Cirba for workload control and
routing in hybrid cloud
Cirba scientifically balances infrastructure supply and application demand – creating a demand-driven approach to virtual and cloud operations that maximizes efficiency and cost savings, while reducing risk.
Cirba provides support for workloads running on a variety of physical platforms, virtualization platforms such as VMware®, Microsoft® Hyper-V®, KVM and IBM® PowerVM® and public clouds including Microsoft® Azure™, Amazon Web Services (AWS) and IBM® SoftLayer®.
Explore our solution by use case
Demand Management for Real-Time Automation of Workload Routing
Cirba enables organizations to gain a complete view of new demand, whether capturing new requests through integrations to a third-party self-service portal or by analyzing the transformation. This visibility enables organizations to accurately understand what resources are required to serve the incoming demand and where new workloads should go.
Model New Demand
Many organizations rely on vendor recommendations when it comes to determining application requests. Sometimes this is the only place to start, but most of the time this approach leads to highly inefficient infrastructure. There are other approaches that can to ensure a low risk efficient deployment. Cirba provides options for modeling the requirements of new demand – whether providing workload profiles of similar applications to leverage for initial analysis, or assessing new workloads staged in a dev/test environment or transforming from physical to virtual environments.
When it comes to platform migration, physical to virtual transformation or virtual environment refresh, Cirba's analytics have set the bar in transformation planning, enabling organizations to make moves faster, with less risk and significantly lower infrastructure cost. Only Cirba's analytics determine which workloads can be transformed, how big to make the VMs (or which cloud containers are the right fit), which workloads can go together on the same hosts from a business and operational policy perspective, as well as how to optimally fit workloads together on physical hardware to optimize density and minimize infrastructure costs. This analysis provides a comprehensive view of potential transformation candidates, as well as the workload placements, resource, and infrastructure requirements to achieve the optimal balance between financial objectives, risk tolerance, and performance.
Real Time, Intelligent Routing – Determine the best environment for your workloads
IT organizations are now faced with a new challenge – they need to be able to determine where to host new workloads. However, most rely on spreadsheets and are ill-equipped to determine which internal environment is suitable, or whether the workload should be hosted externally. Others may look to cloud management platforms that have rudimentary routing logic like round-robin placement options. Neither of these approaches can ensure that workloads get access to the resources they need or that placements comply with policies.
This challenge is even greater in hybrid cloud environments as organizations want to ensure workloads get access to the required resources at the right price without compromising on compliance, security and performance.
Cirba enables real-time, automated selection of the best execution venue for workloads with its Reservation Console. The Console examines the requirements of a workload and evaluates the suitability of target environments based on relative cost, occupancy levels, and fit for purpose criteria such as operating system, software licensing requirements, storage requirements, and operating policies.
In 2016, Cirba will also release for automating routing decisions for hybrid cloud infrastructure including Microsoft® Azure™, Amazon Web Services (AWS) and IBM® SoftLayer®. Click here to learn more about routing for hybrid cloud from Cirba.
Learn more about automating workload routing.
Guarantee Resource Availability with Capacity Reservations
Most requests for new workload placements do not require immediate fulfillment. For these requests, users don't need rapid access to capacity, but instead benefit from a reservation that holds the required resources in the targeted host environment until they are required. Capacity reservations not only ensure availability of the resources, but they also are a key input to enable accurate forecasting. This requires a reservation system that not only determines the optimal environment for a workload (see above), but also takes a forward looking view and holds the capacity for that time frame. Through its Reservation Console, Cirba:
- Profiles all inbound workload placement requests (from self-service portals, transformation projects, etc.) to determine their business requirements, applicable policies, and capacity requirements.
- Determines where the workload should go and whether there is capacity available to fulfill the request on the planned
- Reserves the capacity for that workload.
- If there isn't enough capacity available, determines what action is required to fulfill the request, such as adding a new host, increasing memory or CPU.
Infrastructure Supply Optimization through Densification
Most organizations today combat risk in virtualized infrastructure by significantly over-provisioning hardware. Excess capacity is a cost that most feel they have no choice but to incur because of the complexity of these new school environments. But this inefficiency can be avoided by optimizing workload placements and right-sizing VM allocations to combat both risk and capacity waste.
Most organizations today combat performance risk by significantly over-provisioning hardware and over-sizing VMs – both VMs hosted in internal and public clouds. Allocating excessive resources is a cost that most feel they have no choice but to incur because of the complexity of managing workloads in these new school environments. But this inefficiency can be avoided by optimizing workload placements and right-sizing VM allocations to combat both risk and waste. Cirba does this for internal and hybrid environments.
Optimize Placements to Densify Infrastructure
Cirba optimizes VM placements in virtual and cloud environments with its multi-dimensional analysis that considers all technical, business, and utilization factors and constraints. Cirba accurately determines initial placements, and will also continuously adjust these placements across physical servers to rebalance workloads, avoid hot spots, and "densify" capacity to free up space for new instances.
