward; we can move our application to AWS and it basically just works.  That works really well.  The challenges, I guess, come around the database model, when quite a few of the existing traditional databases tend to go for a more vertical scaling approach, when they don’t tend to take advantage of the cloud features, which I think is, like, being able to add --

BOEHM:    Got it.  Yup.

GAGE:     -- new instances or new hardware -- bring new hardware online dynamically to add to the capacity of your system.

BOEHM:    OK.  So I guess that -- you know, so as you think about this move and the role of the database, I think you’ve kind of answered this, but how do you -- you know, why couldn’t you just take your existing database technologies and simply run them in the cloud?

GAGE:     Well, I mean, that -- that was an option.

BOEHM:    Sure.

GAGE:     And it remains an option.  We did look at that, but there’s -- like I said, they just -- the scaling elements of the traditional databases don’t work particularly well in the cloud.  They’re not particularly well-suited or architected to work in the cloud.  So we decided to look at other options.

BOEHM:    OK.  And I guess overall, how important would you say the database is to your product architecture, whether -- I guess whether that’s on-premises or in the cloud, but especially as you’re moving to the cloud, how -- sort of what role does the database play?  How important is that?

GAGE:     Oh, it’s crucial, right?  The database is absolutely foundational to the system.  Ultimately, it’s a transactional system that handles financial contracts, and the database is the basis on which we build the application, and we need something that actually is completely solid and dependable.

BOEHM:    Got it.  So as you were thinking about the shift and the move to the cloud and the challenges with existing database technologies, I guess, what -- what database technologies did you consider?  What did you look at as options for the database for your cloud-based deployments?

GAGE:     So we’ve been looking at options for a little while -- we’ve been keeping an eye on the market.

BOEHM:    Sure.

GAGE:     We have considered things like Cassandra, and it was an option, and many of the things that Alfa requires of a database are fulfilled by things like Cassandra and MongoDB.  The problem is they’re not fully -- they’re not full SQL implementations.

BOEHM:    Sure.

GAGE:     So supporting something like Cassandra would have required a major re-architecting of the system, and we didn’t want to go down that road, because in doing that we would have probably lost the ability to run on the more traditional databases.  So what we were looking for was something that actually had a full SQL platform support.

BOEHM:    Got it.  And how did you first come across NuoDB?

GAGE:     So we were looking at these options, I guess --

BOEHM:    Yeah.

GAGE:     -- and we were looking at the various new SQL options --

BOEHM:    Yeah.

GAGE:     -- which are coming along, and NuoDB came up and we looked at it in more detail and we liked what we saw.

BOEHM:    Good.  So let’s talk about sort of your engagement with NuoDB, then.  Talk about -- how did you start evaluating NuoDB?  What did you do to start taking a look at it?

GAGE:     Right.  So there’s a Community Edition on the website, which we downloaded -- that was where we got started.  We read some of the documentation, had a bit of a play with it, and we decided that it did look like it was going to be a pretty good fit for our product, so I asked one of the guys on my team to have a go at writing a dialect, so we have an abstraction layer between the system and the database, and we got one of the guys -- I got one of the guys on the team to have a go at writing this dialect.  That went very well,  It took them about 10 days or so to get the majority of the support in, and basically we had the system up and running, and that went very well.  We got, basically, pretty good results out of some basic testing, so we decided to move on to doing a full evaluation, and at that point we got in touch and requested an Enterprise trial license.

BOEHM:    Great.  And so that first phase, then, was purely on your own, using the Community Edition really without any support from NuoDB per se.

GAGE:     That’s right, yeah.  Mm-hmm.

BOEHM:    OK.  And you were able to do that on your own?

GAGE:     That’s right, yeah.  The Community Edition requires almost no configuration.

BOEHM:    Yeah.

GAGE:     You could download it, install it, and it just works.  It works out of the box.  It’s very good.

BOEHM:    OK.  So then, when you moved on to more of an Enterprise trial, talk a little bit about that.  What changed then, or what additional testing you were able to do, or how did you go about that?

GAGE:     Yeah.  So deploying the Enterprise version with NuoDB with a minimally redundant configuration does require a little bit more configuration.

BOEHM:    Sure.

GAGE:     So we had a bit of help and assistance from some NuoDB consultants who came in and helped us out, got us started, basically.  They essentially taught us how to set up the system and on multiple nodes.  And then we went through a process of various kinds of testing that we wanted to do.  We basically built up the volume, so we had a sentence -- we put in a realistic scale of database, something moderately large.  We then started to push a load at the system using Web services and using the UI, and we’ve also got batch processes that we need to run.  So we went -- basically ran through a series of tests.  The kind of thing we did, we put use to this stuff, we do performance tests all the time.  That gave us pretty good results.  One thing, actually, we were particularly impressed by was the fact that the system requires very little configuration to get to the point where it’s running really well, and our deployment model is relatively simple, and almost out of the box, the performance is very good, very impressive.

