MiniWebinar: Overview of NSX Advanced Load Balancer

hi and welcome to this bite-size session on VMware and as X advanced load balancer my name is Federico duct I am a senior solution engineer from the network and security business unit at VMware operating out of the B locks area in today's enterprises there is a tremendous amount of stress on IT in general the main driver is all about time to market or in other words a continuous battle to keep the time it takes to bring new applications or even changes to existing applications as short as possible IT has to cope with an explosion of containerized applications and the many different private or public cloud infrastructures to host them all and they have to do all of that with cost efficiency in mind so of course IT is looking into ways of automating these tasks we already solved the riddle of virtualizing compute resources a long time ago in recent years we've added storage and network virtualization in the mix all of which has already helped many organizations in achieving a shorter time to market but a critical component was missing or at least this component doesn't benefit yet from the high level of automation as these others do while it's great to be able to spin up entire networks storage and resources in record time providing a secure redundant and highly scalable form of connectivity to the end-users who after all are consuming these applications remains the primary responsibility of a load balancer or also often referred to as an application delivery controller with the increase in IT demands organizations are looking into a scalable agile and flexible solution not really pervasive elements in today's load balancers and as such load balancing remains a key blocker for digital transformation traditionally the load balancing question has always been answered with hardware appliances while they are in most cases feature-rich they remain physical boxes even in their virtualized form they are often deployed in pairs as a single front end of an application since physical boxes by definition have a set amount of resources available they are typically over provisioned to accommodate any future peaks in traffic that might occur this results in a poor use of active resources where the active node is waiting for that peak to occur and the passive node is waiting for the first one to fail in itself not a terrible problem but it becomes quite apparent when organizations start deploying different sets of pairs all around the data center they do this to accommodate the load balancing needs for different tenants or different business units that require that kind of functionality of their own this however is an example of poor resource planning and results in different isolated islands of capacity and then at the time of peak traffic these silos have no means of gathering resources from the other islands saying they are all deployed as standalone pairs if the central management of these appliances like any form of role based access control or multi-tenancy support all of these appliances need to be managed and configured individually which doesn't quite help with the automation challenge or even a simple thing like bulk upgrades traditional load balancers were designed during a time where networks and the workload state support were pretty much static nowadays we have to deal with different forms of private and public clouds as well as different form factors like containers in other words the traditional load balancers were not designed with these modern infrastructures in mind when we think about modern infrastructures we often refer to cloud like environments surely the public cloud has solved a load balancing challenge right not really while the cloud-based solutions offer a great deal of automation and elasticity they typically lack the enterprise great features that organizations have grown accustomed to so it again becomes a balancing act to choose between a feature-rich solution or an elastic solution and not only that but different public clouds imply a different set of operational tools you can't just pick up the AWS load balancer and have it run on the Google cloud for instance so can I have the best of both worlds yes you can introducing the NSX advanced load balancer formerly known as RV with the NSX advanced load balancer we have adopted the same software-defined philosophy as we did with for instance in Essex this has resulted in a unique architecture meaning we've separated a control plane away from the data plane a control plane is typically the brains of a solution a place where you define how the solution should operate or simply put what the policy is as a brain it's also capable of learning what the data plane encounters the data plane on the other hand simply performs as instructed by the control plane when we apply this logic to the nsx advanced load balancer this means you have a central management to the controller in charge of the full lifecycle management of different instances running across different infrastructures we call these instances the service engines and they make up the data plane of the solution these service engines come in different form factors according to the clout in which they operate all while being centrally governed by the controller who in turn regulates the resources required to have the best load balancing experience for instance adding service engines when traffic ramps up or shutting them down when business is a bit slower this results in an elastic management of load balancing services across different infrastructures this in itself is already an enormous differentiator when compared to traditional load balancers the service engines can be ran on traditional platforms like vSphere but just as easily in public clouds or in combination with container management platforms typically based around kubernetes and obviously we don't stop there because of its function in the network remember a load balancer sits in front of an application and typically provides redundant connectivity for the users to access that application we are actually in an ideal place to provide analytics as to how your application is performing for instance how many milliseconds does it take for a user to access the application how long does it take for a web page to load