Skip to main content

ThinkAgile CP High Availability Architecture

This technical brief describes the High-Availability (HA) Architecture of the ThinkAgile CP platform. The composable platform is designed to give you a true public cloud experience but inside your own datacenter. The ThinkAgile CPinnovation is in its software, which is designed to run on standard hardware; however, to ensure the very best customer experience, our offering includes both software and hardware. As will become clear, there are no single points of failure in ThinkAgile CP.

Download PDF

Refer to the following topics:

  • ThinkAgile CP platform overview
    The ThinkAgile CP platform has three hardware blocks - the compute block, the storage block and the interconnect switches.
  • ThinkAgile CP Abstraction
    In place of a cluster, ThinkAgile CP supports the concept of a Migration Zone. A migration zone is comprised of a set of compute nodes among which an application instance (or virtual machine) may migrate. Compute nodes (and application instances that run on these nodes) can be assigned categories and tags.
  • Hardware scaling
    ThinkAgile CP supports non-disruptive hardware upgrades and non-disruptive scaling of both compute and storage.
  • Software upgrades
    ThinkAgile CP has a software push upgrade capability, therefore, most of our upgrades to new releases of our software can be done remotely and non-disruptively. In some cases, the upgrade will require maintenance windows, but this still can be done remotely.
  • High availability support
    ThinkAgile CP provides proactive support with ThinkAgile CP Early Insights. Our health monitoring support system automatically alerts the ThinkAgile CP support team of potential issues and failures. We can often see failures and initiate replacement service requests before the customer even knows there is a problem. A chat interface allows a customer to connect directly with the support team. Support mode allows us to help the customer directly without needing a technician to go on-site. The customer can easily disable the support mode at any time.
  • Storage high availability
    The two storage controllers in a storage block operate in active-passive mode, with automatic failover when the failure of a storage controller is detected.
  • Compute node or hypervisor high availability
    Each ThinkAgile CP compute node, storage controller, and network interconnect communicates with the management portal every 60 seconds, sending regular metadata updates on statistical information, such as CPU usage, memory usage, and power usage by node and application instance.
  • Network High Availability
    The network part of the ThinkAgile CP solution consists of redundant SDN-enabled ONIE (Open Network Install Environment) interconnects, each with 48x10 GbE and 6x40 GbE ports, running the Pica8 NOS (PicOS network operating system). We call these internal interconnects, since only servers and storage internal to ThinkAgile CP can connect directly to these ports.
  • Non-disruptive migration of existing application instances
    For customers interested in non-disruptively migrating their current applications from other hypervisors, such as VMware into ThinkAgile CP, we provide the ThinkAgile CP Migration Manager.
  • Minimizing catastrophic failures in ThinkAgile CP
    Instead of clusters, we use the notion of a migration zone (MZ) in ThinkAgile CP. A migration zone is a set of compute nodes among which an application (VM) may migrate.