a project supported by

Nowadays, huge amounts of Copernicus data are freely available in the cloud, and anyone can process them using computing resources offered by various cloud providers. However, if you need to build a cost-efficient computing platform to automate the processing of satellite data in the cloud, and you want to minimize your expenses, it may not be an easy game. Indeed, it requires skilled personnel and efficient designing/maintenance approaches, and often repetitive manual operations have to be done. Further, cloud providers offer a large multitude of types of (virtual) machines, with very different computing power and very different cost per hour. Thus, the users are forced to navigate in the maze of up to hundreds types of machines for finding the most cost-efficient ones for each task.

Lockless, with the support of Copernicus Incubation Programme, developed XClouder, a smart tool that automatizes and makes cost-efficient the deployment and management of applications and services that process satellite data in the cloud. XClouder is able of minimizing the cost of the processing platform through the identification of the most cost-efficient cloud resources, and frees the users from the burden of performing onerous technical management tasks. XClouder facilitates the scale up of the users’ businesses, supporting the efficient deloying in the cloud and the continuos operation of applications and services that use Copernicus data.

XClouder is able to autonomously pursuite the following goals:

1) DISCOVER

Finding in the cloud the
most cost-efficient machine
for your computations

2) BUILD

Creating on-the-fly and dynamically resizing your pools of machines in
cloud to comply with your computation execution plan

3) RUN

Automatically fetching data
and running the computations on the right machines to meet your processing deadlines

4) SAVE COST

Drastically reducing the operating and maintenaince cost of your cloud resources

VIDEO DEMO

DEMO 1

This demo shows XClouder running the k-means algorithm over an area selected by the user. The user wants the computation to complete within 10 minutes. XClouder uses computing resources from CREODIAS. XClouder decides how many and which types of machines to create to meet the deadline and to minimize the cost. Then, it automatically creates and configures all the machines, loads all data frames, runs the computation, releases the machines and returns the results.

DEMO 2

This demo shows XClouder running the k-means algorithm on a large area selected by the user (a portion of Western Europe). The user wants the computation to complete within 25 minutes. In this demo XClouder uses resources from three different cloud platforms: CREODIAS, MUNDI and AWS. XClouder decides how many and which types of machines to create on each cloud platforms to meet the deadline and to minimize the cost. Then, it automatically creates and configures all the machines, loads all data frames, runs the computation, releases the machines and returns the results.

DEMO 3

This demo shows XClouder running a change detection algorithm over an area selected by the user. The user wants the computation to complete within 30 minutes. XClouder uses computing resources from CREODIAS. XClouder decides how many and which types of machines to create to meet the deadline and to minimize the cost. Then, it automatically creates and configures all the machines, loads all data frames, runs the computation, releases the machines and returns the results.-

Do you want to know more about XClouder? Contact us using the form below!