My filter-media job runs nightly (cron job owned by the user dspace), reporting success, but the PDF extracted text files that it claims to have created cannot be found the next morning.
When I re-run the command manually for some of these handles, it will create the extracted text files, and I can view the bitstreams in the item record immediately, and the contents show up in search results immediately.
Any ideas about what could be happening to the ones it thinks it created overnight?
Thanks!
~Amy