2020.2 Migration Guide
To operate on an Office file or an image, PSPDFKit Server needs to convert the file to a PDF during the document import process. For documents created by uploading a file and for documents with a remote URL, conversion happens completely transparently.
No matter the document source, PSPDFKit Server will store the PDF counterpart of the original file in its configured asset storage. This is particularly relevant if you plan to use documents with remote URLs that point to either Office files or images.
All subsequent actions (e.g. annotations or document operations) are applied to the PDF version of the document to guarantee compatibility with all features included in your license.
For this reason, it’s only possible to export a document in PDF format. However, you can download the originally uploaded file via the PDF download API endpoint using the
source=true query string parameter.
The conversion from Office to PDF is handled by a headless LibreOffice instance included in the PSPDFKit Server Docker image. Depending on the size and complexity of the file, the conversion process can take a significant amount of CPU and memory. We recommend testing your setup with a wide range of documents and using the monitoring facilities offered by your deployment solution to make sure the load is compatible with your current infrastructure.
Starting with PSPDFKit Server 2020.2, we recommend using PostgreSQL version 11 or higher.
We plan to guarantee compatibility with version 9.6 for the next two major releases (2020.3 and 2020.4).
The process of upgrading a PostgreSQL database to a newer version depends on the version you use and your infrastructure setup. We recommend looking at the PostgreSQL upgrade guides if you’re using a self-hosted deployment or checking the documentation of your cloud provider for the best course of action.
For a full list of changes, check out the Server changelog.
For more information, please take a look at the PSPDFKit for Web 2020.2 Migration Guide.