How PrizmDoc Viewer
Load Balancing Works
Software Architect, PrizmDoc
If the people using your web application need to view, search, redact, or annotate documents
right in their browser, PrizmDoc Viewer is an amazing option. It lets you present Office, PDF,
TIFF, email, and many other kinds of documents as part of your web application. Check out
some of the demos if you've never seen it in action.
To make all of this possible, there are basically two sides to the PrizmDoc Viewer architecture:
1. The HTML viewer itself, running in the browser
2. A powerful backend which converts documents, page by page, to SVG for viewing in the
Your web server sits between these two, acting as a proxy for the viewer to ask the backend for
the pages it needs to display:
your web page
your web server
PrizmDoc Viewer Backend
(PrizmDoc Application Services)
One of the advantages of this architecture is that we can deliver the first page of the document
as soon as it's ready, even while the rest of the document is still being converted. However,
setting up and maintaining the backend is not trivial.
Fortunately, Accusoft can handle all of that for you with PrizmDoc Cloud. Sign up, get an API
key, and simply connect your web application to our already-running, fully-managed PrizmDoc
Viewer backend. It's a great option, especially if you're just getting started with PrizmDoc Viewer.
But, of course, using an Accusoft-hosted backend may not work for your business. Maybe you
are not allowed to ever let documents leave your network, even temporarily. If that's the case,
you'll need to host and manage the backend yourself. As customers start looking into what it
takes to do that, we get a lot of questions about how load balancing works. How is the compute
workload spread across the servers? How are HTTP requests routed to the correct machines?
What sort of load balancer(s) should I be using? Those