Over the past two weeks, I put together a “status page” for our library services. It’s not live yet, but I hope to make it public soon. Basically the page gathers HTTP response codes by submitting GET requests for each of our outside services (website, collections, and other services like ILLiad, Primo, and EZproxy), and displays them to the user, helpfully color-coded to show which ones have failed.

The point is that librarians and other users – on our campus and elsewhere – can see what is working and what is not. The use case may be as straightforward as asking “is the website down?”, or perhaps may be more complicated, such as debugging eresource access problems.

One interesting thing that I learned is that a lot of our services return 3xx series responses, which are redirects. I’m not sure why this is yet. Perhaps our vendors just move their pages around a lot. Or maybe the URLs we have stored in our CMS are just particularly old. In any case it is curious and I intend to investigate further. I have set Python’s httpx library to dig a bit deeper and follow redirects to see where they lead. One hopes they ultimately lead to a 2xx, but it is clear that sometimes they don’t :(

I also had to spoof some headers to avoid unwarranted 4xx series responses. Some vendors were returning 403s if it was clear that it wasn’t a browser making the request. So I used a Firefox user-agent header and, for the most part, got more reliable responses. Some vendors continue to give 403s nonetheless. I wonder what it is about my GET request that is causing them to continue to return “Forbidden”. I will puzzle over this some more. Once I get the whole thing in decent shape, I will share it with the CUNY eresources librarians.