VMworld 2018 - Uila
Transcript
[Music] everybody welcome to CTO DOS mark may and I have of course Keith Townsend we have a special guest I'm really excited to talk to you today you want to introduce yourself yeah hi everybody cha Chi Quan from Willa I'm the CEO co-founder for Willow were five year old you know we still call the starter company talking with you guys second time love to do that every time yeah when we had a good conversation what was that six months ago maybe a year a little bit
more than six months ago yeah months ago yeah so you know what I think what I wanted to talk about today was as you guys have evolved and customers have evolved from you know it's on-prem infrastructure to hybrid cloud and journey to the cloud how are you guys adapting and helping your customers adapt more importantly right right so as you know our company does data centers started out from five years ago monitoring the private data center with the application focus right it's all private but for
the infrastructure guy we provide them application level visibility so that's the thing that we do primarily started out as you mentioned nowadays people go hybrid or even public with that journey we go with our customer so the first step with them is you know the data that we collected will help them understand the data center today the challenge first step of challenge about migration or even thinking about the public or even private is do I understand enough about my current data center how is the application
put together if I even want to move one service how is that service all right so that's number one that we do analytics that we own and we do a lot of those application dependency mapping that's usually the first step in RIT usually the first challenge so I'm interested in the reverse we've heard about born in the cloud companies that move back on prim you guys have the born in cloud companies that are going through that reverse scenario where they were born in the cloud they're
figuring out you know what this work a little too stable secure really doesn't need the elasticity a cloud it's best to move some of this stuff internally yeah DC or you're right they weren't born in the cloud they decided to go all in all the cloud early for whatever reason you know it's gonna save money maybe it didn't pan out so now they're starting to pull it back I like that that's a very interesting idea yeah definitely I mean on our side we've certainly heard a
lot of those but we on our customer base it's a little bit different from that but in spirit is similar it is there they go to the cloud from private and then come back right for various reasons yeah yeah and the reason for them to come back is either for cost reasons right which is easy to explain and secondly is for not have planted well from the first place and the operation actually doesn't work out for performance reason they have to come back hmm so have
you guys helped like in that decision-making tree to say oh you know what the problem is the uncertainty a cloud or the problem is something that's native to cloud you know cloud is obviously I can build a resilient performing app one cloud but a lot of that boils down to if I have unlimited funds like I can go across multiple AWS regions blah blah blah blah blah blah and build a resilient app but there's these there's key pieces of data that need to make that decision
you guys help in that data yeah absolutely after so as we find out to the cloud the cost ways people are understanding resourcing the cloud cost you money the CPU the compute power the memory discharge everybody understand that and there are lots of to help you understand that the most neglected two things that give you costs number one is the traffic between your clouds your private or public to public or private to public we're that VPN connection you move some of your workload to the cloud
where you don't know you drag alone something that's still in-house or you know they exist you don't know how much traffic that is that VPN tunnel that costs you a lot of money and then sometime cause you not only the money but the performance so I would imagine I know you guys giveaway I saw this coupon for a free like application assessment right I would imagine that is something that should help people avoid that absolutely knowing ok I'm moving this to the public cloud it's gonna
talk a lot to this that I don't want to move to the public cloud or maybe I want to move it to a different cloud provider right we have many customers that has you know tens of thousands of virtual machines being their data center they made the decision to go hybrid or go public the first step is ok how do we even start right what do I have in my data center so then we run our analytic report which is we call the migration assessment report
we identify all those inter connectivity between all these services then separate it out into different services ok ok I'm gonna move this particular service it's not ten thousand VM is a hundred and twenty PM still big right but then we'll help there analyze break it down into pieces do they move this and not that or do this or not that and then here's that here's this how much it's going to cost there how much traffic is gonna be here and then this used to be like
multiple weeks if not months of effort in our case once you deploy our solution in a few hours and then one click voila this would come out so so key I know that application assessment kind of idea has definitely mirrored what I've seen from customers and peers right is that what you've seen in your experience oh yeah well that what to do that the application assessment I think it was a huge part of the one most people just don't have a map of how their applications
