The downloads page contains PDF multimedia components, so I suspect this is the issue. When setting up Solr to index PDFs there are a lot of additional configuration settings and dependent libraries to add in. Double check your configuration related to binaries in the [DXA docs][1], in particular for step 8. that the path to the binary extraction libraries (`contrib/extraction/lib`) matches that of your Solr installation.

  [1]: http://docs.sdl.com/LiveContent/content/en-US/SDL%20Tridion%20Reference%20Implementation-v1/GUID-09DB266E-ABC6-490A-82CF-2306DF0391F3