CliQr's Runtime Performance on the Cloud
Because clouds are elastic and dynamic environments, they require both initial and continued management to optimize application provisioning and scaling, leading to advantages that include business agility, enhanced performance/throughput, and economics. The CliQr CloudCenter™ platform provides the necessary visibility and flexible control to optimize the price/performance benefits of migrating an application to the cloud. CliQr's performance technologies and features are designed to add value at every stage of deploying and managing an application on the cloud—from dynamic provisioning to auto-scaling to tearing down the application and infrastructure resources if not in use.
Optimizing On-boarding of an Application to the Cloud
Rapid On-boarding—Move any application onto and between any cloud without modification or the need to create cloud-specific scripts or virtual machine (VM) images.
App Orchestration and profiles—CliQr's CloudCenter provides packaged Application profiles that support the majority of application types, including big data, big compute (e.g., parallel, cluster, Hadoop, etc.), batch, web apps, and enterprise suites. The App Orchestration profiles inherently capture the most optimal steps to deploy the different application types. As a component of CliQr's CloudCenter Manager environment, these packaged Application profiles also guide the IT end user through a characterization of the application's rules and behaviors, including primitives related to the application's requirements for compute, network, storage, and security, as well as infrastructure support. The App Orchestration profiles are utilized by the App Orchestrator to invoke these rules and behaviors by coordinating with the CloudCenter Orchestrator, ensuring optimal application performance.
Cloud Orchestration—CliQr's CloudCenter Orchestrator technology encapsulates best-practice rules and behaviors for both generic cloud infrastructures and specific cloud providers. Working with the App Orchestrator, the Cloud Orchestrator transparently and automatically selects the optimal cloud candidates (cloud(s) that can support application requirements), configurations (e.g., cluster size, memory, etc.), and general best-practice rules and behaviors, including compute, network, storage, and security. The Cloud Orchestrator dynamically provisions the application onto the cloud, enabling the application to get the best use of available cloud resources while adding no performance overhead.
Optimizing Runtime Performance on the Cloud
Once the application is on-boarded, CliQr offers a variety of features and management environments, many of which are used in the deployment process. These tools also add value during runtime management optimization.
In addition to optimizing performance, CliQr provisions applications on clouds to an optimal configuration and infrastructure resource implementation, without the need to modify the application. CliQr works with an application's existing distributed and scaling architecture to enable it to run efficiently and optimally on a cloud environment. Along with optimizing runtime management, CliQr continually tunes and manages the application runtime to take advantage of the very best cloud scaling performance and economics.
App and Cloud Orchestration for Runtime Optimization—The interaction and coordination of the CliQr App Orchestrator and Cloud Orchestrator continue to play an important role in cloud resource optimization, adding value beyond deployment. These two CliQr product components intelligently use policies to either manually or automatically scale-up and scale-down cloud resources as needed to deliver both price and performance based on the application's unique criteria. CliQr can utilize policies to manually or automatically scale any scalable application to dynamically spin-up more infrastructure resources and add them to the application cluster in order to enhance performance, or intelligently scale-down if extra resources are not needed. CliQr's scale-down process further balances price and performance by maintaining "hot availability" of these discontinued cloud resources after use and up to the limit of the committed billing period, making new capacity available immediately with no spin-up delay and no additional cost.
Benchmarking for Runtime Optimization—Benchmarking helps businesses ensure that they to continue to get the best of the cloud. Policies can be established to alert the user of changes in benchmarking results and then manually or automatically switch the application to the best cloud based on various criteria, including price and performance. For example, high-value design applications might be willing to pay a premium for ultrafast performance, while monthly batch billing systems might want to find the lowest-price cloud services available.
Policy-Driven Management —The CliQr CloudCenter dashboard provides the IT end user with a complete policy management environment that both guides and enables the development and management of polices in order to manually or automatically align cloud execution with price, performance, and other criteria. Some policies examples include:
- Storage Management—Set storage management and use policies such as auto-resizing storage or automatically shutting down storage when not in use to avoid storage runtime costs.
- VM Management—Enable/disable VM pooling options such as "hot availability" to maintain a standby pool of VMs.
- Sync Management—Set up scheduled data syncs between on-premises and cloud storage.
- Security—Enable/disable security features such as storage encryption and disable key downloads.
- Application Lifecycle —Application lifecycle policies center around application deployments and runtime management. Policies include automatically terminating deployments if certain criteria have been met (such as a time limit), auto-scaling, auto-bursting, kill rogue instances, etc.
- Cloud-bursting - With CliQr's cloud-bursting capabilities, applications can move from one cloud environment to another based on schedule or performance-based polices. Cloud-bursting examples include:
- Bursting from a private or public cloud to another private or public cloud based on need for additional capacity
- Bursting specific scalable tiers of multi-tier applications, such as web apps, to improve performance
- Bursting new instances of applications to new/multiple clouds
- Application- or job-specific busting to place specific application jobs on specific clouds; for example:
- Mission-critical jobs on a high-performance cloud configuration
- Commodity batch jobs on low-cost clouds
- Staging of dev/test vs. production on different clouds
- Price/Performance—Run or burst based on price, performance, or price-performance benchmarking results.
- Tag—Dev/Test—Application deployments and clouds can be tagged with keywords so that an application tagged with a certain keyword will be run only on a cloud with the same keyword. As an example, a policy can be set so a deployment tagged with "dev" may run only on a "dev" cloud, while a "production" run will be launched only on a "production" cloud.
- Cost Management—Do not run any or all applications if a certain cost or budget limit has been reached.
- Scaling—CliQr provides complete control over application scaling. A user can scale an application manually or set up defined policies in the CliQr Policy Management Environment to auto-scale applications based of variables, including workload/performance, schedule, and costs. CliQr allows users to create minimum/maximum scaling range limits to achieve the appropriate balance of price and performance for the application.