don't know and when you and a scenario I saw on Twitter of all places put it's a great place to get real information sometimes sometimes sometimes is a scenario where someone takes a service that they think will save them money you know I have a friend that has a database and small application and it's really hard to decide if RDS is the right move from the me but even beyond that if I've taken the application I've seen the dependencies based off of the reporting that I've
gotten from wheeler and I say oh there is a bunch of traffic between this application server and this database server I can break down this application server into micro services and give that to lambda functions and allow those lambda functions to do the processing and theory maybe I'll save some ec2 costs and this be great but in reality was what some customers find based on the scenario they end up like your VPN example they end up increasing cost in the area like the API gateway right
and the cost becomes X or burnt and instead of saving cost they've tripled their cost right you guys help identify that type of dependency and and break-ins I know first originally you guys look at the application from the network stack can you guys inject yourself into the cloud native type services like surely certainly so there we measures is you know the piety despite the or in addition to the dependency who's talking to who is a lot of quantitative information so number one is traffic volume right
how much the traffic volume from flow from A to B if that's VPN particular these VPN how much that is and second thing that we also want people to may pay attention to is how many transactions write transactions is not directly a map of traffic volume that's important too that shows your app to application visibility or application kind of a volume in their activities the number three is probably even more important is the application performance from point A to point B or even proved to for
example Amazon provided services RDS or these SAS based services and then even though you don't own those services you're getting it from Amazon or other services salesforce.com we still want you to know your applications impacted because of those external services that you depend on so all these mer measurements come into one help you not only on cost front but also application performance delivery from so we talked a little bit about cost and how that can impact you in the cloud do you actually show what the
cost would be do you never integrate with the Amazon public pricing and say here's how much it would cost or as your public pricing right right that's a tricky one so today yeah there's there's businesses there there's like whole solutions around exactly so our business is not primarily that so we based on the Amazon published kind of own discounted and own specialized deal right we're based on that and every enterprise have their special deals with Amazon different pricing we certainly don't know so we kind of
just based on the public information and a user change the numbers so they can tweak it to their own liking Nadia today right yes so we're at VMworld 2018 VMware has made some key announcements one of which i think is me and mark both agree is great game change might be to have a podcast just on this we're just RDS on vSphere so is on Prem on Prince fair on on printer so not VMware cloud on AWS but RDS on vSphere on Prem as you look
at cloud services where the control plane is in the cloud AWS runs RDS still but the compute is on local holes how does that kind of get your imagination or juices flowing from a marketing perspective either what monitoring challenges that's going to introduce or what new opportunities that may introduce I think that surely is the opportunity for us and that certainly it's a very important pipe to monitor on and because that's just important from those costs and performance anymore and again that's another agent wouldn't help
kind of scenario because you don't own that right particular service right you use it so a lot of thing has to be done from the networking site which our roots is based at and monitor on that and that is the critical critical piece for the entire operation certainly so you had made three points and I wanted to talk about the second one a little bit about your visibility into the applications is that a native function that you guys do are you integrating with an application an
APM tool right so we complement with the APM tool with ourself out of the box we discover application and find the dependency measure their performance primarily through the packet inspection right so it's agentless solution we use our virtual traffic tab it's all virtual and it is without implanting agent into the application server but at the virtual switch level that we get all the packets the packet inspection understanding what application actually runs on which server and it depends on which other server which other physical server or
even remote services and then measure their performance transaction volume and that's how we kind of build this whole thing together so this has been a great conversation the mark I would love to find out more about wheeler and if I was to want to do that love to be on your platform and every time we work with you guys I talk to you guys will learn a lot of new things it's been a great pleasure no thank you for having us and where can they find
you guys yeah so we're physically based in Santa Clara website wise were at wwe.com UI la Hawaiian words mean lightning the cloud it lights up the cloud so cloud gets visibility UI la Willa dot-com this isn't been a great conversation you know thanks everybody this has been great and you can find us at CTO dose comm if you'd like more information