Only Cirba can place workloads considering their patterns, personalities, profiles, and applicable policies in order to safely maximize density. With Cirba, wasted capacity is significantly reduced and organizations save an average of 33% on hardware.
As environments grow and as workloads change, what was once optimal may no longer be. Cirba optimizes virtual and cloud infrastructure by rebalancing and right-sizing workloads as conditions change. Unlike a load balancer, Cirba's placements take a broader perspective and longer term view to avoid volatility and the hidden costs of unnecessary motioning.
Receive Daily Actions to Reduce Operational Risks
Cirba's Control Console provides a simple way of seeing where risks lie in an environment – any entity in the "Too Little Infrastructure" or red zone is at risk according to workload requirements or policies. Cirba also provides specific, detailed recommended actions that can be passed through to third party systems to automate remediation in accordance with change management policies for:
- Under-allocated VMs - Cirba enables you to right-size VMs whether you are using a standard instance catalog for clouds or custom VM allocations.
- Host and Environment Imbalances - VM rebalancing ensures that each workload gets the resources it requires to avoid potential performance issues.
- Inadequate capacity at the cluster and environment levels - Cirba also tells you when a cluster or environment is running the risk of not having enough capacity.
Receive Daily Actions to Maximize VM Density and Efficiency
Alongside recommended actions to reduce risk, Cirba produces daily actions that enable you to increase efficiency by reclaiming resources and optimizing use of existing capacity such as:
- Over-allocated VMs - Over-sized VMs are a common area of waste, particularly in cloud environments where self-service models allow users to determine allocations. Cirba ensures that service levels are met according to policies and specifies exactly how much resource each workload requires so you can confidently reclaim and reallocate resources.
- Inefficient workload placements - Cirba analytics determines the precise workload placements required to safely "densify" server capacity and fit the workloads onto the minimum amount of infrastructure.
Forecast Capacity Requirements
Cirba provides an accurate view of when a capacity shortfall will arise. Looking into the future, Cirba indicates when storage resources, environments, clusters, and hosts will experience shortfalls and helps you determine when to take action. Cirba also enables you to determine how many VMs you can fit into an environment to "use up" spare capacity and make better use of your infrastructure.
Cirba also provides Universal Sensors to analyze additional resources and quickly reveal, for example, which are low on capacity, which are appropriately utilized, and those that have excess capacity. Users can also select a resource, such as a datastore, to identify which Hosts and VMs are related to it.
Optimize Resource Allocations
VMs must be sized properly in order to service the applications within them. In cloud environments, this means selecting the most suitable (and economical) instance size from a cloud catalog. Although most clouds require the end user to select this, they often get it wrong because they may not know what their demands will be (or may not know how to express their needs in the language of Gigahertz and Gigabytes). Cirba optimizes VM sizing for custom allocations and cloud containers.
Workload Routing and Control for Hybrid Cloud
As capacity becomes commoditized and public cloud options more reliable and secure, hybrid cloud is inevitable for most organizations. However, determining how to leverage public cloud resources in conjunction with internal infrastructure is extremely challenging. Not only do detailed application requirements need to be assessed against the security, cost, and technical capabilities of available public clouds, these decisions need made in the context of how to best leverage all available resources - internal and external. Without analytics driving application hosting decisions, organizations cannot automate the process nor can they effectively determine how to meet application requirements without risk or excessive cost.
Decision Control for Routing to Public Cloud
Cirba provides the necessary decision control point for determining where applications can safely run in hybrid environments. Cirba evaluates Microsoft® Azure™, Amazon Web Services (AWS) and IBM® SoftLayer® alongside internal infrastructures to determine the best execution venue for applications.
Reservation Console hybrid routing capabilities
Cirba determines the best execution venue for workloads by looking at detailed workload requirements against the available resources, technical capabilities, cost and fit for purpose of all available infrastructures. For more information, see details of our Reservation Console.
Management Control for Workloads in Public Cloud
Cirba provides visibility into workloads operating in evaluate Microsoft® Azure™, Amazon Web Services (AWS) and IBM® SoftLayer® providing a single point of control for all an organization's workloads.
Visibility into workloads in public clouds
- Visibility across all workloads in hybrid cloud
Cirba provides discovery that enables you to find all the workloads in AWS and IBM SoftLayer, ensuring no workload goes unmanaged! In addition, Cirba provides visibility across all your workloads to show you where all your workloads live and operate.
- Right-sizing for workloads in public cloud
Cirba analyzes workload performance and sizing characteristics to determine the optimal sizing decision from the host environment's catalogue. This ensures workloads get the right amount of resource, but not too much to keep costs under control.
- Routing analysis for repatriation of workloads
In its analysis of workloads operating in the cloud, users may determine that it is best to bring a workload back inhouse. Cirba enable workload data gathered from operations in public clouds to be leveraged for analysis of the optimal internal execution venue for those workloads.