BOEHM:    Good.  Good.

GAGE:     Which is good.

BOEHM:    OK.

GAGE:     Yeah.  So then once we’d got through that testing phase, we moved on to some failover testing, in order to test the various features that NuoDB has, like the ability to recover from a node failure.  So we start to do things like kill transaction engines and storage managers and check that the system recovered correctly from those scenarios. We did encounter a few issues doing this, which was interesting, but we -- we got good support from NuoDB.  We had the consultants available.  A couple of things required configuration changes to the system, to NuoDB or to Alfa, and then we retested and the problems went away.  We did actually find a bug in NuoDB, and we had pretty good interactions with the support team, and it was great to test that out, actually, and we got some good results.  And that bug has now been fixed and is --

BOEHM:    Good.  Glad to hear it.

GAGE:     -- now it works, yeah.

BOEHM:    Good, good.  So, I guess, given your testing and given your experiences, what led you to select NuoDB?  What were sort of the key reasons that made you decide NuoDB was the right option for Alfa?

GAGE:     Right.  So firstly, it is able to support our product, all right?  We have a complex and large products, and like I said earlier on, not all of the other platforms out there are able to support the full Alfa set of SQL standards that we run, all the functionality that we have.  So that was the initial step.  Secondly, it has this native support for -- it’s designed for running in the cloud, on AWS, on that kind of horizontal scaling hardware.  So that was a great fit for the kind of thing we were looking for for our hosted deployment.  Thirdly, all the technical advantages -- so things like the active-active deployment model, the elastic scaling dynamic -- ability to add nodes at run time, all that kind of stuff, you know, it’s all good features for us.  But there’s one thing, actually, we’re particularly excited about, which is that there’s -- downstream, we’re not there yet, but we might get to the place where we can have cross-region deployments.  So we might have users of the system who are both in -- let’s say, for example, New York and in London, and these guys want to have access to the local system, right?  They want fast performance, so they get really good response times, but they still want access to the global data set.  And NuoDB offers the possibility of having that kind of data set distributed over the -- across the Atlantic, in that example.

BOEHM:    Great.  So given your experiences so far, I guess, how do you foresee customer deployments and the database options evolving over time?  Is this -- obviously getting your -- this was selected as part of starting to deploy customers in the cloud.  Is that where you principally see NuoDB being used, or do you see it being used on-premises as well?  And how do you see that shift happening?

GAGE:     So one of the good things for us about NuoDB is the ability for us to deploy it both in the cloud and on-premises for future customers or, potentially, existing ones.  So I think we’re going to see most of our hosted offerings go -- quite possibly all of our hosted offerings being deployed on NuoDB from later this year.  Then we might start to see -- well, we’re also going to start to see new clients coming on, new prospects basically for us, that may well be very interested in NuoDB, and we expect that -- expect them to be interested in that stuff.  And then in the longer-term, we might see some of the existing clients want to move over onto the platform, depending on when they reach their end of their investments and their current platforms.  But I think that will depend, and quite a few of them having existing investments in technology stacks and new skills and all that kind of stuff.  So it will vary.  It depends on the client.

BOEHM:    And do you foresee customers, some of your on-premises customers, moving to the cloud then as well, and some staying on-premises, some moving to the cloud?

GAGE:     So we started exploring that with some of our existing clients, and they are absolutely interested in that kind of thing.  There’s all sorts of technical hurdles that come up from that kind of thing.  But I think it might well happen; I think it will depend entirely on the circumstances and on the individual clients in question.

BOEHM:    OK.  All right.  And, you know, what -- how have your experiences been so far?  So you went through the evaluation, then obviously you purchased licenses then from NuoDB, and you’re going through the process, I believe, now of getting ready to deploy it.  And I guess -- how have your experiences been since that evaluation process, as you’ve now been moving closer to production?

GAGE:     Yeah, that’s right.  So we’ve not yet in production on NuoDB in our host environment; we expect to be basically late this summer.  We’re still in the planning stages, but we’ve had a good experience.  We’ve -- yeah, we’ve been working with NuoDB on choosing a configuration that’s going to be appropriate for us and appropriate for the client the way -- starting with, starting fairly simple, and then we’re going to have the option to ramp that stuff up later, as soon as we -- as we increase the scales and as we bring on other clients.