as part of that application how many connections per second does this application need to handle under normal circumstances or under heavy load the service engines are uniquely positioned to deliver that level of detail and it's not only about connectivity it's also very much about security we can just as easily determine whether an application is under attack or not simply by leveraging the built-in distributed denial-of-service and web application capabilities and because of our high level of automation as a result of those analytics we can automatically spin up more service engines if needed or face them out when the dust settles so what Famer is bringing to the market with the NSX advanced load balancer is a consistent Enterprise great application service and Web Application Firewall across multi cloud environments all of the different services you see depicted on the left are now applicable to all the different cloud environments you see on the right and of course the NSX is vaunt load balancer is best friends with the already existing solutions from VMware for instance virtual desktop environments can leverage the power of load balancing and single sign-on without the need of third-party brokers and a-six environments can be enriched with the additional security of a Web Application Firewall just to name a few this all ties in neatly with VMware's vision to be able to deliver any application on any cloud with a single platform feature wise the nsx advanced load balancer comes with all enterprise great features you would expect from a modern load balancer things like local or global load balancing as well as content switching and caching or compression with one big difference compared to traditional hardware appliances and that is our true auto scaling capabilities from a security point of view we will write a web application firewall in this tree or denial of service protection and layer 7 based rules and policies as well as more traditional features like as L termination rate limiting and single sign-on authentication the more you move to the right the more we start diversifying ourselves we take the help and response of the applications very serious the more you know about your applications the better you can connect and protect them for instance we visualize collected data back to the administrators by providing an application map service health scores and general application performance which makes proactive scaling and general troubleshooting extremely easy on a far right we found our platform capabilities a central management which is capable of providing all of these insights from different cloud environments as well as deploying service engines on different clouds or tying into different own primer of Prem automation solutions it also supports a self-service portal with complete multi-tenancy so no more of those islands of capacity in our data centers I hope this gave you a good understanding of what the NSX advanced load balancer can deliver next up is a quick demo in which I demonstrate the ease of deployment and ease of configuration just to show you how easy it is to get started with the nsx advanced load balancer in this example we are going to deploy the nsx advanced load balancer in a vSphere environment so the first thing we do is log on onto vSphere and start deploying the controller blowing the controller is as easy as deploying a template so that's exactly what we're going to do here just point to the right file hosting our controller we give it a relevant name we choose on which host controller will end up we review our settings we define our storage profile and very important we define on which network the management interface of our controller should sit in my case it's a management poor group and then we just need to provide the IP details of our controller and then we click next we review our settings and a new task is spawning deploying our controller while we wait for the task to finish we might as well take a look at the two servers we're trying to load balance to so the first one is called wind web zero one and as a specific IP address please take note of the IP address in the 50 range and winweb zero two in the same range of course with a different IP address once the task has finished we can boot up our controller and as soon as it's booted we can go to the IP address we specified earlier we need to put an admin account in place with a proper password I'm not going to put an email address for now I also need to configure some DNS settings and an NTP server and here we see an overview immediately of all the environments in which we can deploy our service engines in this example I'm going to opt for the vSphere environment so I need to provide the proper credentials as soon as that connection is made the controller reads out the vCenter environment and displays a list of all the data centers in my example it's only one so I'm going to opt for that one and I'm also going to use static IP addresses to deploy the service engines I also need to provide a management network in which the management interfaces of my service engines will end up nice case I'm going to use the same as the controller and I'm going to put a static pool in place I'm presented with the virtual service interface so all that's left is to define my back-end resources so my web server pull and link a virtual service to it this can all be done by simply clicking on create virtual service and basic setup in which we need to provide a name for our service a virtual IP address select our back-end servers in this case our Windows web servers the only thing left is to go to our pool and set it on the algorithm we wish to deploy we click Save the service engines automatically get deployed in our environment and as soon as they're up and reporting back to the controller as being available our virtual service will light up green which means that we should be able to load balance to our back-end services in the round robin fashion that we decided upon earlier with that we've come to the end of this quick overview on the nsx advanced load balancer a thank you for your time you

Channel: VMware IT Academy

Views: 550

Keywords: #vmware, #nsx, #load, #balancer, #demo, #it

Id: oa_YZb9blPQ

Length: 14min 50sec (890 seconds)

Published: Thu Jun 04 2020