Control for Physical Storage and Datastores
Cirba's analytics optimize VM placements considering all the required factors including physical storage (including NetApp®) and datastore requirements. This enables organizations to balance demand across storage assets and ensure the right match between workload requirements and storage tiers. Cirba enables organizations to:
Automate VM routing involving complex storage decisions
Cirba's analytics consider all workload requirements when determining the optimal host environment for a workload, even complex storage needs such as multi-disk requirements storage, tiering, and interconnect bandwidth usage. This ensures that your workloads get access to the resources required without over-specifying and over-allocating.
VM placements that balance demand across storage resources
Cirba's routing decisions can also help free up stranded storage resources by placing VMs appropriately to use up capacity or avoid bottlenecks while considering all the other demands on resources.
Visibility into Storage Status and Connected VMs
Through Universal Sensors in the Cirba Control Console, Cirba provides visibility into physical storage and datastore health and the VMs connected to those resources. The Control Console's spectrum visualization reveals, which storage resources have excess capacity, which are appropriately utilized and those that are low on capacity so that users can understand which VMs may be at risk of a shortfall.
Accurate storage forecasting based on the demand pipeline
Through capacity reservations that include physical storage reservations, Cirba can provide a comprehensive picture of demand on storage assets today, and actual planned demand in the future. This enables an accurate view into when storage resources are going to run out so that purchasing plans can be confidently made without over-provisioning.
Watch this video to learn how Cirba optimizes EMC Storage through intelligent vm placements and reservations
Controlling Software License Costs with Placement
Data center class software licensing provides the opportunity to license an entire physical host server and run an unlimited number of instances on that machine. To take advantage of this requires VM placements that minimize host licensing requirements and maximize density.
When factoring software licensing considerations into Cirba's placement decisions, organizations can save an average of 55% per processor licensing packages such as Microsoft® Windows® Server - Data Center Edition, Microsoft® SQL® Server, IBM® WebSphere®, or Oracle® Database, Oracle® Weblogic®.
Cirba's Software License Control module is an add-on to our award-winning Control Console that:
- Reduces the number of processors/hosts requiring licenses.
By maximizing VM density and isolating licensed VMs from those not requiring the licenses.
- Avoids sprawl by containing the licensed VMs on the licensed physical servers.
By restricting placements of licensed VMs to the designated physical serverss ensuring ongoing licensing compliance and efficiency.
- Leverages policies to program VMware DRS rules to ensure ongoing compliance automatically.
Any new VMs entering an environment are placed on appropriately licensed servers, while taking into account the pipeline of demand to understand potential growth.
Control for Software-Defined Networking
Software-Defined Networking (SDN) solutions like VMware NSX® and Cisco® ACI® enable organizations to rapidly define networks in software which provides tremendous advantages in terms of agility, security and efficiency. These technologies, however, also create a new management challenge. With SDN, VxLANs can span across clusters introducing new-found freedom in terms of VM mobility. Cirba's intelligent analytics for controlling VM placements become critical in these environments as factors such as latency, storage connectivity, compliance and software licensing all need to be considered when determining where VMs should go.
Optimize Infrastructure on Enterprise-wide Through Intelligent Routing and Re-Routing
Tearing down the traditional barriers that determined VM placements opens up the opportunity to play a "bigger game of Tetris®" across enterprise environments. Cirba's analytics can analyze a group of workloads or entire applications to optimize placements considering all required factors while also maximizing efficiency. Cirba optimizes initial placements and also re-routes workloads to new environments when current placements have become sub-optimal or in order to further optimize infrastructure by reducing licensing costs.
Network Function Virtualization
Network Function Virtualization (NFV) enables specialized networking functions, such as firewalls and load balancers, to be implemented in software instead of hardware. NFV still requires hardware to run, but you will be replacing dedicated networking gear with cabinets full of servers. Just like any other group of VMs, placement is a critical determinant of how efficient and how performant the overall environment is.
Optimize Placements According to Unique Requirements
NFV workloads typically have specific hardware or feature requirements or characteristics that require special consideration when determining where they should be hosted such as minimum processor core speeds, hardware cryptography and hardware packet acceleration.
Cirba can account for these requirements when routing and placing these VMs, for example:
- NFV VMs may be part of a larger group of VMs that form a multi-tier application, or may be grouped to create pure network services such as VPN endpoints or consumer/subscriber services.
- Virtual firewalls must be placed in an environment and on a specific host that has sufficient I/O bandwidth to service the workload. VM placements must consider other workloads in the environment so as to not create resource contention.
- An NFV component may have packet acceleration (DPDK) and cryptography requirements that require that the VM to be placed on a system with these capabilities. If it is part of a broader application group, then the entire group will be intelligently placed with this requirement in mind.
Manage and Control NFV Groups by Policy
Cirba policies can be used to manage and control different NFV components as different groups. For example, Cisco ACI supports the notion of Endpoint Groups (EPGs), which share a common policy (such as all load balancers having specific ports open). These EPGs can be mirrored in Cirba, and separate policies applied to them, such as de-concentration and allocation/scaling policies. This enables a more precise level of control over the components.