CELAR - Automatic, multi-grained elasticity-provisioning for the Cloud

What user need or pain point is your project addressing?

CELAR targets both users that use cloud infrastructures to deploy and operate their applications and cloud IaaS operators that want to offer extra features to their customers.

Cloud users find it hard to detect, deploy utilize the correct amount of needed resources in order to offer services to their end users, whereas IaaS operators would like to offer more advanced services than simple resource provisioning. Over-provisioning is costly, whereas under-provisioning is not efficient, especially in a dynamic setting where the workload changes constantly during time. Internet serving applications built on top of IaaS clouds such as sites, services, etc, typically encounter these problems.

CELAR targets the problem by offering a fully automated fine grained elasticity mechanism that can be utilized by cloud applications to adjust their resources according to user-defined policies when resources need changing. The project’s results are open-sourced and offered through a simple installation package that can be deployed on top of any IaaS cloud.

Find out more


Open Source software available on

AppHub Directory

Project Start: 
Project End: 

How will your solution/service benefit the end-user?

The CELAR software can be utilized by end users to deploy, manage, monitor and elastically scale their applications on top of any openstack compliant cloud stack, such as Amazon, Rackspace, etc. Users will be able to define complex elasticity rules (i.e., to dictate when and how the system will scale its resources) utilizing CELAR’s powerful decision making module by taking into account application specific KPIs or generic monitoring metrics.

Apart from legacy coarse-grained elasticity actions, such as virtual machine additions or removals that are already supported by other cloud vendors, CELAR offers fine-grained elasticity actions (i.e., specific resource scaling such as CPUs, RAM, hard disc space, etc), in order to take more informed decisions and meet user defined SLAs. All these features can be deployed in any openstack compliant IaaS cloud, thus avoiding vendor lock-in in a public cloud setting, or enabling in-house elasticity management in a private cloud setting.