Eratos subscriptions and Sidecar

Accessing premium datasets, creating additional spaces and publishing to the marketplace requires an Eratos subscription. Pricing information can be found at the link below:

Eratos Subscription Pricing Information

Please refer to Eratos Sidecar Pricing Information page for more detail on Sidecar pricing information.

Usage charges

Usage charges are incurred when you use more storage and computation resources than your subscription allocation allows for.

The table below outlines the prices for additional usage. Explanations for how and when these charges are incurred

Data storage

Resource file storage

The Eratos platform supports storage of a linked graph of Resources. These Resources can store datasets, which can contain many file types, and metadata that describes the stored datasets. The size of dataset files are measured and accumulated over a month to calculate the Resource file storage usage metric. Dataset files stored for only part of the month will only be charged for that period.

Platform services that can incur this usage charge:

• Workspace (and Workspace API)
• Eratos SDK (through Gateway Node API)

Unit price

Usage metrics

• Resource File Storage (GB-mon): This metric is a measure of data storage over time. If you store 1 GB for 1 month, this will be 1 GB-mon. If you store 0.5 GB for 2 months, this is also 1 GB-mon.

Examples:

Rachel uploads 300GB of data files as Resources using the Eratos Workspace API on the 12th of April. The usage charge for this storage can be calculated as follows.

The usage metric for resource file storage can be calculated as:

• Resource file storage = (number of days of storage in April) / (days in April) x (dataset size)
• Resource file storage = ((30-11) / 30 )* 300 = 190 GB-mon
• Usage price = (Resource file storage) x (price)
• Usage price = 190 x $0.34

Usage price = $64.60

Additional resources

• Upload, Download and Manage your Files in Eratos
• Upload Dataset
• Eratos Workspace API (v1)
• Eratos SDK Documentation

Sensor data storage

The Eratos platform provides a dedicated service, Sensor Data API, for the storage and management of time series data collected from live data sources such as IoT and sensors. The Sensor Data API stores this data as observations, which can be various data types, such as scalar, vector, document, geolocation and image. This usage charge covers all of these data types except images, which are covered by the sensor image storage usage charge described below.

Platform services which can incur this usage charge

• Sensor Data API
• Data Source API

Unit prices

Usage metrics

Sensor data storage (GB): This usage metric is calculated at the end of each month from total data size of all observations (excluding image observations) stored in data streams in the Sensor Data API. The GB size is calculated from the number and type of observations stored.

Sensor data storage: images (GB): This usage metric is calculated at the end of each month from the total data size of all images stored in Sensor Data API.

Example

Tom connects a streaming sensor using an MQTT data source from the 15th of March. The sensor generates approximately 10MB of sensor data per day and is continuously adding observations.

The sensor data storage storage is measured at the end of the month so Tom's sensor will have uploaded around 150MB of data by the end of the month.

• Usage price = (sensor data storage) x (price)
• Usage price = 0.15 x $2.45
• Usage price = $0.3675

Additional resources:

• Creating Datastreams
• Connecting your Sensors to Senaps

THREDDS dataset storage

The open source THREDDS service is used to provide a geospatial catalogue with specific features supporting netCDF hosting and querying for Analysis Service workflows. The storage of data within this service is calculated at the end of the month.

Platform services which can incur this usage charge

• THREDDS

Unit price

Usage metrics

• THREDDS dataset storage (GB): The size in GB of datasets that are stored in THREDDS. This is calculated at the end of the month in GB.

Example

Jane uploads a 100GB dataset to the THREDDS catalogue using the TDM API on the 15th March. The dataset remains in place at the end of the month.

• THREDDS Dataset Storage = 100GB

The usage price for March is calculated as:

• Usage price = 100 x $0.56
• Usage price = $56

Additional resources

• Uploads a new data file to Thredds

Compute

Workflow compute

Workflow compute usage is incurred when any model, operator or workflow is executed on the platform. This will include all operators executed from Workspace Marketplace and Studio, it also includes Workflows executed using the Analysis Service and those triggered using the Senaps dashboard. The compute time is calculated based on the running time of the compute infrastructure used to host the operation. In some cases complex workflow with multiple operators will execute in parallel, the timed usage will be sum of all operator time to complete the workflow. For example, a workflow with 10 operators executing in parallel, each taking 1 minute to complete, will incur 10 minutes of the compute time usage metric.

The executing operators and models have different compute profiles in terms of CPU and memory resources. The execution time of each operator is accumulated and charged monthly.

Analysis Service also provides a restartable configuration flag which enables the use of Spot infrastructure at the expense of the task potentially being terminated by the cloud infrastructure provider and restarted.

Platform services which can incur this usage charge:

• Workspace Studio and API
• Analysis Service API
• Senaps dashboard

Unit Prices

Usage metrics

• Operator run-time: The execution time of operators executed on Eratos managed infrastructure over the given month, in hours.
• Operator CPUs: The amount of vCPU used by an individual operator
• Operator Memory GB: The amount of memory in GB allocated to an operator

Example

John runs a scheduled job on Analysis Service which takes 2 hours to complete every day. His workflow only uses one operator. He has not enabled the restartable option and he has used the compute profile 'M'. To calculate his March usage:

• CPU On-demand price = (Operator run-time) x (Operator CPUs) x $0.0938

The 'M' compute profile uses 0.5 vCPU and 1 GB memory

• CPU On-demand price = 31 days x 2 hours x 0.5 cpus x $0.0938
• CPU On-demand price = $2.9078
• Memory On-demand price = (Operator run-time) x (Operator GB) x $0.0103
• Memory On-demand price = 31 days x 2 hours x 1 GB x $0.0103
• Memory On-demand price = $0.6386
• Total usage price = CPU on-demand price + memory on-demand price
• Total usage price = $3.5464

Additional resources

• Run a Model on Eratos
• How to Build an Eratos Model
• Eratos Workspace API (v1)
• Analysis Service API

Transfers

Resource file transfers

A transfer usage charge applies for all resource file transfers. This usage charge applies to data files uploaded or downloaded externally and from compute operators executed in the Eratos platform.

Relevant services

• Workspace (and Workspace API)
• Eratos SDK

Unit price

Usage metric

• Resource file transfers (GB): This metric is accumulated during the month as resource files are uploaded or downloaded.

Example

Joe uploads a 300GB file during the month of March.

• usage price = 300 x $0.20
• usage price = $60

Glossary

vCPU - A vCPU is a virtualized representation of a CPU core, allowing multiple tasks or containers to share the underlying physical CPU resources.

THREDDS - Thematic Real-time Environmental Distributed Data Services is a data server system designed to serve and share large datasets, particularly those related to environmental and atmospheric science, including data such as weather, climate, and oceanographic models. It allows users to access, discover, and retrieve data from remote datasets over the web.

MQTT - Message Queuing Telemetry Transport, is a lightweight, publish-subscribe messaging protocol designed for resource-constrained devices and low-bandwidth, high-latency, or unreliable networks, commonly used in Internet of Things (IoT) applications.