BOEHM:    Great.  So as you anticipate these rollouts, then, what sort of advantages do you foresee for your customers?  What benefits will they experience by having NuoDB as the underlying database?

GAGE:     Right.  So the -- our customers don’t see the database directly, typically.  So the kind of things that they’re going to see are the active-active deployment model, giving them increased uptime, longer and better resilience, the ability to cope with, you know -- uptime, resilience, active-active, all that kind of good stuff that actually comes out of using NuoDB.

BOEHM:    Yup.

GAGE:     The other factor is that the lower -- potentially lower licensing costs for NuoDB --

BOEHM:    Oh, OK.

GAGE:     -- might mean that we can offer the product more cheaply, Alfa as a whole more cheaply, as the total cost of ownership might be lower for those guys.

BOEHM:    Good.  And so what about for Alfa?  I mean, what do you foresee as the benefits of going with NuoDB for you, for your organization?

GAGE:     So we think that ultimately, we should be able to ship a better product.

BOEHM:    OK.

GAGE:     So by -- like I said earlier on, we’re -- the database is the foundation on which we build, and if we have a stronger foundation with all this extra support for active-active continuous uptime, then ultimately what we’re delivering to our clients is a better product, and will be better than what our competitors can currently offer.

BOEHM:    That’s right.

GAGE:     But also, we’ve had great support from the NuoDB team so far, in our work with them. We had worked through a few issues on both sides, and we’ve really enjoyed that.  It’s great people to work with, and we’re enjoying it, and it feels like a true partnership.

BOEHM:    OK.  Good.  If you -- as you think about your experiences so far with NuoDB, as you talk to your colleagues within Alfa or if you were to talk to people outside of Alfa, you know, other companies in your industry or potentially even your customers, if you talk to them about the database -- how would you -- how do you describe NuoDB?  What do you think of as -- what is NuoDB?

GAGE:     I guess the key elements for us is it’s a full SQL implementation of a database, and we see the -- you know, the active-active continuous availability and dynamic scaling, all of these great features.  Actually, one thing that’s worth saying is that quite a few of our engineers at Alfa are really interested by this product --

BOEHM:    Oh, good.

GAGE:     -- right?  They’re really enjoying the idea and the prospect of actually working on this stuff, working on this new technology.  And that’s an important thing for us, right?

BOEHM:    Absolutely.

GAGE:     We like to keep our engineers happy, and we want to get them working on --

BOEHM:    So it almost becomes a help for you for recruiting or retention, to say --

GAGE:     That’s right.  Absolutely. Which is --

BOEHM:    -- you’re working on this new generation technology.

GAGE:     -- which is always a challenge, yeah.

BOEHM:    Yeah.  Oh, that’s great to hear.  Good.  We will be opening this up for live questions in just a few minutes, so if you do have any questions for either myself or Tim, feel free to type them into the question box, and I will be looking at them here on my monitor and asking them.  But before we get to that, I guess, what recommendations would you have for other software development organizations, whether they be software companies such as yourselves or organization -- software development teams within larger companies?  What sorts of recommendations would you have as they consider a move to the cloud or SAAS architecture?

GAGE:     Sure.  So I think it’s really very difficult to generalize this kind of thing, and I think it’s going to vary enormously depending on the particular piece of software that you’re moving and the way that you use it, and your users and how you want to deploy, all that kind of stuff.  But there’s -- I guess there’s a few things that we did learn and that we perhaps -- it might help to pass on.  First of all, understand what you want to get out of this process, make sure what you know what your goals are, and it can really help with actually driving the decisions you make.  Secondly, security is a key, right?

BOEHM:    Sure.

GAGE:     So security is hard to get right, and we put a lot of investment in, as we did this model, to getting security right.  And we actually think we’ve got a very good, very secure solution now.  In fact, it may even be better than many our clients’ on-premises deployments.  So I think that’s a really important thing, and, you know, you consider getting external help on getting that stuff right.

BOEHM:    Yup.

GAGE:     And thirdly, consider NuoDB as a platform.

BOEHM:    OK.  (laughter)

GAGE:     It works.

BOEHM:    Thank you.

GAGE:     It works well.

BOEHM:    (laughter) Good.  I’m glad to hear that.  So let’s see -- I do see some questions coming in here, and again, if you do have any questions out there, feel free to type them into the question box.  There’s one here about your configuration for your test deployment.  You mentioned that it was a minimally-redundant system.  Can you talk a little bit about the NuoDB architecture -- as you obviously know, there’s this notion of separating the transaction processing from the storage processing, into something we call transaction engines and storage managers.  What sort of configuration were you testing out, especially with the Enterprise trial version of the product?

