The AnyLogic Cloud version 2.2.x releases have given organizations that use simulation modeling an integrated environment for model management and execution. The new features in these releases upgrade simulation accessibility, security, and performance.
This blog post will look at how simulation models are managed by AnyLogic Cloud and highlight various new features up to the current 2.2.2 release.
Simulation model life cycle
The simulation model lifecycle links model development with model usage. During the lifecycle, a model will start with conceptualization and coding, in the development stage, and progress into a model that is ready for deployment — the usage stage.
In the usage stage, you can design experiments with the model, add user interfaces, setup control dashboards, and integrate the model with other software applications. AnyLogic 9, presented at WinterSim 2020, unites the two stages and allows the entire simulation model lifecycle to move to cloud and server environments.
Cloud simulation
AnyLogic Cloud lets you run models and collect outputs straight from a web browser. The public Software as a Service (SaaS) version is available for anyone to try for free at cloud.anylogic.com
Models run in the cloud and their animations render in your browser. For intensive models, this lets the server-side manage the computation and eases the load on your computer. If you are only interested in model outputs, you can run a simulation as fast as possible without animation. To do this, the AnyLogic Cloud dashboard lets you configure an experiment from your browser and run it in the cloud.
Furthermore, each run is stored in a database so that there is no need to repeat computation for previously run experiments (input configurations). A feature of AnyLogic Cloud is its scalable capacity to carry out multiple runs in parallel. This helps speed up multi-run experiments such as Monte Carlo and 2nd order Monte Carlo, Parameter Variation, and Parameter Variation with Replications. These experiments require many runs and can make use of the resources a cloud installation has to offer.
Features of Private Cloud simulation
AnyLogic Private Cloud and Private cloud Lite install on client infrastructure. They do not require connection to the external web, and clients have complete control over cloud data and internal network connections.
Private cloud products provide tools for custom branding and custom web interfaces. The Enterprise edition of AnyLogic Cloud allows an unlimited number of execution nodes in your cluster – providing effective resource utilization, load balancing, and redundancy. Cloud Lite, on the other hand, installs on just one powerful execution node. It easily installs with a few clicks. AnyLogic Cloud products are GDPR compliant and prompt to use encrypted HTTPS connections on installation.
Once set up, you can manage your cloud installation through an updated graphical user interface (GUI). The administration panel shows a general cloud health report, version number, licensing, and more:
- The Customization tab lets you customize the dashboard with logomarks, logotypes, and a favicon.
- The Services and Tasks tab shows each component status.
- Running Tasks lets you monitor cloud load, see how many execution cores are available, and terminate them on demand.
- The Nodes admin page shows how many nodes are involved in the cloud environment.
Sharing simulation models
One of the benefits of using a cloud platform is easy sharing. AnyLogic Cloud lets you upload models, manage versions, and share them. Models are accessible in web browsers through their own page, through custom interfaces, or via RESTful API (JavaScript, Python, Java). It is also possible to comment on models and download the model source file locally.
Cloud models can use a wide variety of built-in presentation options, but for those wanting to do post-processing or make use of data visualization software, it is also possible to export data. You can download experiments as Excel files that contain both the input and output data, or, to get all the results from all a model’s experiments, you can download a JSON file.
Simulation modeling cloud example video (feat. Orville and Wilbur Wright)
Learn about AnyLogic Cloud and the simulation modeling lifecycle with AnyLogic Senior Customer Engineer Gregory Monakhov in this presentation from WinterSim 2020 — including audience Q&A with Alexander Rakulenko, Head of the AnyLogic Cloud Development Team.
If you are looking to deploy simulation modeling in the cloud, please contact us and request an AnyLogic Private Cloud trial.