Storing and publishing contents of a content store

Abstract

Aspects are disclosed for publishing contents of a content store. A storage operation is performed, and a completion of the storage operation is detected. Here, the storage operation redundantly stores contents of a content set onto instances associated with a content store. The contents stored in the instances are then published in response to the completion of the storage operation. In another aspect, a dataset table is generated to facilitate storing contents of a content set, which include payload and metadata. The payload is stored onto a payload table, and the metadata is stored onto a metadata table. For this embodiment, the dataset table includes a first foreign key to the payload table, whereas the metadata table includes a second foreign key to the dataset table. The dataset table is monitored to determine a storage status of the contents, and the contents are subsequently published based on the storage status.

Description

Background technique [0001] By way of background, content storage devices provide storage for raw content that is subsequently transformed and displayed as web pages to website users after transformation processing via a server. For web pages where large amounts of content are served, by having multiple instances of the content storage device, read requests to the content storage device can be load balanced. For example, each instance may have the content of the entire set capable of handling a subset of the request load. [0002] While a multi-instance design can solve the read load problem for a website, it can create problems for adding, updating, or removing content. Conventionally, content publishers rely on file copying tools that have several undesirable limitations. For example, conventional content publishers do not guarantee data integrity across (across) instances of a content storage device (i.e., some content may be available on a particular instance of a storag...

AI Technical Summary

Problems solved by technology

For example, conventional content publishers do not guarantee data integrity across (across) instances of a content storage device (i.e., some content may be available on a particular instance of a storage device but not on other instances)

Furthermore, since conventional content publishers do not support parallel publishing, it is undesirable to process simultaneous publishing requests one request at a time, in an arbitrary order

For this reason, publishing times are often unpredictable as simultaneous publishing requests are processed arbitrarily one at a time

Also, if content is being updated, such content is not available for reading for the duration of the update

Images