GAGE:     That’s right.  So the minimally-redundant configuration, which is hard to say, (laughter) has two transaction engines and two storage managers.  And so I do have -- that is there.  So if you lose any -- either a transaction manager or a storage -- either a transaction engine or a storage manager, then the system continues to run seamlessly.

BOEHM:    OK.

GAGE:     So that’s what we were aiming for for our base deployment, and we would do that across two availability zones within one AWS region.

BOEHM:    Oh, OK.  So you’re in AWS -- within one AWS region, but across two different availability zones.

GAGE:     That’s right.  So if one availability zone goes away --

BOEHM:    Oh, great.

GAGE:     -- then the system continues. So that’s the active-active goal.

BOEHM:    Right.  Good.  OK.  And what sort of -- next question is, what sort of volumes did you test?  What sort of -- you mentioned sort of basic database load, but what -- can you share any statistics or numbers on the sort of database volumes that you were testing?

GAGE:     Sure.  So I guess it’s kind of difficult to -- to take numbers from our system and then apply them generally, but in our testing, we were talking about databases that are of the order of a few hundred gigabytes.  We’re pushing through loads of up to a few hundred transactions per second.  That’s -- this is a medium-size deployment for us, and we actually would like -- would very much like to push that volume up, and our testing showed that NuoDB has the capability to go faster and further, but that was -- and I -- basically, our target was to show that we had support for that kind of initial volumes.

BOEHM:    Got it.  OK.  Question here: what sort of schema and indexing changes did you need to make to your application in moving from -- and actually, we didn’t touch on this.  Maybe if you want to touch a little bit on sort of how you’ve architected your system to sit on top of -- are you specific to a relational database system, or are you using more sort of standard -- ANSI standard SQL? How did you -- you know, sort of how did you -- what sort of schema and indexing changes did you need to make?

GAGE:     So the way we -- so we built the system to be agnostic of the databases underneath it.  And the way we do that is we have an abstraction layer between us and the database.  We actually use Hibernate for a lot of the internal modeling and queries.  And that takes away a lot of the complexity.  We’ve also got a -- excuse me.  We’ve got a SQL model that we can use.  We’ve basically got a SQL DSL you can write that you want to -- if a select statement, it selects these fields, and then at runtime we can then translate that using the dialect into a SQL statement that will run on the target database in question.  So that was the piece that we had to do, the piece that we had to write in order for us to have support for NuoDB from our product.  So I guess in terms of what -- the question of what sort of schema and index changes --

BOEHM:    Yeah.

GAGE:     -- the overall schema and index changes were -- the schema and index is the same, right?  There’s the -- we have -- we basically had to, just had to map from types on other systems and choose -- for every type in our model, we have to choose a type on the target platform.

BOEHM:    Sure.

GAGE:     And NuoDB has types that match the -- support all the things we needed.  So although the schema doesn’t -- right.  There -- basically, from a -- sort of an abstraction layer point of view, the schema that we’ve got and the indexes that are present in the database are exactly the same as they are on any of our other platforms.

BOEHM:    OK.  Good.  And I’m guessing, based on what you were saying earlier, that had you gone with a Cassandra or a NoSQL-type system, there would have been considerably more work on that side of things to make your application run.

GAGE:     Right.  So -- yeah.  I mean, I think the challenges with things like Cassandra is that you can’t do joins across multiple tables.  So we have got that kind of query in abundance in our systems, so that was the kind of challenge we had here.  So that kind of -- we couldn’t -- you can’t map that.  Our query model was not mappable onto a Cassandra database model.

BOEHM:    OK.  Question here on performance: in your testing so far, how did the performance, or how have you seen the performance of NuoDB stacking up against other traditional relational database systems that you’ve been using?

GAGE:     Yeah.  So I think that the -- I guess the question is, on comparable hardware, what was the performance going to be?  How does the performance compare?  I think it -- the performance basically is very good.  It’s basically -- it’s very comparable, in some cases better.  And in other cases, we had some -- we’ve had a few places where the network latency and things like that has a small effect, but basically by and large, the performance is very comparable and is very good.

BOEHM:    OK.

GAGE:     So it’s kind of a difficult question to answer in a general way, but --

BOEHM:    And a full apples-to-apples, too, because --

GAGE:     Right.

