Don’t assume a CDN is safe or fast just because it’s an industry standard
A CDN (Content Delivery Network) refers to an external or hosted network used to load third-party JavaScript libraries, stylesheets, and assets (such as jQuery, DataTables, or Chart.js) required to render custom report interfaces.
If you build large amounts of WebReports in OpenText Content Server, chances are you pull in third party JS libraries like jQuery, DataTables, Chart.js and so on. And most of the time, just drop in a CDN link and move on. It’s quick and it works. But over time it comes with risks that do not get talked about enough.
The risks
- Internet dependency. Content Server is often hosted inside a corporate network, sometimes with restricted or no outbound internet access. If the server or the user’s browser can’t reach the CDN, the report breaks. Charts don’t load, tables lose their formatting, buttons stop working.
- CDN downtime. CDNs are reliable, but they are not immune to outages. When one goes down, every report depending on it goes down with it, and you have no control over when it comes back.
- Security risk. You are running someone else’s code inside your Content Server session. If a CDN gets compromised or the library file is tampered with, malicious JavaScript can execute in the context of a logged in user. That’s a potential supply chain attack. Unless you use Sub resource Integrity (SRI) hashes, you have no way to know the file hasn’t changed but even then it can be too late.
- Version changes. If you reference a latest version or a loosely pinned one, the library can change underneath. A report that worked fine last month can suddenly behave differently after a library update. Debugging this is painful because nothing changed on your side.
- Content Security Policy and firewall issues. Many organisations enforce CSP headers or block external domains at the firewall or proxy level. A CDN that works in dev might simply be blocked in production, and you only find out after going live.
- Compliance and privacy. Every time a user loads the report, their browser makes a request to a third-party server. That shares IP address and other request metadata with an external party. For organisations with strict data residency or privacy rules (GDPR for example), this can be a problem.
- Performance is not guaranteed. CDNs are usually fast, but not always from inside a corporate network. Proxies, VPNs and geography can make external loading slower than serving the file locally.
- Hard to maintain and audit. When CDN links are scattered across dozens of WebReports, nobody knows which library and which version is used where. Upgrading or patching a vulnerable library becomes a longer task and a difficult search for the areas to work on.
- Library removal or URL changes. Sometimes a CDN removes older versions or changes the URL structure. A report just stops working one day with no warning.
What we can do about it
The good news is that this is easy to fix, and the answer can be simple and convenient. There are two approaches that work well:
- Host the libraries inside Content Server itself. Store the JS files in the Support directory or in the Support Asset volume (works best if you are on OT cloud) and refer to them from your WebReport. You control the version, there’s no external dependency, and it works even in locked down environments.
- Keep the library code in a separate WebReport and call it as a sub WebReport. Put your common JS in one dedicated WebReport and call it from the main reports. This gives you one central place to maintain things, and when you need to upgrade a library you change it once and every report picks it up.
Both approaches give you control, stability and better security, and they make life easier when it’s time to upgrade or audit.
CDNs are fine for quick prototypes and POCs. But for anything that goes to production should always look to host things within Content Server.
