The Enterprise File Fabric provides a global file system through which users and applications can transparently access corporate file systems and object storage no matter where they or the storage is located. To provide universal browsing, search, and other services the File Fabric scans and smartly indexes file and object metadata. In this way, no data is copied or replicated, and the metadata is used to build out the high-performance global file system.
Unlike other global file system solutions, the File Fabric does not prevent direct access to the storage – what we call bi-model access. Data can be accessed and updated concurrently through the File Fabric and directly through native file and object-APIs. When objects are updated through native APIs the File Fabric metadata should also be updated to keep indexes and other metadata synchronized.
An earlier blog article outlined three general approaches for keeping metadata up to date, including a specific example using Amazon S3 object notifications. See Using AWS Lambda to Automatically Sync Metadata With S3 Events. This article describes a similar approach we have developed for file systems using the same framework we call Jibe.by