1. Welcome to InfinStor¶
InfinStor is a multi cloud Machine Learning Platform with an Enterprise grade MLflow service at its core. InfinStor includes storage enhancements for cloud storage snapshots ( InfinSnap and InfinSlice). The InfinStor Compute Engine ICE is capable of running complex Machine Learning pipelines across different clouds and different regions of a cloud.
InfinStor may be consumed in the following ways:
- SaaS mode: The InfinStor service components reside in our AWS account and the data/compute resides in your Cloud accounts.
- Enterprise Licence: Every aspect of InfinStor operates in your own Cloud account(s)
InfinStor provides full enterprise support for both modes of operation.
Note that in the case of the Enterprise License option, references to
mlflowui.infinstor.com, etc. should be replaced with your own instance of the service. For example, if the service is installed in your own AWS account and available at subdomain
ai.acme.com, then replace
This documentation website is organized as two sections, the Admin Guide and the Data Scientist Guide. Use the Navigation bar of this website to get quick access to these guides.
1.1. Admin Guide¶
This section of the documentation is for adminstrators responsible for making the InfinStor service available to Data Scientists. This includes activities such as:
- Signing up for the InfinStor hosted service, or alternatively, installing the InfinStor service in your own AWS account
- Adding, deleting and managing Data Scientist Users
- Managing the MLflow service included by InfinStor
- Managing the storage enhancements offered by InfinStor
- Managing the compute capabilities offered by InfinStor
1.2. Data Scientist Guide¶
This section of the documentation is for Data Scientists interested in using the InfinStor platform. It includes documentation on:
- Authenticating to the system
- Using the MLflow service
- Using the storage enhancements of the InfinStor platform
- Using the Compute enhancements included in InfinStor