BOEHM:    -- as you talk about --

GAGE:     As you --

BOEHM:    -- sort of adding nodes and scaling that out --

GAGE:     Yeah, that’s right.  So you can’t do comparable hardware, because I’ve -- actually have got multiple EC2 instances running your NuoDB database, and you can’t do that with the traditional platforms.  You have to actually have one very large instance, and that is a fundamentally different beast.  So --

BOEHM:    Yeah.  So in some ways, you need to think about performance testing in a completely different way, or --

GAGE:     Yeah.  Yeah.

BOEHM:    -- or comparable testing --

GAGE:     I suppose another way you can compare it is actually based on the performance per cost. 

BOEHM:    OK.

GAGE:     I think at that level, NuoDB certainly is extremely strong.

BOEHM:    Yep.  Yep.

GAGE:     Because you can -- you can deploy it on commodity hardware and you get very good performance out of that.

BOEHM:    Great.  A question, and you may have a limited view on this, just given how you’ve used it: the question is, what sorts of applications do you think are best-suited for NuoDB?

GAGE:     That’s -- yeah.  I’ve --

BOEHM:    If you have a specific view on -- applications.

GAGE:     -- yeah, we have a particular view.  So our application is suited to working on NuoDB.  I think the -- some of the key features are that it is based on a SQL platform, it does use a SQL platform, and we’ve got this abstraction layer.  We’ve got -- we already had the ability to run against multiple databases that we needed in order to support various clients that are running on different platforms.  And because we had that, it was relatively easy for us to move across.  So I think it depends.

BOEHM:    Yup.

GAGE:     There’s no simple answer to that.

BOEHM:    Sure.  Sure.  You touched on this a little bit earlier, but do you envision supporting multiple platforms with NuoDB?  And I think what the question means by that is, you’ve talked about, you know, sort of on-premises versus AWS.  Do you imagine having NuoDB running the same in both -- in both areas?

GAGE:     I think we’d like to keep it as similar as possible.  All right, so for us, if we can come up with a configuration that works well for Alfa, for our product, then we’ll deploy it and AWS will deploy it with our clients on-premises.  And the closer we can get those things, the easier it is for us to support, right?  And we’d like to move to a model where our architecture is the same everywhere, so that we’ve got the simplest possible support mechanism for that.

BOEHM:    Sure.  OK.  All right.  And again, if there are -- I think I’ve addressed most of the questions at this point that have come in through the webinar, but if there are other questions, feel free to type them into the box.  I guess another question, along the lines of the final question I asked you earlier around recommendations, was -- what recommendations do you have for evaluating or testing NuoDB?  So I think it kind of gets back to your experience with --

GAGE:     Yeah.

BOEHM:    -- how you tested it, but what would you recommend if somebody wanted to evaluate or test out NuoDB?  How would you recommend they go about doing that?

GAGE:     I think the process we followed was pretty good.  So you can get started on the Community Edition very quickly; you can just go ahead and download it.  And that will give you a good idea of whether the overall -- can you build your application against -- can you make it work against this platform?  Because it has the same features, right?  The Community Edition has all the features of the Enterprise edition, it just doesn’t have the ability to distribute it across multiple instances.

BOEHM:    Sure.

GAGE:     So once you’ve -- if you’ve gotten to the point where you’re happy with it, the Community Edition, and they think you want to move forward, then get in touch and then go for the Enterprise license trial, and work through the various things that we did -- do some performance testing, do some load testing, and do some failover testing, to make sure that all the promises hold true.

BOEHM:    OK.  All right.  Good.  Well, I think I have addressed all the questions that have come in, and certainly, I don't know if there’s any final words you have, any final words of wisdom for our audiences that are considering a --

GAGE:     I don’t think so.  I wouldn’t dream of offering any major words of wisdom.

BOEHM:    (laughter) So certainly appreciate your time today, and hopefully the audience found this insightful as we learned about Alfa’s journey to the cloud, and sort of where you are and the database considerations you had for that.

GAGE:     Yup.

BOEHM:    And certainly if you want to learn more about NuoDB, you can visit our website or download the Community Edition, as Tim said, as a place to get started.  And also, if you want to learn more about Alfa, you can go to alfasystems.com and certainly learn more about your company and the sorts of customers you’re working with and the products that you’re providing.  And --

GAGE:     Yeah, that’s right, and the jobs and roles that we have available for people that want to come and work for us.  We’ve got interesting stuff.

BOEHM:    And the jobs and roles that -- and they can go work on NuoDB, at --

GAGE:     Yeah.  That’s right.

BOEHM:    -- at Alfa.  So.  Good.  OK.  Well, thank you -- again, thank you for your time today, and thank you all for joining us, and we will call that a wrap.  Have a good day.