As mentioned yesterday here and here, we are very proud of Citrix CloudPlatform and the ability to manage both cloud native scale-out application workloads as well as traditional scale-up application workloads. Let’s break that statement down into more detail.  We’ll start with cloud native scale-out application workloads.

Cloud Native Scale-Out Applications

Let’s take a look at the underlying infrastructure commonly used to support a cloud native scale-out application. In this scenario, the infrastructure is often built using commodity hardware that is not designed for failure or high availability. Any node could fail at any time. Resilience is not built into the infrastructure; it is built into the application. On top of this commodity hardware software defined services are often used such as Software Defined Networking (SDN) and Software Defined Storage (SDS), typically Object Storage. In CloudPlatform we wrap all of this into an Availability Zone as shown below:

New in CloudPlatform 4.2 is the ability to use Object Storage for volume snapshots as well as image templates. In addition, Object Storage can be shared between availability zones for greater operations efficiency.

An additional feature in CloudPlatform 4.2 for cloud native applications is the concept of Regions. CloudPlatform has introduced regions as an additional logical hierarchy element above zones. Regions may contain more than one availability zone. Each region will have a dedicated management infrastructure to allow for geographic disbursement and disaster recovery scenarios.  Regions also allow for additional scalability, higher availability, as well as increased compliance. In addition, object storage can be configured to replicate across regions.

Lastly, new to CloudPlatform 4.2 for cloud native applications we have a number of enhancements to the daily operations of storage and networking:

  • Dynamic Primary Storage Provisioning provides storage resource flexibility within cloud environments by automating common storage management tasks and streamlining the provisioning process
  • Advanced cloud networking enabling cloud operators to create advanced cloud networking configurations and Network-as-a-Service offerings including new comprehensive portable IP capabilities, advanced load balancing scenarios (GSLB , Advanced zones) and AWS like Health Checks to ensure application availability, dedicated network resources and VLANS.

Traditional Scale-Up Applications 

Next up we have traditional scale-up applications. Let’s start by examining the underlying infrastructure as we did above. The infrastructure typically used for this type of application is very different. Resiliency IS built into the infrastructure; it IS NOT built into the application. As you can see, the infrastructure requirements for a traditional application are vastly different from a cloud native application. In this case the underlying hardware is built for redundancy and resiliency and the application expects this environment to maintain operations at all times. A hypervisor management server is often deployed, VLAN’s for Layer 2 networking, hypervisors are grouped into clusters, and SAN storage is utilized. In CloudPlatform we wrap all of this into an Availability Zone as follows:

While this topology is not new, CloudPlatform 4.2 brings some significant enhancements to this architecture that are worth talking about.

The first major enhancement to CloudPlatform 4.2 is better support for multi-tier (n-teir) applications and virtual private clouds (VPC). With n-tier support we are able to segregate and firewall as needed to provide better security in between each access layer. This functionality when combined with existing load balancing, and anti-affinity/affinity rules allow for an amazing amount of flexibility, reliability, and compliance for n-tier applications, all in a secure environment provided by VPC. Here is an example of a typical three-tier application providing web, application, and database tiers:

There are many situations (compliance, physical security, departmental budgets, performance, licensing) where dedicated resources may be required to support traditional application workloads. New in CloudPlatform 4.2 is the ability to offer dedicated resources at the host, cluster, and pod level. CloudPlatform supports dedicated CPU’s, memory, network IP’s, and disks

Lastly we have a few additional enhancements to existing CloudPlatform features that are worth mentioning in support of traditional applications:

  • VMware integrations for running enterprise application workloads in the cloud and leveraging your existing VMware investments, configurations and expertise. CloudPlatform supports dynamic scaling of CPU and memory on running VMs as well as dvSwitch and PVLAN networking support for VMware switching and VM isolation on the same network. In addition, CloudPlatform enables VMware vSphere features like virtual hardware hot add, vMotion, VMware HA & DRS, CPU & RAM overcommit, VM snapshots, storage migration and zone-wide block storage.
  • Enterprise grade networking for supporting the performance, scale and SLA requirements of the most demanding production application workloads. CloudPlatform direct integrates with external enterprise class Application Delivery Controllers (ADC) and networking devices including Citrix NetScaler, Cisco hardware and software (UCS, Nexus 1000v, ASA1000v), Juniper firewalls and VMware Distributed Virtual Switch. CloudPlatform also offers built-in support for DHCP, NAT, Gateway, VPN, Firewall, Port Forwarding, VLAN, Elastic IP, routing and load balancing.

As you can see Citrix CloudPlatform 4.2 has many new features and enhancements to allow management of BOTH your cloud native as well as traditional application workloads.

Here are some great NEW resources for more information.  Take a Look!